gtk2/gtk/gtksignal.c

1978 lines
48 KiB
C
Raw Normal View History

1997-11-24 22:37:52 +00:00
/* GTK - The GIMP Toolkit
* Copyright (C) 1995-1997 Peter Mattis, Spencer Kimball and Josh MacDonald
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
1997-11-24 22:37:52 +00:00
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public
* License along with this library; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 02111-1307, USA.
1997-11-24 22:37:52 +00:00
*/
#include <stdarg.h>
#include <string.h>
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
#include <stdio.h>
1997-11-24 22:37:52 +00:00
#include "gtksignal.h"
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
#include "gtkargcollector.c"
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
#define SIGNAL_BLOCK_SIZE (100)
#define HANDLER_BLOCK_SIZE (200)
#define EMISSION_BLOCK_SIZE (100)
#define DISCONNECT_INFO_BLOCK_SIZE (64)
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
#define MAX_SIGNAL_PARAMS (31)
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
enum
{
EMISSION_CONTINUE,
EMISSION_RESTART,
EMISSION_DONE
};
1997-11-24 22:37:52 +00:00
#define GTK_RUN_TYPE(x) ((x) & GTK_RUN_BOTH)
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
typedef struct _GtkSignal GtkSignal;
typedef struct _GtkSignalHash GtkSignalHash;
typedef struct _GtkHandler GtkHandler;
typedef struct _GtkEmission GtkEmission;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
typedef struct _GtkEmissionHookData GtkEmissionHookData;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
typedef struct _GtkDisconnectInfo GtkDisconnectInfo;
1997-11-24 22:37:52 +00:00
typedef void (*GtkSignalMarshaller0) (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer data);
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
struct _GtkSignal
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
GtkType object_type;
gchar *name;
guint function_offset;
GtkSignalMarshaller marshaller;
GtkType return_val;
guint signal_flags : 16;
guint nparams : 16;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkType *params;
GHookList *hook_list;
1997-11-24 22:37:52 +00:00
};
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
struct _GtkSignalHash
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkType object_type;
GQuark quark;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
1997-11-24 22:37:52 +00:00
};
struct _GtkHandler
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint id;
guint blocked : 20;
guint object_signal : 1;
guint after : 1;
guint no_marshal : 1;
guint16 ref_count;
guint16 signal_id;
GtkSignalFunc func;
gpointer func_data;
GtkSignalDestroy destroy_func;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkHandler *prev;
GtkHandler *next;
};
1997-11-24 22:37:52 +00:00
struct _GtkEmission
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkObject *object;
guint16 signal_id;
guint in_hook : 1;
GtkEmission *next;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
};
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
struct _GtkEmissionHookData
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkObject *object;
guint signal_id;
guint n_params;
GtkArg *params;
};
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
struct _GtkDisconnectInfo
{
GtkObject *object1;
guint disconnect_handler1;
guint signal_handler;
GtkObject *object2;
guint disconnect_handler2;
1997-11-24 22:37:52 +00:00
};
static guint gtk_signal_hash (gconstpointer h);
static gint gtk_signal_compare (gconstpointer h1,
gconstpointer h2);
1997-11-24 22:37:52 +00:00
static GtkHandler* gtk_signal_handler_new (void);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static void gtk_signal_handler_ref (GtkHandler *handler);
static void gtk_signal_handler_unref (GtkHandler *handler,
GtkObject *object);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static void gtk_signal_handler_insert (GtkObject *object,
1997-11-24 22:37:52 +00:00
GtkHandler *handler);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static void gtk_signal_real_emit (GtkObject *object,
guint signal_id,
GtkArg *params);
static GtkHandler* gtk_signal_get_handlers (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_type);
static guint gtk_signal_connect_by_type (GtkObject *object,
guint signal_id,
1997-11-24 22:37:52 +00:00
GtkSignalFunc func,
gpointer func_data,
GtkSignalDestroy destroy_func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gint object_signal,
gint after,
gint no_marshal);
static guint gtk_alive_disconnecter (GtkDisconnectInfo *info);
static GtkEmission* gtk_emission_new (void);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
static void gtk_emission_add (GtkEmission **emissions,
1997-11-24 22:37:52 +00:00
GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_type);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
static void gtk_emission_remove (GtkEmission **emissions,
1997-11-24 22:37:52 +00:00
GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_type);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
static gint gtk_emission_check (GtkEmission *emissions,
1997-11-24 22:37:52 +00:00
GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_type);
static gint gtk_handlers_run (GtkHandler *handlers,
GtkSignal *signal,
GtkObject *object,
GtkArg *params,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gint after);
static gboolean gtk_emission_hook_marshaller (GHook *hook,
gpointer data);
static gboolean gtk_signal_collect_params (GtkArg *params,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint nparams,
GtkType *param_types,
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
GtkType return_type,
va_list var_args);
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
#define LOOKUP_SIGNAL_ID(signal_id) ( \
signal_id > 0 && signal_id < gtk_n_signals ? \
(GtkSignal*) gtk_signals + signal_id : \
(GtkSignal*) 0 \
)
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static GtkSignalMarshal global_marshaller = NULL;
static GtkSignalDestroy global_destroy_notify = NULL;
1997-11-24 22:37:52 +00:00
static guint gtk_handler_id = 1;
static guint handler_quark = 0;
static GHashTable *gtk_signal_hash_table = NULL;
static GtkSignal *gtk_signals = NULL;
static guint gtk_n_signals = 0;
static GMemChunk *gtk_signal_hash_mem_chunk = NULL;
static GMemChunk *gtk_disconnect_info_mem_chunk = NULL;
static GtkHandler *gtk_handler_free_list = NULL;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
static GtkEmission *gtk_free_emissions = NULL;
1997-11-24 22:37:52 +00:00
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
static GtkEmission *current_emissions = NULL;
static GtkEmission *stop_emissions = NULL;
static GtkEmission *restart_emissions = NULL;
1997-11-24 22:37:52 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static GtkSignal*
gtk_signal_next_and_invalidate (void)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
static guint gtk_n_free_signals = 0;
register GtkSignal *signal;
register guint new_signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
/* don't keep *any* GtkSignal pointers across invokation of this function!!!
*/
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (gtk_n_free_signals == 0)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
register guint i;
register guint size;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
/* nearest pow
*/
size = gtk_n_signals + SIGNAL_BLOCK_SIZE;
size *= sizeof (GtkSignal);
i = 1;
while (i < size)
i <<= 1;
size = i;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signals = g_realloc (gtk_signals, size);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_n_free_signals = size / sizeof (GtkSignal) - gtk_n_signals;
memset (gtk_signals + gtk_n_signals, 0, gtk_n_free_signals * sizeof (GtkSignal));
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
new_signal_id = gtk_n_signals++;
gtk_n_free_signals--;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
g_assert (gtk_n_signals < 65535);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal = LOOKUP_SIGNAL_ID (new_signal_id);
if (signal)
signal->signal_id = new_signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return signal;
}
void
gtk_signal_init (void)
{
if (!handler_quark)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
{
GtkSignal *zero;
zero = gtk_signal_next_and_invalidate ();
g_assert (zero == NULL);
handler_quark = g_quark_from_static_string ("gtk-signal-handlers");
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_hash_mem_chunk =
g_mem_chunk_new ("GtkSignalHash mem chunk",
sizeof (GtkSignalHash),
sizeof (GtkSignalHash) * SIGNAL_BLOCK_SIZE,
G_ALLOC_ONLY);
gtk_disconnect_info_mem_chunk =
g_mem_chunk_new ("GtkDisconnectInfo mem chunk",
sizeof (GtkDisconnectInfo),
sizeof (GtkDisconnectInfo) * DISCONNECT_INFO_BLOCK_SIZE,
G_ALLOC_AND_FREE);
gtk_handler_free_list = NULL;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gtk_free_emissions = NULL;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_hash_table = g_hash_table_new (gtk_signal_hash,
gtk_signal_compare);
}
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint
gtk_signal_newv (const gchar *r_name,
new functions gtk_selection_data_copy and gtk_selection_data_free. Tue Jun 9 01:57:23 1998 Tim Janik <timj@gtk.org> * gtk/gtkselection.h: * gtk/gtkselection.c: new functions gtk_selection_data_copy and gtk_selection_data_free. * gtk/gtkwidget.c (gtk_widget_class_init): fixed gtk_signal_new() call for "selection_received", which was completely bogus. * other fixups to gtk_signal_new() calls all over the place. * gtk/gtktypebuiltins.h: types as variables (formerly macros). * gtk/gtktypebuiltins_vars.c: type variable implementations. * gtk/gtktypebuiltins_ids.c: array entries for builtin type declarations. * gtk/gtktypebuiltins_evals.c: enum value arrays. * gtk/gtk-boxed.defs: gtk and gdk structure definitions, used to build gtk.defs. * gtk/gtk.defs: generated file with scheme syntax for type definitions of gtk and gdk structures and enums. * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reworked type ids, so they are variables not macros anymore (this fixes binary incompatibility with new enum definitions). * gtk/gtkwidget.c (gtk_widget_real_key_press_event): proccess possible key bindings for this widget. * gtk/gtkwindow.c (gtk_window_key_press_event): chain parent class' handler. * gtk/gtkobject.h: * gtk/gtkobject.c: removed gtk_object_class_new_user_signal_no_recurse() again. new functions gtk_object_class_user_signal_new () and gtk_object_class_user_signal_newv (), to feature the GtkSignalRunType flag on the signal creation. Mon Jun 8 20:52:21 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: new signal GtkContainer::set_focus_child.
1998-06-09 07:11:55 +00:00
GtkSignalRunType signal_flags,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkType object_type,
guint function_offset,
GtkSignalMarshaller marshaller,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkType return_val,
guint nparams,
GtkType *params)
{
1997-11-24 22:37:52 +00:00
GtkSignal *signal;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkSignalHash *hash;
GQuark quark;
guint i;
gchar *name;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_val_if_fail (r_name != NULL, 0);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
g_return_val_if_fail (marshaller != NULL, 0);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
g_return_val_if_fail (nparams < MAX_SIGNAL_PARAMS, 0);
if (nparams)
g_return_val_if_fail (params != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (!handler_quark)
1997-11-24 22:37:52 +00:00
gtk_signal_init ();
name = g_strdup (r_name);
g_strdelimit (name, NULL, '_');
quark = gtk_signal_lookup (name, object_type);
if (quark)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_newv(): signal \"%s\" already exists in the `%s' class ancestry\n",
r_name,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_type_name (object_type));
g_free (name);
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (return_val != GTK_TYPE_NONE &&
new functions gtk_selection_data_copy and gtk_selection_data_free. Tue Jun 9 01:57:23 1998 Tim Janik <timj@gtk.org> * gtk/gtkselection.h: * gtk/gtkselection.c: new functions gtk_selection_data_copy and gtk_selection_data_free. * gtk/gtkwidget.c (gtk_widget_class_init): fixed gtk_signal_new() call for "selection_received", which was completely bogus. * other fixups to gtk_signal_new() calls all over the place. * gtk/gtktypebuiltins.h: types as variables (formerly macros). * gtk/gtktypebuiltins_vars.c: type variable implementations. * gtk/gtktypebuiltins_ids.c: array entries for builtin type declarations. * gtk/gtktypebuiltins_evals.c: enum value arrays. * gtk/gtk-boxed.defs: gtk and gdk structure definitions, used to build gtk.defs. * gtk/gtk.defs: generated file with scheme syntax for type definitions of gtk and gdk structures and enums. * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reworked type ids, so they are variables not macros anymore (this fixes binary incompatibility with new enum definitions). * gtk/gtkwidget.c (gtk_widget_real_key_press_event): proccess possible key bindings for this widget. * gtk/gtkwindow.c (gtk_window_key_press_event): chain parent class' handler. * gtk/gtkobject.h: * gtk/gtkobject.c: removed gtk_object_class_new_user_signal_no_recurse() again. new functions gtk_object_class_user_signal_new () and gtk_object_class_user_signal_newv (), to feature the GtkSignalRunType flag on the signal creation. Mon Jun 8 20:52:21 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: new signal GtkContainer::set_focus_child.
1998-06-09 07:11:55 +00:00
(signal_flags & GTK_RUN_BOTH) == GTK_RUN_FIRST)
{
Added a modular client-message-filter mechanism, that is used for the DND Sun Oct 18 18:16:39 1998 Owen Taylor <otaylor@gtk.org> * gdk/gdk.c gdkprivate.h: Added a modular client-message-filter mechanism, that is used for the DND messages. Removed all the old DND code. * gdk/gdkcolormap.c gdk/gdkcolormap.h: Add a function to get the visual of a given colormap. * gtk/gtkcolorsel.c: Conversion to new DND, drag a color-swatch. * gdk/gdk.h gdk/gdkdnd.c: The low-level X oriented portions of drag and drop protocols. Sending and receiving client messages, and navigating window trees. * gdk/gdkimage.c: added a gdk_flush() when destroying SHM images to hopefully make it more likely that X will gracefully handle the segment being destroyed. * gdk/gdkprivate.h gtk/gtkdebug.h: Add new DND debugging flags. * gtk/gtkeditable.[ch]: Updates for the selection handling changes. * gtk/gtkselection.[ch]: Added GtkTargetList, a refcounted data structure for keeping track of lists of GdkAtom + information. Removed selection_handler_add in favor of a "drag_data_get" signal. * gtk/gtkdnd.[ch] gtk/gtk.h: New files - highlevel (event loop dependent) parts of the DND protocols, display of drag icons, drag-under highlighting, and the "default handlers". * gtk/gtkinvisible.[ch]: New widget - InputOnly offscreen windows that are used for reliable pointer grabs and selection handling in the DND code. * gtk/testdnd.c: New test program for new DND. (Old DND tests in testgtk still need to be converted.) * gtk/testselection.c: Use the new selection API. * docs/dnd_internals: Start at describing how all the new code works inside. * docs/Changes-1.2.txt: New file describing source-incompatible changes in GTK+-1.2. Sat Oct 17 22:50:34 1998 Owen Taylor <otaylor@gtk.org> * gdk/gdkwindow.c (gdk_window_remove_filter): Free the right list node. * gdk/gdkwindow.c (gdk_window_init): Add gdk_root_parent to the XID table so we can receive events on it. Wed Oct 14 12:57:40 1998 Owen Taylor <otaylor@redhat.com> * gdk/gdk.c gdk/gdk.h (gdk_event_get_time): New function to get the timestamp from a generic event. Fri Oct 9 13:16:04 1998 Owen Taylor <otaylor@redhat.com> * gtk/gtkwidget.c (gtk_widget_add_events): Added function that safely adds additional events to a widget's event mask, even if the widget has previously been realized. (We can do this, but not remove events from the event mask). Fri Oct 2 17:35:35 1998 Owen Taylor <otaylor@redhat.com> * gdk/gdkproperty.c (gdk_property_get): Allow type == 0, for AnyPropertyType. Fri Oct 2 10:32:21 1998 Owen Taylor <otaylor@redhat.com> * gdk/gdkproperty.c (gdk_atom_intern): Add client-local hashing. Thu Sep 24 20:33:54 1998 Owen Taylor <otaylor@redhat.com> * gdk/gdk.c (gdk_event_send_clientmessage_toall): serial isn't a timestamp. Thu Sep 17 14:23:03 1998 Owen Taylor <otaylor@redhat.com> * gdk/gdk.c (gdk_event_translate): Removed printing of unknown window lookup warnings. (Made it a GDK_NOTE) - they happen in many circumstances.
1998-10-18 22:51:24 +00:00
g_warning ("gtk_signal_newv(): signal \"%s\" - return value `%s' incompatible with GTK_RUN_FIRST",
name, gtk_type_name (return_val));
g_free (name);
return 0;
}
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal = gtk_signal_next_and_invalidate ();
/* signal->signal_id already set */
signal->object_type = object_type;
signal->name = name;
1997-11-24 22:37:52 +00:00
signal->function_offset = function_offset;
signal->marshaller = marshaller;
signal->return_val = return_val;
signal->signal_flags = signal_flags;
1997-11-24 22:37:52 +00:00
signal->nparams = nparams;
signal->hook_list = NULL;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
if (nparams > 0)
{
signal->params = g_new (GtkType, nparams);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
for (i = 0; i < nparams; i++)
signal->params[i] = params[i];
1997-11-24 22:37:52 +00:00
}
else
signal->params = NULL;
/* insert "signal_name" into hash table
*/
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
hash = g_chunk_new (GtkSignalHash, gtk_signal_hash_mem_chunk);
hash->object_type = object_type;
hash->quark = g_quark_from_string (signal->name);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
hash->signal_id = signal->signal_id;
g_hash_table_insert (gtk_signal_hash_table, hash, GUINT_TO_POINTER (hash->signal_id));
/* insert "signal-name" into hash table
*/
g_strdelimit (signal->name, NULL, '-');
quark = g_quark_from_static_string (signal->name);
if (quark != hash->quark)
{
hash = g_chunk_new (GtkSignalHash, gtk_signal_hash_mem_chunk);
hash->object_type = object_type;
hash->quark = quark;
hash->signal_id = signal->signal_id;
g_hash_table_insert (gtk_signal_hash_table, hash, GUINT_TO_POINTER (hash->signal_id));
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return signal->signal_id;
}
guint
gtk_signal_new (const gchar *name,
new functions gtk_selection_data_copy and gtk_selection_data_free. Tue Jun 9 01:57:23 1998 Tim Janik <timj@gtk.org> * gtk/gtkselection.h: * gtk/gtkselection.c: new functions gtk_selection_data_copy and gtk_selection_data_free. * gtk/gtkwidget.c (gtk_widget_class_init): fixed gtk_signal_new() call for "selection_received", which was completely bogus. * other fixups to gtk_signal_new() calls all over the place. * gtk/gtktypebuiltins.h: types as variables (formerly macros). * gtk/gtktypebuiltins_vars.c: type variable implementations. * gtk/gtktypebuiltins_ids.c: array entries for builtin type declarations. * gtk/gtktypebuiltins_evals.c: enum value arrays. * gtk/gtk-boxed.defs: gtk and gdk structure definitions, used to build gtk.defs. * gtk/gtk.defs: generated file with scheme syntax for type definitions of gtk and gdk structures and enums. * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reworked type ids, so they are variables not macros anymore (this fixes binary incompatibility with new enum definitions). * gtk/gtkwidget.c (gtk_widget_real_key_press_event): proccess possible key bindings for this widget. * gtk/gtkwindow.c (gtk_window_key_press_event): chain parent class' handler. * gtk/gtkobject.h: * gtk/gtkobject.c: removed gtk_object_class_new_user_signal_no_recurse() again. new functions gtk_object_class_user_signal_new () and gtk_object_class_user_signal_newv (), to feature the GtkSignalRunType flag on the signal creation. Mon Jun 8 20:52:21 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: new signal GtkContainer::set_focus_child.
1998-06-09 07:11:55 +00:00
GtkSignalRunType signal_flags,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkType object_type,
guint function_offset,
GtkSignalMarshaller marshaller,
GtkType return_val,
guint nparams,
...)
{
GtkType *params;
guint i;
va_list args;
guint signal_id;
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
g_return_val_if_fail (nparams < MAX_SIGNAL_PARAMS, 0);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (nparams > 0)
{
params = g_new (GtkType, nparams);
va_start (args, nparams);
for (i = 0; i < nparams; i++)
params[i] = va_arg (args, GtkType);
va_end (args);
}
else
params = NULL;
signal_id = gtk_signal_newv (name,
new functions gtk_selection_data_copy and gtk_selection_data_free. Tue Jun 9 01:57:23 1998 Tim Janik <timj@gtk.org> * gtk/gtkselection.h: * gtk/gtkselection.c: new functions gtk_selection_data_copy and gtk_selection_data_free. * gtk/gtkwidget.c (gtk_widget_class_init): fixed gtk_signal_new() call for "selection_received", which was completely bogus. * other fixups to gtk_signal_new() calls all over the place. * gtk/gtktypebuiltins.h: types as variables (formerly macros). * gtk/gtktypebuiltins_vars.c: type variable implementations. * gtk/gtktypebuiltins_ids.c: array entries for builtin type declarations. * gtk/gtktypebuiltins_evals.c: enum value arrays. * gtk/gtk-boxed.defs: gtk and gdk structure definitions, used to build gtk.defs. * gtk/gtk.defs: generated file with scheme syntax for type definitions of gtk and gdk structures and enums. * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reworked type ids, so they are variables not macros anymore (this fixes binary incompatibility with new enum definitions). * gtk/gtkwidget.c (gtk_widget_real_key_press_event): proccess possible key bindings for this widget. * gtk/gtkwindow.c (gtk_window_key_press_event): chain parent class' handler. * gtk/gtkobject.h: * gtk/gtkobject.c: removed gtk_object_class_new_user_signal_no_recurse() again. new functions gtk_object_class_user_signal_new () and gtk_object_class_user_signal_newv (), to feature the GtkSignalRunType flag on the signal creation. Mon Jun 8 20:52:21 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: new signal GtkContainer::set_focus_child.
1998-06-09 07:11:55 +00:00
signal_flags,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
object_type,
function_offset,
marshaller,
return_val,
nparams,
params);
g_free (params);
return signal_id;
}
guint
gtk_signal_lookup (const gchar *name,
GtkType object_type)
{
GtkSignalHash hash;
GtkType lookup_type;
gpointer class = NULL;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
g_return_val_if_fail (name != NULL, 0);
g_return_val_if_fail (gtk_type_is_a (object_type, GTK_TYPE_OBJECT), 0);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
relookup:
lookup_type = object_type;
hash.quark = g_quark_try_string (name);
if (hash.quark)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
{
while (lookup_type)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
{
guint signal_id;
hash.object_type = lookup_type;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = GPOINTER_TO_UINT (g_hash_table_lookup (gtk_signal_hash_table, &hash));
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (signal_id)
return signal_id;
lookup_type = gtk_type_parent (lookup_type);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
}
}
if (!class)
{
class = gtk_type_class (object_type);
goto relookup;
}
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return 0;
1997-11-24 22:37:52 +00:00
}
GtkSignalQuery*
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
gtk_signal_query (guint signal_id)
{
GtkSignalQuery *query;
GtkSignal *signal;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_val_if_fail (signal_id >= 1, NULL);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal = LOOKUP_SIGNAL_ID (signal_id);
if (signal)
{
query = g_new (GtkSignalQuery, 1);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
query->object_type = signal->object_type;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
query->signal_id = signal_id;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
query->signal_name = signal->name;
query->is_user_signal = signal->function_offset == 0;
new functions gtk_selection_data_copy and gtk_selection_data_free. Tue Jun 9 01:57:23 1998 Tim Janik <timj@gtk.org> * gtk/gtkselection.h: * gtk/gtkselection.c: new functions gtk_selection_data_copy and gtk_selection_data_free. * gtk/gtkwidget.c (gtk_widget_class_init): fixed gtk_signal_new() call for "selection_received", which was completely bogus. * other fixups to gtk_signal_new() calls all over the place. * gtk/gtktypebuiltins.h: types as variables (formerly macros). * gtk/gtktypebuiltins_vars.c: type variable implementations. * gtk/gtktypebuiltins_ids.c: array entries for builtin type declarations. * gtk/gtktypebuiltins_evals.c: enum value arrays. * gtk/gtk-boxed.defs: gtk and gdk structure definitions, used to build gtk.defs. * gtk/gtk.defs: generated file with scheme syntax for type definitions of gtk and gdk structures and enums. * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reworked type ids, so they are variables not macros anymore (this fixes binary incompatibility with new enum definitions). * gtk/gtkwidget.c (gtk_widget_real_key_press_event): proccess possible key bindings for this widget. * gtk/gtkwindow.c (gtk_window_key_press_event): chain parent class' handler. * gtk/gtkobject.h: * gtk/gtkobject.c: removed gtk_object_class_new_user_signal_no_recurse() again. new functions gtk_object_class_user_signal_new () and gtk_object_class_user_signal_newv (), to feature the GtkSignalRunType flag on the signal creation. Mon Jun 8 20:52:21 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: new signal GtkContainer::set_focus_child.
1998-06-09 07:11:55 +00:00
query->signal_flags = signal->signal_flags;
query->return_val = signal->return_val;
query->nparams = signal->nparams;
query->params = signal->params;
}
else
query = NULL;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
return query;
}
1997-11-24 22:37:52 +00:00
gchar*
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
gtk_signal_name (guint signal_id)
1997-11-24 22:37:52 +00:00
{
GtkSignal *signal;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_val_if_fail (signal_id >= 1, NULL);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal = LOOKUP_SIGNAL_ID (signal_id);
1997-11-24 22:37:52 +00:00
if (signal)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return signal->name;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
return NULL;
}
void
gtk_signal_emitv (GtkObject *object,
guint signal_id,
GtkArg *params)
{
GtkSignal *signal;
g_return_if_fail (object != NULL);
g_return_if_fail (signal_id >= 1);
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_if_fail (signal != NULL);
g_return_if_fail (gtk_type_is_a (GTK_OBJECT_TYPE (object), signal->object_type));
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
if (signal->nparams > 0)
g_return_if_fail (params != NULL);
gtk_signal_real_emit (object, signal_id, params);
}
void
1997-11-24 22:37:52 +00:00
gtk_signal_emit (GtkObject *object,
guint signal_id,
1997-11-24 22:37:52 +00:00
...)
{
GtkSignal *signal;
va_list args;
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
GtkArg params[MAX_SIGNAL_PARAMS + 1];
gboolean abort;
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (signal_id >= 1);
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_if_fail (signal != NULL);
g_return_if_fail (gtk_type_is_a (GTK_OBJECT_TYPE (object), signal->object_type));
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
va_start (args, signal_id);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
abort = gtk_signal_collect_params (params,
signal->nparams,
signal->params,
signal->return_val,
args);
va_end (args);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
if (!abort)
gtk_signal_real_emit (object, signal_id, params);
}
void
gtk_signal_emitv_by_name (GtkObject *object,
const gchar *name,
GtkArg *params)
{
guint signal_id;
g_return_if_fail (object != NULL);
g_return_if_fail (name != NULL);
g_return_if_fail (params != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (signal_id >= 1)
{
GtkSignal *signal;
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_if_fail (signal != NULL);
g_return_if_fail (gtk_type_is_a (GTK_OBJECT_TYPE (object), signal->object_type));
gtk_signal_real_emit (object, signal_id, params);
}
else
{
g_warning ("gtk_signal_emitv_by_name(): could not find signal \"%s\" in the `%s' class ancestry",
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
}
1997-11-24 22:37:52 +00:00
}
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_emit_by_name (GtkObject *object,
const gchar *name,
1997-11-24 22:37:52 +00:00
...)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (name != NULL);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (signal_id >= 1)
1997-11-24 22:37:52 +00:00
{
GtkSignal *signal;
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
GtkArg params[MAX_SIGNAL_PARAMS + 1];
va_list args;
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
gboolean abort;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_if_fail (signal != NULL);
g_return_if_fail (gtk_type_is_a (GTK_OBJECT_TYPE (object), signal->object_type));
va_start (args, name);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
abort = gtk_signal_collect_params (params,
signal->nparams,
signal->params,
signal->return_val,
args);
1997-11-24 22:37:52 +00:00
va_end (args);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
if (!abort)
gtk_signal_real_emit (object, signal_id, params);
1997-11-24 22:37:52 +00:00
}
1998-03-08 10:41:48 +00:00
else
{
g_warning ("gtk_signal_emit_by_name(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1998-03-08 10:41:48 +00:00
}
1997-11-24 22:37:52 +00:00
}
void
gtk_signal_emit_stop (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gint state;
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (signal_id >= 1);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
state = gtk_emission_check (current_emissions, object, signal_id);
if (state > 1)
g_warning ("gtk_signal_emit_stop(): emission (%u) for object `%s' cannot be stopped from emission hook",
signal_id,
gtk_type_name (GTK_OBJECT_TYPE (object)));
else if (state)
{
if (!gtk_emission_check (stop_emissions, object, signal_id))
gtk_emission_add (&stop_emissions, object, signal_id);
}
1998-03-08 10:41:48 +00:00
else
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_warning ("gtk_signal_emit_stop(): no current emission (%u) for object `%s'",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
}
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_emit_stop_by_name (GtkObject *object,
1997-11-24 22:37:52 +00:00
const gchar *name)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
g_return_if_fail (name != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (signal_id)
gtk_signal_emit_stop (object, signal_id);
1998-03-08 10:41:48 +00:00
else
g_warning ("gtk_signal_emit_stop_by_name(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
}
guint
gtk_signal_n_emissions (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id)
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkEmission *emission;
guint n;
g_return_val_if_fail (object != NULL, 0);
g_return_val_if_fail (GTK_IS_OBJECT (object), 0);
n = 0;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
for (emission = current_emissions; emission; emission = emission->next)
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (emission->object == object && emission->signal_id == signal_id)
n++;
}
return n;
}
guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_n_emissions_by_name (GtkObject *object,
const gchar *name)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
guint n;
g_return_val_if_fail (object != NULL, 0);
g_return_val_if_fail (GTK_IS_OBJECT (object), 0);
g_return_val_if_fail (name != NULL, 0);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (signal_id)
n = gtk_signal_n_emissions (object, signal_id);
else
{
g_warning ("gtk_signal_n_emissions_by_name(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
n = 0;
}
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return n;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect (GtkObject *object,
const gchar *name,
1997-11-24 22:37:52 +00:00
GtkSignalFunc func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer func_data)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
g_return_val_if_fail (GTK_IS_OBJECT (object), 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (!signal_id)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_connect(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id,
1997-11-24 22:37:52 +00:00
func, func_data, NULL,
FALSE, FALSE, FALSE);
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect_after (GtkObject *object,
const gchar *name,
GtkSignalFunc func,
gpointer func_data)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (!signal_id)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_connect_after(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id,
1997-11-24 22:37:52 +00:00
func, func_data, NULL,
FALSE, TRUE, FALSE);
1997-11-24 22:37:52 +00:00
}
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint
gtk_signal_connect_full (GtkObject *object,
const gchar *name,
GtkSignalFunc func,
GtkCallbackMarshal marshal,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer func_data,
GtkDestroyNotify destroy_func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gint object_signal,
gint after)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (!signal_id)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_connect_full(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (marshal)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id, (GtkSignalFunc) marshal,
func_data, destroy_func,
object_signal, after, TRUE);
else
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id, func,
func_data, destroy_func,
object_signal, after, FALSE);
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect_object (GtkObject *object,
const gchar *name,
1997-11-24 22:37:52 +00:00
GtkSignalFunc func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkObject *slot_object)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
/* slot_object needs to be treated as ordinary pointer
*/
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (!signal_id)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_connect_object(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id,
1997-11-24 22:37:52 +00:00
func, slot_object, NULL,
TRUE, FALSE, FALSE);
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint
1997-11-24 22:37:52 +00:00
gtk_signal_connect_object_after (GtkObject *object,
const gchar *name,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkSignalFunc func,
1997-11-24 22:37:52 +00:00
GtkObject *slot_object)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id = gtk_signal_lookup (name, GTK_OBJECT_TYPE (object));
if (!signal_id)
1997-11-24 22:37:52 +00:00
{
1998-03-08 10:41:48 +00:00
g_warning ("gtk_signal_connect_object_after(): could not find signal \"%s\" in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
name,
gtk_type_name (GTK_OBJECT_TYPE (object)));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return gtk_signal_connect_by_type (object, signal_id,
1997-11-24 22:37:52 +00:00
func, slot_object, NULL,
TRUE, TRUE, FALSE);
1997-11-24 22:37:52 +00:00
}
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect_while_alive (GtkObject *object,
const gchar *signal,
GtkSignalFunc func,
gpointer func_data,
GtkObject *alive_object)
{
GtkDisconnectInfo *info;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (object != NULL);
g_return_if_fail (GTK_IS_OBJECT (object));
g_return_if_fail (signal != NULL);
g_return_if_fail (func != NULL);
g_return_if_fail (alive_object != NULL);
g_return_if_fail (GTK_IS_OBJECT (alive_object));
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
info = g_chunk_new (GtkDisconnectInfo, gtk_disconnect_info_mem_chunk);
info->object1 = object;
info->object2 = alive_object;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
info->signal_handler = gtk_signal_connect (object, signal, func, func_data);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
info->disconnect_handler1 =
gtk_signal_connect_object (info->object1,
"destroy",
GTK_SIGNAL_FUNC (gtk_alive_disconnecter),
(GtkObject*) info);
info->disconnect_handler2 =
gtk_signal_connect_object (info->object2,
"destroy",
GTK_SIGNAL_FUNC (gtk_alive_disconnecter),
(GtkObject*) info);
}
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect_object_while_alive (GtkObject *object,
const gchar *signal,
GtkSignalFunc func,
GtkObject *alive_object)
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
{
GtkDisconnectInfo *info;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
g_return_if_fail (object != NULL);
g_return_if_fail (GTK_IS_OBJECT (object));
g_return_if_fail (signal != NULL);
g_return_if_fail (func != NULL);
g_return_if_fail (alive_object != NULL);
g_return_if_fail (GTK_IS_OBJECT (alive_object));
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
info = g_chunk_new (GtkDisconnectInfo, gtk_disconnect_info_mem_chunk);
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
info->object1 = object;
info->object2 = alive_object;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
info->signal_handler = gtk_signal_connect_object (object, signal, func, alive_object);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
info->disconnect_handler1 =
gtk_signal_connect_object (info->object1,
"destroy",
GTK_SIGNAL_FUNC (gtk_alive_disconnecter),
(GtkObject*) info);
info->disconnect_handler2 =
gtk_signal_connect_object (info->object2,
"destroy",
GTK_SIGNAL_FUNC (gtk_alive_disconnecter),
(GtkObject*) info);
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
}
1997-11-24 22:37:52 +00:00
void
gtk_signal_disconnect (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint handler_id)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (handler_id > 0);
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (handler->id == handler_id)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler->id = 0;
handler->blocked += 1;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
gtk_signal_handler_unref (handler, object);
1997-11-24 22:37:52 +00:00
return;
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler->next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_warning ("gtk_signal_disconnect(): could not find handler (%u)", handler_id);
1997-11-24 22:37:52 +00:00
}
void
gtk_signal_disconnect_by_func (GtkObject *object,
GtkSignalFunc func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer data)
{
GtkHandler *handler;
gint found_one;
g_return_if_fail (object != NULL);
g_return_if_fail (func != NULL);
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
while (handler)
{
GtkHandler *handler_next;
handler_next = handler->next;
if ((handler->id > 0) &&
(handler->func == func) &&
(handler->func_data == data))
{
found_one = TRUE;
handler->id = 0;
handler->blocked += 1;
gtk_signal_handler_unref (handler, object);
}
handler = handler_next;
}
if (!found_one)
g_warning ("gtk_signal_disconnect_by_func(): could not find handler (0x%0lX) containing data (0x%0lX)", (long) func, (long) data);
}
1997-11-24 22:37:52 +00:00
void
gtk_signal_disconnect_by_data (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer data)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler;
1997-11-24 22:37:52 +00:00
gint found_one;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler_next;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler_next = handler->next;
if ((handler->id > 0) &&
(handler->func_data == data))
1997-11-24 22:37:52 +00:00
{
found_one = TRUE;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler->id = 0;
handler->blocked += 1;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
gtk_signal_handler_unref (handler, object);
1997-11-24 22:37:52 +00:00
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler_next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
if (!found_one)
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
g_warning ("gtk_signal_disconnect_by_data(): could not find handler containing data (0x%0lX)", (long) data);
1997-11-24 22:37:52 +00:00
}
void
gtk_signal_handler_block (GtkObject *object,
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint handler_id)
1997-11-24 22:37:52 +00:00
{
GtkHandler *handler;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (handler_id > 0);
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
if (handler->id == handler_id)
1997-11-24 22:37:52 +00:00
{
handler->blocked += 1;
1997-11-24 22:37:52 +00:00
return;
}
handler = handler->next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_warning ("gtk_signal_handler_block(): could not find handler (%u)", handler_id);
1997-11-24 22:37:52 +00:00
}
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_handler_block_by_func (GtkObject *object,
GtkSignalFunc func,
gpointer data)
{
GtkHandler *handler;
gint found_one;
g_return_if_fail (object != NULL);
g_return_if_fail (func != NULL);
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
while (handler)
{
if ((handler->id > 0) &&
(handler->func == func) &&
(handler->func_data == data))
{
found_one = TRUE;
handler->blocked += 1;
}
handler = handler->next;
}
if (!found_one)
g_warning ("gtk_signal_handler_block_by_func(): could not find handler (0x%0lX) containing data (0x%0lX)", (long) func, (long) data);
}
1997-11-24 22:37:52 +00:00
void
gtk_signal_handler_block_by_data (GtkObject *object,
gpointer data)
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler;
1997-11-24 22:37:52 +00:00
gint found_one;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
if ((handler->id > 0) &&
(handler->func_data == data))
1997-11-24 22:37:52 +00:00
{
found_one = TRUE;
handler->blocked += 1;
1997-11-24 22:37:52 +00:00
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler->next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
if (!found_one)
GTK_RESIZE_NEEDED is a private flag now. Mon Feb 2 04:15:08 1998 Tim Janik <timj@gimp.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: GTK_RESIZE_NEEDED is a private flag now. (gtk_container_register_toplevel): new function. (gtk_container_unregister_toplevel): new function. * gtk/gtkmain.c: GTK_LEAVE_PENDING is a private flag now. * gtk/gtkmenu.c: call gtk_container_register_toplevel in gtk_menu_class_init instead of this dirty gtk_widget_set_parent(,NULL) hack. new default handler gtk_menu_destroy for calling gtk_container_unregister_toplevel. removed GTK_ANCHORED, GTK_UNMAPPED. * gtk/gtkobject.h: macro cleanups, added GTK_DESTROYED flag. * gtk/gtkobject.c: only emit DESTROY signal if !GTK_OBJECT_DESTROYED (object). * gtk/gtkprivate.h: new file that will not be automatically included. it holds the private flags for GtkWidget along with it's SET/UNSET and examination macros. * gtk/gtkwidget.c: private flags: GTK_RESIZE_NEEDED, GTK_REDRAW_PENDING, GTK_RESIZE_PENDING, GTK_IN_REPARENT, GTK_USER_STYLE. GTK_ANCHORED is replaced by GTK_TOPLEVEL. added missing UNSET for GTK_IN_REPARENT. removed the gtk_widget_set_parent(, NULL) hack for toplevels. upon destroy free memory for widgets with GTK_WIDGET_HAS_SHAPE_MASK. * gtk/gtkwidget.h: split up the widget flags into a public and a private portion. added an extra field private_flags to GtkWidget without making it bigger by using an alignment gap of 16 bit. macro cleanups. * gtk/gtkwindow.c: removed GTK_ANCHORED. new function gtk_window_destroy for calling gtk_container_unregister_toplevel. removed the gtk_widget_set_parent(,NULL), call gtk_container_register_toplevel instead. remove GTK_UNMAPPED. GTK_RESIZE_NEEDED is private now. * gtk/gtksignal.c (gtk_signal_disconnect): removed a bug on removal that cut off the handler list -> living_objects == 0 with testgtk. made some warnings more descriptive. new function gtk_signal_connect_object_while_alive, which will automatically destroy the connection once one of the objects is destroyed. didn't include this before removal of the above mentioned bug. * reflected refcounting revolution in ChangeLog
1998-02-02 04:54:25 +00:00
g_warning ("gtk_signal_handler_block_by_data(): could not find handler containing data (0x%0lX)", (long) data);
1997-11-24 22:37:52 +00:00
}
void
gtk_signal_handler_unblock (GtkObject *object,
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint handler_id)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_if_fail (handler_id > 0);
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
if (handler->id == handler_id)
1997-11-24 22:37:52 +00:00
{
if (handler->blocked > 0)
handler->blocked -= 1;
else
g_warning ("gtk_signal_handler_unblock(): handler (%u) is not blocked", handler_id);
1997-11-24 22:37:52 +00:00
return;
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler->next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_warning ("gtk_signal_handler_unblock(): could not find handler (%u)", handler_id);
1997-11-24 22:37:52 +00:00
}
void
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_handler_unblock_by_func (GtkObject *object,
GtkSignalFunc func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer data)
{
GtkHandler *handler;
gint found_one;
g_return_if_fail (object != NULL);
g_return_if_fail (func != NULL);
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
while (handler)
{
if ((handler->id > 0) &&
(handler->func == func) &&
(handler->func_data == data) &&
(handler->blocked > 0))
{
handler->blocked -= 1;
found_one = TRUE;
}
handler = handler->next;
}
if (!found_one)
g_warning ("gtk_signal_handler_unblock_by_func(): could not find blocked handler (0x%0lX) containing data (0x%0lX)", (long) func, (long) data);
}
1997-11-24 22:37:52 +00:00
void
gtk_signal_handler_unblock_by_data (GtkObject *object,
gpointer data)
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handler;
1997-11-24 22:37:52 +00:00
gint found_one;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
found_one = FALSE;
handler = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (handler)
1997-11-24 22:37:52 +00:00
{
if ((handler->id > 0) &&
(handler->func_data == data) &&
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
(handler->blocked > 0))
1997-11-24 22:37:52 +00:00
{
handler->blocked -= 1;
1997-11-24 22:37:52 +00:00
found_one = TRUE;
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler->next;
1997-11-24 22:37:52 +00:00
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
if (!found_one)
g_warning ("gtk_signal_handler_unblock_by_data(): could not find blocked handler containing data (0x%0lX)", (long) data);
1997-11-24 22:37:52 +00:00
}
void
gtk_signal_handlers_destroy (GtkObject *object)
{
GtkHandler *handler;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
/* we make the "optimization" of destroying the first handler in the last
* place, since we don't want gtk_signal_handler_unref() to reset the objects
* handler_key data on each removal
*/
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
handler = gtk_object_get_data_by_id (object, handler_quark);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (handler)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = handler->next;
while (handler)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *next;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
next = handler->next;
if (handler->id > 0)
gtk_signal_handler_unref (handler, object);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler = next;
1997-11-24 22:37:52 +00:00
}
handler = gtk_object_get_data_by_id (object, handler_quark);
if (handler->id > 0)
gtk_signal_handler_unref (handler, object);
1997-11-24 22:37:52 +00:00
}
}
void
gtk_signal_set_funcs (GtkSignalMarshal marshal_func,
GtkSignalDestroy destroy_func)
{
global_marshaller = marshal_func;
global_destroy_notify = destroy_func;
1997-11-24 22:37:52 +00:00
}
static guint
gtk_signal_hash (gconstpointer h)
1997-11-24 22:37:52 +00:00
{
Removed g_object_pointer_hash, which was just g_direct_hash. Tue Jun 9 18:44:57 1998 Owen Taylor <otaylor@gtk.org> * gtk/gtkobject.c: Removed g_object_pointer_hash, which was just g_direct_hash. Mon May 25 19:54:20 1998 Owen Taylor <otaylor@gtk.org> * configure.in: x_libs=, not $x_libs=. Enough said. (Case only hit for --disable-xshm) Mon May 25 12:08:14 1998 Owen Taylor <otaylor@gtk.org> * configure.in (LDFLAGS): Add to $CFLAGS and $LDFLAGS when testing for X libraries, don't replace them. Because the user might have specified the path to the X libraries themself before running configure. * examples/**.c: Changed all gpointer * to gpointer Sat May 23 21:54:05 1998 Owen Taylor <otaylor@gtk.org> * configure.in (LDFLAGS): Bomb out with a moderately helpful message if detection of X libraries fails. Sat May 23 18:57:06 1998 Owen Taylor <otaylor@gtk.org> [ Combination of: gtk-rrh-980412-0.patch (Raja R Harinath <harinath@cs.umn.edu>) gtk-jbuhler-980516-0 (Jeremy Buhler <jbuhler@cs.washington.edu>) ] * gdk/gdk.h gdk/gdkcc.c gdk/gdkfont.c gtk/gtkmain.c gtk/gtksignal.c gtk/gtktext.c: Fixups for warnings from adding const to type of GHashFunc, GCompareFunc * gtk/gtkcombo.c (gtk_combo_entry_key_press): Minor style/ ansi-warnings fixups. Tue Jun 9 17:47:33 1998 Owen Taylor <otaylor@gtk.org> * glib.h: Remove #error - HP/UX. Sat May 23 19:00:01 1998 Owen Taylor <otaylor@gtk.org> [ Combination of: gtk-rrh-980412-0.patch (Raja R Harinath <harinath@cs.umn.edu>) gtk-jbuhler-980516-0 (Jeremy Buhler <jbuhler@cs.washington.edu>) ] * glib.h ghash.c gstring.c gdataset.c gutils.c: - Added new typedef g_const_pointer; expunged all incorrect uses of 'const gpointer'. - Fixed up warnings that that created, - Changed GHashFunc and GCompareFunc to take g_const_pointer arguments. (Necessary, but will cause warnings in existing code until fixed) - Added other new const in harmless positions.
1998-06-09 23:18:11 +00:00
register const GtkSignalHash *hash = h;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return hash->object_type ^ hash->quark;
1997-11-24 22:37:52 +00:00
}
static gint
gtk_signal_compare (gconstpointer h1,
gconstpointer h2)
1997-11-24 22:37:52 +00:00
{
Removed g_object_pointer_hash, which was just g_direct_hash. Tue Jun 9 18:44:57 1998 Owen Taylor <otaylor@gtk.org> * gtk/gtkobject.c: Removed g_object_pointer_hash, which was just g_direct_hash. Mon May 25 19:54:20 1998 Owen Taylor <otaylor@gtk.org> * configure.in: x_libs=, not $x_libs=. Enough said. (Case only hit for --disable-xshm) Mon May 25 12:08:14 1998 Owen Taylor <otaylor@gtk.org> * configure.in (LDFLAGS): Add to $CFLAGS and $LDFLAGS when testing for X libraries, don't replace them. Because the user might have specified the path to the X libraries themself before running configure. * examples/**.c: Changed all gpointer * to gpointer Sat May 23 21:54:05 1998 Owen Taylor <otaylor@gtk.org> * configure.in (LDFLAGS): Bomb out with a moderately helpful message if detection of X libraries fails. Sat May 23 18:57:06 1998 Owen Taylor <otaylor@gtk.org> [ Combination of: gtk-rrh-980412-0.patch (Raja R Harinath <harinath@cs.umn.edu>) gtk-jbuhler-980516-0 (Jeremy Buhler <jbuhler@cs.washington.edu>) ] * gdk/gdk.h gdk/gdkcc.c gdk/gdkfont.c gtk/gtkmain.c gtk/gtksignal.c gtk/gtktext.c: Fixups for warnings from adding const to type of GHashFunc, GCompareFunc * gtk/gtkcombo.c (gtk_combo_entry_key_press): Minor style/ ansi-warnings fixups. Tue Jun 9 17:47:33 1998 Owen Taylor <otaylor@gtk.org> * glib.h: Remove #error - HP/UX. Sat May 23 19:00:01 1998 Owen Taylor <otaylor@gtk.org> [ Combination of: gtk-rrh-980412-0.patch (Raja R Harinath <harinath@cs.umn.edu>) gtk-jbuhler-980516-0 (Jeremy Buhler <jbuhler@cs.washington.edu>) ] * glib.h ghash.c gstring.c gdataset.c gutils.c: - Added new typedef g_const_pointer; expunged all incorrect uses of 'const gpointer'. - Fixed up warnings that that created, - Changed GHashFunc and GCompareFunc to take g_const_pointer arguments. (Necessary, but will cause warnings in existing code until fixed) - Added other new const in harmless positions.
1998-06-09 23:18:11 +00:00
register const GtkSignalHash *hash1 = h1;
register const GtkSignalHash *hash2 = h2;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return (hash1->quark == hash2->quark &&
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
hash1->object_type == hash2->object_type);
1997-11-24 22:37:52 +00:00
}
static guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_alive_disconnecter (GtkDisconnectInfo *info)
1997-11-24 22:37:52 +00:00
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
g_return_val_if_fail (info != NULL, 0);
gtk_signal_disconnect (info->object1, info->disconnect_handler1);
gtk_signal_disconnect (info->object1, info->signal_handler);
gtk_signal_disconnect (info->object2, info->disconnect_handler2);
g_mem_chunk_free (gtk_disconnect_info_mem_chunk, info);
return 0;
1997-11-24 22:37:52 +00:00
}
static GtkHandler*
configure.in acheader.h gdk/gdkwindow.c Check for Shape extension both on Sun May 3 13:38:22 1998 Owen Taylor <otaylor@gtk.org> * configure.in acheader.h gdk/gdkwindow.c Check for Shape extension both on the client and server side. (And, more importantly, check for the shape extension so we may include -lXext even when compiling with --disable-xshm) Don't set override_redirect on all shaped windows. It isn't necessary. * gdk/gdkwindow.c: Set ->colormap to NULL for root and foreign windows. Use this to check if we need to get the colormap from X. Fri May 1 22:32:47 1998 Owen Taylor <otaylor@gtk.org> * gtk/gtkbutton.c (gtk_button_paint): Draw the areas between the default and the button always in GTK_STATE_NORMAL. * gtk/gtkrange.c (gtk_range_style_set): Added a style_set callback. Fri May 1 16:40:57 1998 Owen Taylor <otaylor@gtk.org> * gdk/gdkpixmap.c (gdk_pixmap_colormap_create_from_xpmp[_d]): Fix a buffer overflow on pixmaps that claim to have more than 31 characters per pixel. (gdk_pixmap_read_string): Don't wrap around strings longer than half of address space ;-) * gtk/gtk[vh]ruler.c gtk/gtkinputdialog.c: Expand some buffers that were used for printing integers. * */* (almost): Style: All int foo () { ... } changed to int foo (void) { ... } ^^^^^^^ This is why some many files changed Even where there were proper prototypes elsewhere. * gdk/gxid.c (handle_claim_device): Some extra checks. It isn't safe against being fed bad X id's, but at least it should be safe against deleting all your files.
1998-05-03 22:41:32 +00:00
gtk_signal_handler_new (void)
1997-11-24 22:37:52 +00:00
{
GtkHandler *handler;
if (!gtk_handler_free_list)
{
GtkHandler *handler_block;
guint i;
handler_block = g_new0 (GtkHandler, HANDLER_BLOCK_SIZE);
for (i = 1; i < HANDLER_BLOCK_SIZE; i++)
{
(handler_block + i)->next = gtk_handler_free_list;
gtk_handler_free_list = (handler_block + i);
}
handler = handler_block;
}
else
{
handler = gtk_handler_free_list;
gtk_handler_free_list = handler->next;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
handler->id = 0;
handler->blocked = 0;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
handler->signal_id = 0;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler->object_signal = FALSE;
1997-11-24 22:37:52 +00:00
handler->after = FALSE;
handler->no_marshal = FALSE;
handler->ref_count = 1;
1997-11-24 22:37:52 +00:00
handler->func = NULL;
handler->func_data = NULL;
handler->destroy_func = NULL;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler->prev = NULL;
1997-11-24 22:37:52 +00:00
handler->next = NULL;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
return handler;
}
static void
gtk_signal_handler_ref (GtkHandler *handler)
1997-11-24 22:37:52 +00:00
{
handler->ref_count += 1;
}
static void
gtk_signal_handler_unref (GtkHandler *handler,
GtkObject *object)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (!handler->ref_count)
{
/* FIXME: i wanna get removed somewhen */
g_warning ("gtk_signal_handler_unref(): handler with ref_count==0!");
return;
}
handler->ref_count -= 1;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (handler->ref_count == 0)
{
if (handler->destroy_func)
(* handler->destroy_func) (handler->func_data);
else if (!handler->func && global_destroy_notify)
(* global_destroy_notify) (handler->func_data);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (handler->prev)
handler->prev->next = handler->next;
else if (handler->next)
gtk_object_set_data_by_id (object, handler_quark, handler->next);
else
{
GTK_OBJECT_UNSET_FLAGS (object, GTK_CONNECTED);
gtk_object_set_data_by_id (object, handler_quark, NULL);
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (handler->next)
handler->next->prev = handler->prev;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
handler->next = gtk_handler_free_list;
gtk_handler_free_list = handler;
}
1997-11-24 22:37:52 +00:00
}
static void
gtk_signal_handler_insert (GtkObject *object,
GtkHandler *handler)
{
GtkHandler *tmp;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
/* FIXME: remove */ g_assert (handler->next == NULL);
/* FIXME: remove */ g_assert (handler->prev == NULL);
tmp = gtk_object_get_data_by_id (object, handler_quark);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (!tmp)
{
GTK_OBJECT_SET_FLAGS (object, GTK_CONNECTED);
gtk_object_set_data_by_id (object, handler_quark, handler);
}
1997-11-24 22:37:52 +00:00
else
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
while (tmp)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (tmp->signal_id < handler->signal_id)
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
{
if (tmp->prev)
{
tmp->prev->next = handler;
handler->prev = tmp->prev;
}
else
gtk_object_set_data_by_id (object, handler_quark, handler);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
tmp->prev = handler;
handler->next = tmp;
break;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (!tmp->next)
{
tmp->next = handler;
handler->prev = tmp;
break;
}
tmp = tmp->next;
}
1997-11-24 22:37:52 +00:00
}
#ifdef G_ENABLE_DEBUG
/* value typically set via gdb */
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
static GtkObject *gtk_trace_signal_object = NULL;
#endif /* G_ENABLE_DEBUG */
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
static void
1997-11-24 22:37:52 +00:00
gtk_signal_real_emit (GtkObject *object,
guint signal_id,
GtkArg *params)
1997-11-24 22:37:52 +00:00
{
GtkSignal signal;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
GtkHandler *handlers;
GtkSignalFunc signal_func;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkEmission *emission;
/* gtk_handlers_run() expects a reentrant GtkSignal*, so we allocate
* it locally on the stack. we save some lookups ourselves with this as well.
*/
signal = *LOOKUP_SIGNAL_ID (signal_id);
if (signal.function_offset)
signal_func = G_STRUCT_MEMBER (GtkSignalFunc, object->klass, signal.function_offset);
else
signal_func = NULL;
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
#ifdef G_ENABLE_DEBUG
if (gtk_debug_flags & GTK_DEBUG_SIGNALS ||
object == gtk_trace_signal_object)
g_message ("%s::%s emitted (object=%p class-method=%p)\n",
gtk_type_name (GTK_OBJECT_TYPE (object)),
signal.name,
object,
signal_func);
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
#endif /* G_ENABLE_DEBUG */
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (signal.signal_flags & GTK_RUN_NO_RECURSE)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gint state;
state = gtk_emission_check (current_emissions, object, signal_id);
if (state)
{
if (state > 1)
g_warning ("gtk_signal_real_emit(): emission (%u) for object `%s' cannot be restarted from emission hook",
signal_id,
gtk_type_name (GTK_OBJECT_TYPE (object)));
else if (!gtk_emission_check (restart_emissions, object, signal_id))
gtk_emission_add (&restart_emissions, object, signal_id);
return;
}
1997-11-24 22:37:52 +00:00
}
gtk_object_ref (object);
gtk_emission_add (&current_emissions, object, signal_id);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission = current_emissions;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
emission_restart:
if (signal.signal_flags & GTK_RUN_FIRST && signal_func)
{
signal.marshaller (object, signal_func, NULL, params);
if (stop_emissions && gtk_emission_check (stop_emissions, object, signal_id))
{
gtk_emission_remove (&stop_emissions, object, signal_id);
goto emission_done;
}
else if (restart_emissions &&
signal.signal_flags & GTK_RUN_NO_RECURSE &&
gtk_emission_check (restart_emissions, object, signal_id))
{
gtk_emission_remove (&restart_emissions, object, signal_id);
goto emission_restart;
}
}
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (signal.hook_list && !GTK_OBJECT_DESTROYED (object))
{
GtkEmissionHookData data;
data.object = object;
data.n_params = signal.nparams;
data.params = params;
data.signal_id = signal_id;
emission->in_hook = 1;
g_hook_list_marshal_check (signal.hook_list, TRUE, gtk_emission_hook_marshaller, &data);
emission->in_hook = 0;
}
if (GTK_OBJECT_CONNECTED (object))
{
handlers = gtk_signal_get_handlers (object, signal_id);
if (handlers)
{
gint return_val;
return_val = gtk_handlers_run (handlers, &signal, object, params, FALSE);
switch (return_val)
{
case EMISSION_CONTINUE:
break;
case EMISSION_RESTART:
goto emission_restart;
case EMISSION_DONE:
goto emission_done;
}
}
}
if (signal.signal_flags & GTK_RUN_LAST && signal_func)
{
signal.marshaller (object, signal_func, NULL, params);
if (stop_emissions && gtk_emission_check (stop_emissions, object, signal_id))
{
gtk_emission_remove (&stop_emissions, object, signal_id);
goto emission_done;
}
else if (restart_emissions &&
signal.signal_flags & GTK_RUN_NO_RECURSE &&
gtk_emission_check (restart_emissions, object, signal_id))
{
gtk_emission_remove (&restart_emissions, object, signal_id);
goto emission_restart;
}
}
if (GTK_OBJECT_CONNECTED (object))
{
handlers = gtk_signal_get_handlers (object, signal_id);
if (handlers)
{
gint return_val;
return_val = gtk_handlers_run (handlers, &signal, object, params, TRUE);
switch (return_val)
{
case EMISSION_CONTINUE:
break;
case EMISSION_RESTART:
goto emission_restart;
case EMISSION_DONE:
goto emission_done;
}
}
}
emission_done:
if (restart_emissions && signal.signal_flags & GTK_RUN_NO_RECURSE)
gtk_emission_remove (&restart_emissions, object, signal_id);
gtk_emission_remove (&current_emissions, object, signal_id);
gtk_object_unref (object);
1997-11-24 22:37:52 +00:00
}
static GtkHandler*
1997-11-24 22:37:52 +00:00
gtk_signal_get_handlers (GtkObject *object,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
guint signal_id)
1997-11-24 22:37:52 +00:00
{
GtkHandler *handlers;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
handlers = gtk_object_get_data_by_id (object, handler_quark);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
while (handlers)
{
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (handlers->signal_id == signal_id)
1997-11-24 22:37:52 +00:00
return handlers;
handlers = handlers->next;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
return NULL;
}
guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_handler_pending (GtkObject *object,
guint signal_id,
gboolean may_be_blocked)
{
GtkHandler *handlers;
guint handler_id;
g_return_val_if_fail (object != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_return_val_if_fail (signal_id >= 1, 0);
if (GTK_OBJECT_CONNECTED (object))
handlers = gtk_signal_get_handlers (object, signal_id);
else
return 0;
handler_id = 0;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
while (handlers && handlers->signal_id == signal_id)
{
if (handlers->id > 0 &&
yeppers, accelerator changes to fix the gimp. commit message dedicated to yeppers, accelerator changes to fix the gimp. commit message dedicated to sopwith ;) Thu Jun 18 03:30:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new function gtk_accel_label_accelerator_width to request the size of the accelerator portion of an accel label. (gtk_accel_label_size_request): don't request for the accelerators size. (gtk_accel_label_expose_event): only draw the accelerator if we got enough extra space. * gtk/gtkmenuitem.c (gtk_menu_item_size_request): request accelerator width from children. * gtk/gtkmenu.c (gtk_menu_key_press): when adding an accelerator to an object (after removal has been requested) check if there is still an accelerator remaining to avoid adding two accelerators on an object. this can happen for locked accelerators (or accelerator-frozen widgets). (gtk_menu_size_request): feature childrens accelerator width in size requests. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): use gtk_widget_freeze_accelerators() for dynamically created menu items. * gtk/gtksignal.h: * gtk/gtksignal.c: new function gtk_signal_handler_pending_by_func() which will return a handler_id > 0 if the specified function is pending for `signal_id'. * gtk/gtkwidget.h: * gtk/gtkwidget.c: remove gtk_widget_stop_accelerator, which was just a signal handler function to stop accelerator addition. added gtk_widget_freeze_accelerators and gtk_widget_thaw_accelerators which will prevent (undo) any accelerators from being added to or removed from a widget.
1998-06-18 03:22:09 +00:00
(may_be_blocked || handlers->blocked == FALSE))
{
handler_id = handlers->id;
break;
}
handlers = handlers->next;
}
return handler_id;
}
guint
gtk_signal_handler_pending_by_func (GtkObject *object,
guint signal_id,
gboolean may_be_blocked,
GtkSignalFunc func,
gpointer data)
{
GtkHandler *handlers;
guint handler_id;
g_return_val_if_fail (object != NULL, 0);
g_return_val_if_fail (func != NULL, 0);
g_return_val_if_fail (signal_id >= 1, 0);
if (GTK_OBJECT_CONNECTED (object))
handlers = gtk_signal_get_handlers (object, signal_id);
else
return 0;
handler_id = 0;
while (handlers && handlers->signal_id == signal_id)
{
if (handlers->id > 0 &&
handlers->func == func &&
handlers->func_data == data &&
(may_be_blocked || handlers->blocked == FALSE))
{
handler_id = handlers->id;
break;
}
handlers = handlers->next;
}
return handler_id;
}
guint
gtk_signal_add_emission_hook (guint signal_id,
GtkEmissionHook hook_func,
gpointer data)
{
return gtk_signal_add_emission_hook_full (signal_id, hook_func, data, NULL);
}
guint
gtk_signal_add_emission_hook_full (guint signal_id,
GtkEmissionHook hook_func,
gpointer data,
GDestroyNotify destroy)
{
static guint seq_hook_id = 1;
GtkSignal *signal;
GHook *hook;
g_return_val_if_fail (signal_id > 0, 0);
g_return_val_if_fail (hook_func != NULL, 0);
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_val_if_fail (signal != NULL, 0);
if (signal->signal_flags & GTK_RUN_NO_HOOKS)
{
g_warning ("gtk_signal_add_emission_hook_full(): signal \"%s\" does not support emission hooks",
signal->name);
return 0;
}
if (!signal->hook_list)
{
signal->hook_list = g_new (GHookList, 1);
g_hook_list_init (signal->hook_list, sizeof (GHook));
}
hook = g_hook_alloc (signal->hook_list);
hook->data = data;
hook->func = hook_func;
hook->destroy = destroy;
signal->hook_list->seq_id = seq_hook_id;
g_hook_prepend (signal->hook_list, hook);
seq_hook_id = signal->hook_list->seq_id;
return hook->hook_id;
}
void
gtk_signal_remove_emission_hook (guint signal_id,
guint hook_id)
{
GtkSignal *signal;
g_return_if_fail (signal_id > 0);
g_return_if_fail (hook_id > 0);
signal = LOOKUP_SIGNAL_ID (signal_id);
g_return_if_fail (signal != NULL);
if (!signal->hook_list || !g_hook_destroy (signal->hook_list, hook_id))
g_warning ("gtk_signal_remove_emission_hook(): could not find hook (%u)", hook_id);
}
static gboolean
gtk_emission_hook_marshaller (GHook *hook,
gpointer data_p)
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkEmissionHookData *data = data_p;
GtkEmissionHook func;
func = hook->func;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (!GTK_OBJECT_DESTROYED (data->object))
return func (data->object, data->signal_id,
data->n_params, data->params,
hook->data);
else
return TRUE;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
static guint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_signal_connect_by_type (GtkObject *object,
guint signal_id,
1997-11-24 22:37:52 +00:00
GtkSignalFunc func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gpointer func_data,
1997-11-24 22:37:52 +00:00
GtkSignalDestroy destroy_func,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gint object_signal,
gint after,
gint no_marshal)
1997-11-24 22:37:52 +00:00
{
GtkObjectClass *class;
1997-11-24 22:37:52 +00:00
GtkHandler *handler;
gint found_it;
GtkSignal *signal;
1997-11-24 22:37:52 +00:00
g_return_val_if_fail (object != NULL, 0);
g_return_val_if_fail (object->klass != NULL, 0);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
signal = LOOKUP_SIGNAL_ID (signal_id);
1997-11-24 22:37:52 +00:00
/* Search through the signals for this object and make
* sure the one we are adding is valid. We need to perform
* the lookup on the objects parents as well. If it isn't
* valid then issue a warning and return.
fixed an assertment. Sat Jun 6 06:01:24 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c (gtk_signal_emitv): fixed an assertment. * gtk/makeenums.awk: a script to generate the GtkEnumValue arrays from, this should eventually be done by gentypeinfo.el somewhen. * gtk/gtkenumvalues.c: new generated file to hold GtkEnumValue arrays. * gtk/gtktypeutils.h: new function gtk_enum_values() to retrive all the enum values of an enum type. * gtk/gtk.defs: * gtk/gtkcurve.h: * gtk/gtkobject.h: * gtk/gtkprivate.h: * gtk/gtkwidget.h: * gtk/gtkenums.h: brought enum/flags definitions in sync, added a few more enum definitions for bindings and pattern matching. * some more macro and GtkType fixups in various places. * gdk/gdktypes.h (enum): added a new value GDK_AFTER_MASK, which is used as a key-release modifier for the binding system. Fri Jun 5 06:06:06 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenu.h (struct _GtkMenu): removed GList*children, since it was a stale list pointer that is already present in GtkMenuShell. * gtk/gtkmenushell.h (struct _GtkMenuShellClass): added a signal GtkMenuShell::selection_done which is emitted after the menu shell poped down again and all possible menu items have been activated. Thu Jun 4 02:20:42 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenushell.c (gtk_menu_shell_button_release): flush the x-queue before activation of the menuitem, so the menu is actually taken off the screen prior to any menu item activation. * gtk/gtkctree.c (gtk_ctree_get_row_data): allow function invokation for NULL nodes. * gtk/gtkwidget.h: * gtk/gtkwidget.c: new function gtk_widget_stop_accelerator to stop the emission of the "add-accelerator" signal on a widget. this is usefull to prevent accelerator installation on certain widgets. * gtk/gtknotebook.c (gtk_notebook_menu_item_create): keep the menu labels left justified, by setting their alignment. stop accelerator installation for the menu items, since we use dynamic menus. Wed Jun 3 06:41:22 1998 Tim Janik <timj@gtk.org> * gtk/gtkmenufactory.c: adaptions to use the new accel groups. people should *really* use GtkItemFactory. this is only for preserving source compatibility where possible, use of GtkMenuFactory is deprecated as of now. * gtk/gtkobject.h (gtk_object_class_add_user_signal): new function to create user signals of type GTK_RUN_NO_RECURSE. don't know why i missed this possibility when i added gtk_object_class_add_user_signal in late january. * gtk/gtkmain.c (gtk_init): ignore subsequent function calls. Sun May 31 07:31:09 1998 Tim Janik <timj@gtk.org> * gtk/gtkaccelgroup.h: * gtk/gtkaccelgroup.c: new implementation of the accelerator concept. * gtk/gtkaccellabel.h: * gtk/gtkaccellabel.c: new widget derived from GtkLabel whitch features display of the accelerators associated with a certain widget. * gtk/gtkitemfactory.h: * gtk/gtkitemfactory.c: new widget, item factory with automatic rc parsing and accelerator handling. * gtk/gtkmenu.c (gtk_menu_reposition): new function to care for positioning a menu. (gtk_menu_map): removed the allocation code. (gtk_menu_size_allocate): care for redrawing of children and resize our widget->window correctly. (gtk_menu_key_press): feature the new accelerator groups. * gtk/gtkmenuitem.c (gtk_menu_item_size_allocate): reposition the submenu if neccessary. * gtk/gtkmenuitem.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkradiomenuitem.c: use GtkAccelLabel in the *_new_with_label() function variants. * gdk/gdk.c: (gdk_keyval_from_name): (gdk_keyval_name): new functions for keyval<->key-name associations. (gdk_keyval_to_upper): (gdk_keyval_to_lower): (gdk_keyval_is_upper): (gdk_keyval_is_lower): new functions to check/translate keyvalues with regards to their cases. Wed May 27 00:48:10 1998 Tim Janik <timj@gtk.org> * gtk/gtkwidget.c (gtk_widget_class_path): new function to calculate a widget's class path. (gtk_widget_path): new function to calculate a widget's name path. * gtk/gtkrc.c: newly introduced GtkPatternSpec structures to speed up pattern matching, features reversed pattern matches.
1998-06-07 06:48:56 +00:00
* As of now (1998-05-27) this lookup shouldn't be neccessarry
* anymore since gtk_signal_lookup() has been reworked to only
* return correct signal ids per class-branch.
1997-11-24 22:37:52 +00:00
*/
found_it = FALSE;
class = object->klass;
while (class)
{
GtkType parent;
guint *object_signals;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
guint nsignals;
guint i;
object_signals = class->signals;
nsignals = class->nsignals;
for (i = 0; i < nsignals; i++)
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
if (object_signals[i] == signal_id)
{
found_it = TRUE;
break;
}
parent = gtk_type_parent (class->type);
if (parent)
class = gtk_type_class (parent);
else
class = NULL;
}
1997-11-24 22:37:52 +00:00
if (!found_it)
{
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
g_warning ("gtk_signal_connect_by_type(): could not find signal id (%u) in the `%s' class ancestry",
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
signal_id,
gtk_type_name (object->klass->type));
1997-11-24 22:37:52 +00:00
return 0;
}
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
handler = gtk_signal_handler_new ();
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
handler->id = gtk_handler_id++;
handler->signal_id = signal_id;
1997-11-24 22:37:52 +00:00
handler->object_signal = object_signal;
handler->func = func;
handler->func_data = func_data;
handler->destroy_func = destroy_func;
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handler->after = after != FALSE;
1997-11-24 22:37:52 +00:00
handler->no_marshal = no_marshal;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
gtk_signal_handler_insert (object, handler);
return handler->id;
}
static GtkEmission*
configure.in acheader.h gdk/gdkwindow.c Check for Shape extension both on Sun May 3 13:38:22 1998 Owen Taylor <otaylor@gtk.org> * configure.in acheader.h gdk/gdkwindow.c Check for Shape extension both on the client and server side. (And, more importantly, check for the shape extension so we may include -lXext even when compiling with --disable-xshm) Don't set override_redirect on all shaped windows. It isn't necessary. * gdk/gdkwindow.c: Set ->colormap to NULL for root and foreign windows. Use this to check if we need to get the colormap from X. Fri May 1 22:32:47 1998 Owen Taylor <otaylor@gtk.org> * gtk/gtkbutton.c (gtk_button_paint): Draw the areas between the default and the button always in GTK_STATE_NORMAL. * gtk/gtkrange.c (gtk_range_style_set): Added a style_set callback. Fri May 1 16:40:57 1998 Owen Taylor <otaylor@gtk.org> * gdk/gdkpixmap.c (gdk_pixmap_colormap_create_from_xpmp[_d]): Fix a buffer overflow on pixmaps that claim to have more than 31 characters per pixel. (gdk_pixmap_read_string): Don't wrap around strings longer than half of address space ;-) * gtk/gtk[vh]ruler.c gtk/gtkinputdialog.c: Expand some buffers that were used for printing integers. * */* (almost): Style: All int foo () { ... } changed to int foo (void) { ... } ^^^^^^^ This is why some many files changed Even where there were proper prototypes elsewhere. * gdk/gxid.c (handle_claim_device): Some extra checks. It isn't safe against being fed bad X id's, but at least it should be safe against deleting all your files.
1998-05-03 22:41:32 +00:00
gtk_emission_new (void)
1997-11-24 22:37:52 +00:00
{
GtkEmission *emission;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (!gtk_free_emissions)
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkEmission *emission_block;
guint i;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission_block = g_new0 (GtkEmission, EMISSION_BLOCK_SIZE);
for (i = 1; i < EMISSION_BLOCK_SIZE; i++)
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
(emission_block + i)->next = gtk_free_emissions;
gtk_free_emissions = (emission_block + i);
}
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission = emission_block;
}
else
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission = gtk_free_emissions;
gtk_free_emissions = emission->next;
}
1997-11-24 22:37:52 +00:00
emission->object = NULL;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
emission->signal_id = 0;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission->in_hook = 0;
emission->next = NULL;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
return emission;
}
static void
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gtk_emission_add (GtkEmission **emissions,
GtkObject *object,
guint signal_id)
1997-11-24 22:37:52 +00:00
{
GtkEmission *emission;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (emissions != NULL);
g_return_if_fail (object != NULL);
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
emission = gtk_emission_new ();
emission->object = object;
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
emission->signal_id = signal_id;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission->next = *emissions;
*emissions = emission;
1997-11-24 22:37:52 +00:00
}
static void
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gtk_emission_remove (GtkEmission **emissions,
GtkObject *object,
guint signal_id)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
GtkEmission *emission, *last;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
1997-11-24 22:37:52 +00:00
g_return_if_fail (emissions != NULL);
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
last = NULL;
emission = *emissions;
while (emission)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (emission->object == object && emission->signal_id == signal_id)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (last)
last->next = emission->next;
else
*emissions = emission->next;
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
emission->next = gtk_free_emissions;
gtk_free_emissions = emission;
1997-11-24 22:37:52 +00:00
break;
}
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
last = emission;
emission = last->next;
1997-11-24 22:37:52 +00:00
}
}
static gint
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
gtk_emission_check (GtkEmission *emission,
GtkObject *object,
guint signal_id)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
while (emission)
1997-11-24 22:37:52 +00:00
{
reverted marius change to expose the type systems internal type info data Mon Dec 7 03:08:39 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h: * gtk/gtktypeutils.c: reverted marius change to expose the type systems internal type info data to the user. if such functionality is required we should provide wrapped accessors, ala gtk_signal_query(). * gtk/gtksignal.c (gtk_signal_connect_by_type): reverted marius change, since it destroys the possibility to implement automatic marshaller lookups some day, and it also disables third party code's ability to connect to any signal. also the GtkTypeInfo structures are dynamically allocated memory portions, so only the type system is really allowed to access that stuff. Mon Dec 7 01:32:18 1998 Tim Janik <timj@gtk.org> * gtk/gtkfilesel.c (gtk_file_selection_key_press): always intercept the Tab key on the entry. the focus shouldn't get lost even if completion is attempted from an empty entry, since an empty entry string does indeed have a valid completion meaning (complete all). (gtk_file_selection_init): cast the gchar array parameter in calls to gtk_clist_new_with_titles() to quit compiler warnings. (check_dir): the no_stat_dirs struct must not be const, since we do indeed modify its contents. * gtk/testgtk.c (event_watcher): adapted prototype to fit new emission hook semantics. * gtk/gtksignal.h: * gtk/gtksignal.c: changed emission allocation, so we don't use a doubly linked list but link ourselfs (singly linked). changed emission hooks, they get the emision parameters passed as well now and are emitted during the actuall signal emission (after the RUN_FIRST class method, but prior to RUN_FIRST handlers). the existing restrictions do still apply to signal emission hooks, i.e. an emission may not be stopped or restarted from an emission hook. due to possibly huge perfomance impacts, frequent use of emision hooks is also not recommended. (gtk_signal_next_and_invalidate): added an assertments which explicits what the code assumes anyways: a maximum amount of 65535 signals. * gtk/gtkcontainer.h: deprecated gtk_container_foreach_interp(), gtk_container_foreach_full() should be used instead. * gtk/gtkmain.h: deprecated gtk_timeout_add_interp and gtk_idle_add_interp, since we provide _full variants. * gtk/gtksignal.h: deprecated gtk_signal_connect_interp(), we provide gtk_signal_connect_full() for long enough now.
1998-12-07 02:31:19 +00:00
if (emission->object == object && emission->signal_id == signal_id)
return 1 + emission->in_hook;
emission = emission->next;
1997-11-24 22:37:52 +00:00
}
return FALSE;
}
static gint
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gtk_handlers_run (GtkHandler *handlers,
GtkSignal *signal,
GtkObject *object,
GtkArg *params,
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
gint after)
1997-11-24 22:37:52 +00:00
{
/* *signal is a local copy on the stack of gtk_signal_real_emit(),
* so we don't need to look it up every time we invoked a function.
*/
while (handlers && handlers->signal_id == signal->signal_id)
1997-11-24 22:37:52 +00:00
{
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
GtkHandler *handlers_next;
made the <widget>_signals[] arrays of type guint rather than gint. made Mon Mar 9 15:48:10 1998 Tim Janik <timj@gimp.org> * Signal signedness and naming corrections, plus GtkType fixes: * gtk/gtkadjustment.c: * gtk/gtkbutton.c: * gtk/gtkcheckmenuitem.c: * gtk/gtkclist.c: * gtk/gtkcolorsel.c: * gtk/gtkcontainer.c: * gtk/gtkcurve.c: * gtk/gtkdata.c: * gtk/gtkeditable.c: * gtk/gtkentry.c: * gtk/gtkhandlebox.c: * gtk/gtkinputdialog.c: * gtk/gtkitem.c: * gtk/gtklist.c: * gtk/gtkmenuitem.c: * gtk/gtkmenushell.c: * gtk/gtknotebook.c: * gtk/gtkstatusbar.c: * gtk/gtktoolbar.c: * gtk/gtktree.c: * gtk/gtktreeitem.c: * gtk/gtkwidget.c: * gtk/gtktogglebutton.c: * gtk/gtkwindow.c: made the <widget>_signals[] arrays of type guint rather than gint. * gtk/gtkwidget.c (gtk_widget_get_ancestor): made widget_type a GtkType. * gtk/gtkcombo.h: handler ids need to be of type guint (entry_change_id, list_change_id). * gtk/gtkaccelerator.c: changed signal_num to signal_id and typed it guint. * gtk/gtkmain.c: made gtk_ndebug_keys a guint. * gtk/gtkmenu.h: * gtk/gtkmenu.c: (gtk_menu_popup): made button a guint. (gtk_menu_set_active): made index a guint. * gtk/gtkmenuitem.h: * gtk/gtkmenuitem.c: made accelerator_signal a guint. * gtk/gtkoptionmenu.h: * gtk/gtkoptionmenu.c: (gtk_option_menu_set_history): made index a guint. * gtk/gtksignal.h: * gtk/gtksignal.c: * gtk/gtkobject.h: * gtk/gtkobject.c: changed a bunch of prototypes to take guints rather than gints. also made some conversions from guint to GtkType, left over from when the fundamental-types system was introduced. * gtk/gtkobject.h: * gtk/gtkobject.c: made object_data_id_index and obj_count guints. made *signals and nsignals guints in GtkObjectClass.
1998-03-09 15:16:28 +00:00
gtk_signal_handler_ref (handlers);
if (!handlers->blocked && handlers->after == after)
1997-11-24 22:37:52 +00:00
{
if (handlers->func)
{
if (handlers->no_marshal)
(* (GtkCallbackMarshal) handlers->func) (object,
handlers->func_data,
signal->nparams,
params);
1997-11-24 22:37:52 +00:00
else if (handlers->object_signal)
/* don't cast with GTK_OBJECT () */
(* signal->marshaller) ((GtkObject*) handlers->func_data,
handlers->func,
object,
params);
1997-11-24 22:37:52 +00:00
else
(* signal->marshaller) (object,
handlers->func,
handlers->func_data,
params);
1997-11-24 22:37:52 +00:00
}
else if (global_marshaller)
(* global_marshaller) (object,
handlers->func_data,
signal->nparams,
params,
signal->params,
signal->return_val);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
if (stop_emissions && gtk_emission_check (stop_emissions,
object,
signal->signal_id))
1997-11-24 22:37:52 +00:00
{
gtk_emission_remove (&stop_emissions, object, signal->signal_id);
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
gtk_signal_handler_unref (handlers, object);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return EMISSION_DONE;
1997-11-24 22:37:52 +00:00
}
else if (restart_emissions &&
signal->signal_flags & GTK_RUN_NO_RECURSE &&
gtk_emission_check (restart_emissions, object, signal->signal_id))
1997-11-24 22:37:52 +00:00
{
gtk_emission_remove (&restart_emissions, object, signal->signal_id);
gtk_signal_handler_unref (handlers, object);
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return EMISSION_RESTART;
1997-11-24 22:37:52 +00:00
}
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
handlers_next = handlers->next;
gtk_signal_handler_unref (handlers, object);
handlers = handlers_next;
1997-11-24 22:37:52 +00:00
}
ok, there have been several severe bugs in the signal handler referencing Tue Feb 10 07:12:07 1998 Tim Janik <timj@gimp.org> * gtk/gtksignal.h: * gtk/gtksignal.c: ok, there have been several severe bugs in the signal handler referencing and ->next connection stuff. these bugs caused invokations of handlers that are disconnected and - worse - destroyed already. invokation of *destroyd* handlers mean: anything can be executed , because the handler structure can just as well be realocated. at the cost of an extra ->prev field per handler we should have a reasonable stable system now, because of the various places that can cause a handler to be disconnected (*any* handler invokation can cause *any* or *all* handlers to be disconnected, there is no way around a doubly linked list, actually handler disconnection has never worked correctly because of this. handlers are connected together via a *doubly* linked list now, and it is *not* valid to remove a handler out of this list untill all its references have been droped, i.e. handler->ref_count==0. to prevent emissions of disconnected but still referenced handlers, disconnected handlers are simply marked as blocked and get an id of 0 which is an invalid signal handler id. the handler->id has been changed to have 28 significant bits (using alignment gaps), since 65536 (old range: guint16) signal connections (as a total) can easily be reached by complex applications. this whole handler thingy is at least as tedious as writing doubly linked list implementations ;)
1998-02-10 06:53:08 +00:00
reimplemented the signal storage system to use a linear array rather than Wed May 6 02:16:34 1998 Tim Janik <timj@gtk.org> * gtk/gtksignal.c: reimplemented the signal storage system to use a linear array rather than a hash table. be carefull *not* to keep any GtkSignal pointers across invokations of gtk_signal_next_and_invalidate() and therefore gtk_signal_new[v](). general code cleanups, made all allocations through memchunks. (gtk_signal_lookup): we now do the lookup of signals through a key_id for the signal names to avoid multiple hashing of the signal name on lookups for the several hirarchy levels of an object. Tue May 5 15:33:27 1998 Tim Janik <timj@gtk.org> * gtk/gtkmain.c (gtk_init): added gtk_signal_init() to avoid repeatedly checking for signal initialization in gtksignal.c. * gtk/gtktypeutils.c (TYPE_NODES_BLOCK_SIZE): reseted this to 200 after a long debugging period ;) Mon May 4 00:30:11 1998 Tim Janik <timj@gtk.org> * gtk/gtktypeutils.h (GTK_TYPE_IS_A): new macro to speedup gtk_type_is_a(). * gtk/gtktypeutils.c: reimplemented the type storage system to use a linear array rather than a hash table. it actually speeded up testgtk for a *considerable* amount. be carefull *not* to keep any GtkTypeNode pointers across invokations of gtk_type_node_next_and_invalidate() and therefore gtk_type_unique()! * gtk/gtkobject.h (GTK_IS_OBJECT_CLASS): new macro to test for the inheritance of a class pointer. (GTK_OBJECT): modifications for speedups. (GTK_IS_OBJECT): likewise. (GTK_OBJECT_CLASS): likewise. (suggested by owen). * gtk/gtkcontainer.h (GTK_IS_CONTAINER_CLASS): new macro. * gtk/gtkwidget.h (GTK_IS_WIDGET_CLASS): new macro. * gtk/gtk.defs (GtkTooltips): define GtkTooltips as object not as boxed, since its derivation changed 4 month ago ;(. showed up with the unique type name checking code in gtk_type_unique(). * random guint->GtkType and macro fixups.
1998-05-06 01:43:56 +00:00
return EMISSION_CONTINUE;
1997-11-24 22:37:52 +00:00
}
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
static gboolean
gtk_signal_collect_params (GtkArg *params,
guint n_params,
GtkType *param_types,
GtkType return_type,
va_list var_args)
1997-11-24 22:37:52 +00:00
{
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
register GtkArg *last_param;
register gboolean failed = FALSE;
for (last_param = params + n_params; params < last_param; params++)
1997-11-24 22:37:52 +00:00
{
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
register gchar *error;
params->name = NULL;
params->type = *(param_types++);
GTK_ARG_COLLECT_VALUE (params,
var_args,
error);
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
if (error)
1997-11-24 22:37:52 +00:00
{
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
failed = TRUE;
g_warning ("gtk_signal_collect_params(): %s", error);
g_free (error);
1997-11-24 22:37:52 +00:00
}
}
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
params->type = return_type;
params->name = NULL;
return_type = GTK_FUNDAMENTAL_TYPE (return_type);
if (return_type != GTK_TYPE_NONE)
1997-11-24 22:37:52 +00:00
{
if ((return_type >= GTK_TYPE_FLAT_FIRST &&
return_type <= GTK_TYPE_FLAT_LAST) ||
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
(return_type == GTK_TYPE_OBJECT))
{
GTK_VALUE_POINTER (*params) = va_arg (var_args, gpointer);
if (GTK_VALUE_POINTER (*params) == NULL)
{
failed = TRUE;
g_warning ("gtk_signal_collect_params(): invalid NULL pointer for return argument type `%s'",
gtk_type_name (params->type));
}
}
else
{
failed = TRUE;
g_warning ("gtk_signal_collect_params(): unsupported return argument type `%s'",
gtk_type_name (params->type));
}
1997-11-24 22:37:52 +00:00
}
deleted most of the argument handling code, since that is now implemented Fri Jul 10 00:02:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkcontainer.h: * gtk/gtkcontainer.c: deleted most of the argument handling code, since that is now implemented in gtkarg.c. similar to gtk_object_args_collect, we now export a new function gtk_container_child_args_collect(). for consistency with the object arguments, a few functions got renamed: gtk_container_child_arg_get -> gtk_container_child_get, gtk_container_child_arg_set -> gtk_container_child_set, gtk_container_child_arg_getv -> gtk_container_child_getv, gtk_container_child_arg_setv -> gtk_container_child_setv, gtk_container_add_with_argv -> gtk_container_addv. note, gtk_container_add_with_args() remained, because its equivalent would be gtk_container_add(,, ...) which would break all existing code. (gtk_container_add_child_arg_type): similar to gtk_object_add_arg_type, we expect the `arg_name' argument to be a const static string now. (gtk_container_get_child_arg_type): function removed. Thu Jul 9 07:03:04 1998 Tim Janik <timj@gtk.org> * gtk/gtkargcollector.c: new file which holds gtk_arg_collect_value(). this is a static inline function that collects command line arguments from a va_list. this file can just be included in all places that need this functionality. * gtk/gtkarg.h: * gtk/gtkarg.c: new files which implement most of the argument handling stuff from gtkobject.c. also collected a few more gtk_arg_* utility functions from else places. * gtk/gtkobject.h: * gtk/gtkobject.c: moved most of the argument handling code into gtkarg.c. we now export gtk_object_args_collect() as a non-public method with a blind va_list pointer. (gtk_object_add_arg_type): the `arg_name' argument is required to be a const static string now. (gtk_object_get_arg_type): function got removed. * gtk/gtkwidget.c: (gtk_widget_set): (gtk_widget_new): adaptions for gtk_object_args_collect(). * gtk/gtktypeutils.c (gtk_type_init_builtin_types): changed the internal fundamental type name so as to have a valid prefix, e.g. "bool"-> "gboolean", "string"->"GtkString" and somesuch, left "void" as is, though that should probably be something like GtkNone since the type itself is called GTK_TYPE_NONE. even the internal type names need to avoid name clashes and must live in their own namespace, several code portions rely on that. we should relly have typedefs such as typedef gchar* GtkString; so the fundamental type names can be used for code dumpers just like with all the Gtk/Gdk types.
1998-07-09 23:00:34 +00:00
else
GTK_VALUE_POINTER (*params) = NULL;
return failed;
1997-11-24 22:37:52 +00:00
}