AuroraMiddleware/gtk2
1
0
Fork 0
forked from AuroraMiddleware/gtk
Code Issues Pull Requests Projects Releases Wiki Activity
b84d37af4d
gtk2/gtk/gtkstyleproperty.c

3157 lines
102 KiB
C
Raw Normal View History

css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
/* GTK - The GIMP Toolkit
* Copyright (C) 2010 Carlos Garnacho <carlosg@gnome.org>
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser 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
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser 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.
*/
#include "config.h"
css: Rename gtkcssstrfuncs file to gtkstyleproperty The idea is to handle the style property registering etc in this file instead of the styleproperties file. The next commits will move those files.
2011-05-21 01:30:36 +00:00
#include "gtkstylepropertyprivate.h"
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
#include <errno.h>
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
#include <math.h>
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
#include <stdlib.h>
#include <string.h>
#include <gdk-pixbuf/gdk-pixbuf.h>
#include <cairo-gobject.h>
#include "gtkcssprovider.h"
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
#include "gtkcssparserprivate.h"
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
#include "gtkcsstypesprivate.h"
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
/* the actual parsers we have */
#include "gtkanimationdescription.h"
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
#include "gtkbindings.h"
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
#include "gtkborderimageprivate.h"
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
#include "gtkgradient.h"
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
#include "gtkshadowprivate.h"
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
#include "gtkthemingengine.h"
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
#include "gtktypebuiltins.h"
Add initial cut at win32 theme support for CSS We now support -gtk-win32-theme-part(class,part,state) in background and border-image CSS properties. This renders the corresponding theme part using DrawThemeBackground() and acts as a base for a CSS based windows theme. Note that we build the parsing code even on non-win32 so that all themese will parse the same on all arches. We draw pink instead of the actual theme parts on non-win32 though.
2011-11-15 19:52:45 +00:00
#include "gtkwin32themeprivate.h"
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
Include C89 fallback code gtk/gtkborderimage.c, gtk/gtklabel.c and gtk/gtkstyleproperty.c call round() and/or rint(), which was only available in C99 compilers. This adds the inclusion of the fallback implementation (gtk/fallback-c89.c) to define these functions if they are not initially made available by the compiler. Also remove the rint() implementation in gtk/gtklabel.c as it is now in the fallback implmentation.
2011-08-19 06:26:52 +00:00
/* this is in case round() is not provided by the compiler,
* such as in the case of C89 compilers, like MSVC
*/
#include "fallback-c89.c"
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
static GHashTable *parse_funcs = NULL;
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static GHashTable *print_funcs = NULL;
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
static GHashTable *properties = NULL;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
static void
styleproperty: Rename ParseFunc and PrintFunc typedefs Rename them to GtkStyleParseFunc and GtkStylePrintFunc, so we can export them.
2011-05-25 23:32:01 +00:00
register_conversion_function (GType type,
GtkStyleParseFunc parse,
GtkStylePrintFunc print)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (parse)
g_hash_table_insert (parse_funcs, GSIZE_TO_POINTER (type), parse);
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
if (print)
g_hash_table_insert (print_funcs, GSIZE_TO_POINTER (type), print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
styleproperty: Factor out the string print function We print strings in lots of places.
2011-05-29 03:25:05 +00:00
static void
string_append_double (GString *string,
double d)
{
char buf[G_ASCII_DTOSTR_BUF_SIZE];
g_ascii_dtostr (buf, sizeof (buf), d);
g_string_append (string, buf);
}
static void
string_append_string (GString *str,
const char *string)
{
gsize len;
g_string_append_c (str, '"');
do {
len = strcspn (string, "\"\n\r\f");
g_string_append (str, string);
string += len;
switch (*string)
{
case '\0':
break;
case '\n':
g_string_append (str, "\\A ");
break;
case '\r':
g_string_append (str, "\\D ");
break;
case '\f':
g_string_append (str, "\\C ");
break;
case '\"':
g_string_append (str, "\\\"");
break;
default:
g_assert_not_reached ();
break;
}
} while (*string);
g_string_append_c (str, '"');
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
/*** IMPLEMENTATIONS ***/
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
rgba_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkSymbolicColor *symbolic;
GdkRGBA rgba;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
symbolic = _gtk_css_parser_read_symbolic_color (parser);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (symbolic == NULL)
return FALSE;
cssstringfuncs: Read colors as symbolic colors always ... and then try to resolve them, so static colors get resolved in advance, just like it was before.
2011-05-18 11:15:05 +00:00
if (gtk_symbolic_color_resolve (symbolic, NULL, &rgba))
{
g_value_set_boxed (value, &rgba);
gtkstyleproperty: Plug two leaks In rgba_value_parse(), unref the symbolic color once we've resolved it. In gradient_value_parse(), take the GtkGradient so we leave no dangling references.
2011-05-30 17:21:42 +00:00
gtk_symbolic_color_unref (symbolic);
cssstringfuncs: Read colors as symbolic colors always ... and then try to resolve them, so static colors get resolved in advance, just like it was before.
2011-05-18 11:15:05 +00:00
}
else
{
g_value_unset (value);
g_value_init (value, GTK_TYPE_SYMBOLIC_COLOR);
g_value_take_boxed (value, symbolic);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
rgba_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
const GdkRGBA *rgba = g_value_get_boxed (value);
if (rgba == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
char *s = gdk_rgba_to_string (rgba);
g_string_append (string, s);
g_free (s);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
color_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkSymbolicColor *symbolic;
cssstringfuncs: Read colors as symbolic colors always ... and then try to resolve them, so static colors get resolved in advance, just like it was before.
2011-05-18 11:15:05 +00:00
GdkRGBA rgba;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
symbolic = _gtk_css_parser_read_symbolic_color (parser);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (symbolic == NULL)
return FALSE;
cssstringfuncs: Read colors as symbolic colors always ... and then try to resolve them, so static colors get resolved in advance, just like it was before.
2011-05-18 11:15:05 +00:00
if (gtk_symbolic_color_resolve (symbolic, NULL, &rgba))
{
GdkColor color;
color.red = rgba.red * 65535. + 0.5;
color.green = rgba.green * 65535. + 0.5;
color.blue = rgba.blue * 65535. + 0.5;
g_value_set_boxed (value, &color);
}
else
{
g_value_unset (value);
g_value_init (value, GTK_TYPE_SYMBOLIC_COLOR);
g_value_take_boxed (value, symbolic);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
color_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
const GdkColor *color = g_value_get_boxed (value);
if (color == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
char *s = gdk_color_to_string (color);
g_string_append (string, s);
g_free (s);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
static gboolean
symbolic_color_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkSymbolicColor *symbolic;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
symbolic = _gtk_css_parser_read_symbolic_color (parser);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (symbolic == NULL)
return FALSE;
g_value_take_boxed (value, symbolic);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
symbolic_color_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkSymbolicColor *symbolic = g_value_get_boxed (value);
if (symbolic == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
char *s = gtk_symbolic_color_to_string (symbolic);
g_string_append (string, s);
g_free (s);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
static gboolean
font_family_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
GPtrArray *names;
char *name;
/* We don't special case generic families. Pango should do
* that for us */
names = g_ptr_array_new ();
do {
name = _gtk_css_parser_try_ident (parser, TRUE);
if (name)
{
GString *string = g_string_new (name);
g_free (name);
while ((name = _gtk_css_parser_try_ident (parser, TRUE)))
{
g_string_append_c (string, ' ');
g_string_append (string, name);
g_free (name);
}
name = g_string_free (string, FALSE);
}
else
{
name = _gtk_css_parser_read_string (parser);
if (name == NULL)
{
g_ptr_array_free (names, TRUE);
return FALSE;
}
}
g_ptr_array_add (names, name);
} while (_gtk_css_parser_try (parser, ",", TRUE));
/* NULL-terminate array */
g_ptr_array_add (names, NULL);
g_value_set_boxed (value, g_ptr_array_free (names, FALSE));
return TRUE;
}
static void
font_family_value_print (const GValue *value,
GString *string)
{
const char **names = g_value_get_boxed (value);
if (names == NULL || *names == NULL)
{
g_string_append (string, "none");
return;
}
string_append_string (string, *names);
names++;
while (*names)
{
g_string_append (string, ", ");
string_append_string (string, *names);
names++;
}
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
font_description_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
PangoFontDescription *font_desc;
css: Ensure that font shorthand parsing sets font name and size
2011-07-21 00:40:06 +00:00
guint mask;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str;
str = _gtk_css_parser_read_value (parser);
if (str == NULL)
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
font_desc = pango_font_description_from_string (str);
css: Ensure that font shorthand parsing sets font name and size
2011-07-21 00:40:06 +00:00
mask = pango_font_description_get_set_fields (font_desc);
/* These values are not really correct,
* but the fields must be set, so we set them to something */
if ((mask & PANGO_FONT_MASK_FAMILY) == 0)
pango_font_description_set_family_static (font_desc, "Sans");
if ((mask & PANGO_FONT_MASK_SIZE) == 0)
pango_font_description_set_size (font_desc, 10 * PANGO_SCALE);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_value_take_boxed (value, font_desc);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
font_description_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
const PangoFontDescription *desc = g_value_get_boxed (value);
if (desc == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
char *s = pango_font_description_to_string (desc);
g_string_append (string, s);
g_free (s);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
boolean_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (_gtk_css_parser_try (parser, "true", TRUE) ||
_gtk_css_parser_try (parser, "1", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
g_value_set_boolean (value, TRUE);
return TRUE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
else if (_gtk_css_parser_try (parser, "false", TRUE) ||
_gtk_css_parser_try (parser, "0", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
g_value_set_boolean (value, FALSE);
return TRUE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
else
{
_gtk_css_parser_error (parser, "Expected a boolean value");
return FALSE;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
boolean_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
if (g_value_get_boolean (value))
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "true");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "false");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
int_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
gint i;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
win32-theme: Support -gtk-win32-size CSS value
2011-11-16 11:19:20 +00:00
if (_gtk_css_parser_begins_with (parser, '-'))
{
int res = _gtk_win32_theme_int_parse (parser, base, &i);
if (res >= 0)
{
g_value_set_int (value, i);
return res > 0;
}
/* < 0 => continue */
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (!_gtk_css_parser_try_int (parser, &i))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected a valid integer value");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
g_value_set_int (value, i);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
int_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%d", g_value_get_int (value));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
uint_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
guint u;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (!_gtk_css_parser_try_uint (parser, &u))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected a valid unsigned value");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
g_value_set_uint (value, u);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
uint_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%u", g_value_get_uint (value));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
double_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
gdouble d;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (!_gtk_css_parser_try_double (parser, &d))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected a number");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
g_value_set_double (value, d);
return TRUE;
}
styleproperty: Add a helper function for printing doubles We will want to print doubles more often, so add a little helper function for it. We cannot use printf() here, because we want English locale output.
2011-05-26 01:44:09 +00:00
static void
double_value_print (const GValue *value,
GString *string)
{
string_append_double (string, g_value_get_double (value));
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
float_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
gdouble d;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (!_gtk_css_parser_try_double (parser, &d))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected a number");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
g_value_set_float (value, d);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
float_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
styleproperty: Add a helper function for printing doubles We will want to print doubles more often, so add a little helper function for it. We cannot use printf() here, because we want English locale output.
2011-05-26 01:44:09 +00:00
string_append_double (string, g_value_get_float (value));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
string_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str = _gtk_css_parser_read_string (parser);
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (str == NULL)
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
return FALSE;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_value_take_string (value, str);
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
string_value_print (const GValue *value,
GString *str)
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
{
styleproperty: Factor out the string print function We print strings in lots of places.
2011-05-29 03:25:05 +00:00
string_append_string (str, g_value_get_string (value));
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
theming_engine_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkThemingEngine *engine;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str;
str = _gtk_css_parser_try_ident (parser, TRUE);
if (str == NULL)
{
_gtk_css_parser_error (parser, "Expected a valid theme name");
return FALSE;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
engine = gtk_theming_engine_load (str);
if (engine == NULL)
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Themeing engine '%s' not found", str);
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
g_value_set_object (value, engine);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
theming_engine_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkThemingEngine *engine;
char *name;
engine = g_value_get_object (value);
if (engine == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
/* XXX: gtk_theming_engine_get_name()? */
g_object_get (engine, "name", &name, NULL);
styleproperty: An unnamed engine equals the default engine
2011-07-20 10:31:44 +00:00
g_string_append (string, name ? name : "none");
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_free (name);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
animation_description_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkAnimationDescription *desc;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str;
str = _gtk_css_parser_read_value (parser);
if (str == NULL)
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
desc = _gtk_animation_description_from_string (str);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (desc == NULL)
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
{
_gtk_css_parser_error (parser, "Invalid animation description");
return FALSE;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_value_take_boxed (value, desc);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
animation_description_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkAnimationDescription *desc = g_value_get_boxed (value);
if (desc == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
_gtk_animation_description_print (desc, string);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
static gboolean
border_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
GtkBorder border = { 0, };
guint i, numbers[4];
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
for (i = 0; i < G_N_ELEMENTS (numbers); i++)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
win32-theme: Support -gtk-win32-size CSS value
2011-11-16 11:19:20 +00:00
if (_gtk_css_parser_begins_with (parser, '-'))
{
/* These are strictly speaking signed, but we want to be able to use them
for unsigned types too, as the actual ranges of values make this safe */
int res = _gtk_win32_theme_int_parse (parser, base, (int *)&numbers[i]);
if (res == 0) /* Parse error, report */
return FALSE;
if (res < 0) /* Nothing known to expand */
break;
}
else
{
if (!_gtk_css_parser_try_uint (parser, &numbers[i]))
break;
/* XXX: shouldn't allow spaces here? */
_gtk_css_parser_try (parser, "px", TRUE);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (i == 0)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected valid border");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
border.top = numbers[0];
if (i > 1)
border.right = numbers[1];
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
border.right = border.top;
if (i > 2)
border.bottom = numbers[2];
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
border.bottom = border.top;
if (i > 3)
border.left = numbers[3];
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
border.left = border.right;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_value_set_boxed (value, &border);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
border_value_print (const GValue *value, GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
const GtkBorder *border = g_value_get_boxed (value);
if (border == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else if (border->left != border->right)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%d %d %d %d", border->top, border->right, border->bottom, border->left);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else if (border->top != border->bottom)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%d %d %d", border->top, border->right, border->bottom);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else if (border->top != border->left)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%d %d", border->top, border->right);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append_printf (string, "%d", border->top);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
static gboolean
gradient_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkGradient *gradient;
cairo_pattern_type_t type;
gdouble coords[6];
guint i;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, "-gtk-gradient", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected '-gtk-gradient'");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, "(", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected '(' after '-gtk-gradient'");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
/* Parse gradient type */
if (_gtk_css_parser_try (parser, "linear", TRUE))
type = CAIRO_PATTERN_TYPE_LINEAR;
else if (_gtk_css_parser_try (parser, "radial", TRUE))
type = CAIRO_PATTERN_TYPE_RADIAL;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
else
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Gradient type must be 'radial' or 'linear'");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
/* Parse start/stop position parameters */
for (i = 0; i < 2; i++)
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (! _gtk_css_parser_try (parser, ",", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected ','");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (_gtk_css_parser_try (parser, "left", TRUE))
coords[i * 3] = 0;
else if (_gtk_css_parser_try (parser, "right", TRUE))
coords[i * 3] = 1;
else if (_gtk_css_parser_try (parser, "center", TRUE))
coords[i * 3] = 0.5;
else if (!_gtk_css_parser_try_double (parser, &coords[i * 3]))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected a valid X coordinate");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (_gtk_css_parser_try (parser, "top", TRUE))
coords[i * 3 + 1] = 0;
else if (_gtk_css_parser_try (parser, "bottom", TRUE))
coords[i * 3 + 1] = 1;
else if (_gtk_css_parser_try (parser, "center", TRUE))
coords[i * 3 + 1] = 0.5;
else if (!_gtk_css_parser_try_double (parser, &coords[i * 3 + 1]))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected a valid Y coordinate");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
if (type == CAIRO_PATTERN_TYPE_RADIAL)
{
/* Parse radius */
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (! _gtk_css_parser_try (parser, ",", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
_gtk_css_parser_error (parser,
"Expected ','");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (! _gtk_css_parser_try_double (parser, &coords[(i * 3) + 2]))
{
_gtk_css_parser_error (parser,
"Expected a numer for the radius");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
}
if (type == CAIRO_PATTERN_TYPE_LINEAR)
gradient = gtk_gradient_new_linear (coords[0], coords[1], coords[3], coords[4]);
else
gradient = gtk_gradient_new_radial (coords[0], coords[1], coords[2],
coords[3], coords[4], coords[5]);
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
while (_gtk_css_parser_try (parser, ",", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkSymbolicColor *color;
gdouble position;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (_gtk_css_parser_try (parser, "from", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
position = 0;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, "(", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected '('");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
else if (_gtk_css_parser_try (parser, "to", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
position = 1;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, "(", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected '('");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
else if (_gtk_css_parser_try (parser, "color-stop", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, "(", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected '('");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try_double (parser, &position))
{
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected a valid number");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, ",", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected a comma");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
}
else
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Not a valid color-stop definition");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
color = _gtk_css_parser_read_symbolic_color (parser);
if (color == NULL)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_add_color_stop (gradient, position, color);
gtk_symbolic_color_unref (color);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, ")", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected ')'");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
}
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
if (!_gtk_css_parser_try (parser, ")", TRUE))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Rewrite the parser Instead of relying on GScanner and its idea of syntax, code up a parser that obeys the CSS spec. This also has the great side effect of reporting correct line numbers and positions. Also included is a reorganization of the returned error values. Instead of error values describing what type of syntax error was returned, the code just returns SYNTAX_ERROR. Other messages exist for when actual values don't work or when errors shouldn't be fatal due to backwards compatibility.
2011-04-14 02:47:18 +00:00
gtk_gradient_unref (gradient);
_gtk_css_parser_error (parser,
"Expected ')'");
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return FALSE;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
gtkstyleproperty: Plug two leaks In rgba_value_parse(), unref the symbolic color once we've resolved it. In gradient_value_parse(), take the GtkGradient so we leave no dangling references.
2011-05-30 17:21:42 +00:00
g_value_take_boxed (value, gradient);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
gradient_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GtkGradient *gradient = g_value_get_boxed (value);
if (gradient == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
{
char *s = gtk_gradient_to_string (gradient);
g_string_append (string, s);
g_free (s);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
stringfuncs: Remove _gtk_css_parse_url() from header No code changes, just moving the function in the c file so it doesn't need to be forward-declared.
2011-05-21 01:18:00 +00:00
static GFile *
gtk_css_parse_url (GtkCssParser *parser,
GFile *base)
{
gchar *path;
GFile *file;
if (_gtk_css_parser_try (parser, "url", FALSE))
{
if (!_gtk_css_parser_try (parser, "(", TRUE))
{
_gtk_css_parser_skip_whitespace (parser);
if (_gtk_css_parser_try (parser, "(", TRUE))
{
GError *error;
error = g_error_new_literal (GTK_CSS_PROVIDER_ERROR,
GTK_CSS_PROVIDER_ERROR_DEPRECATED,
styleproperty: Improve error message We do dsupport that feature still and don't want to scare people into thinking everything broke.
2011-06-17 05:41:37 +00:00
"Whitespace between 'url' and '(' is deprecated");
stringfuncs: Remove _gtk_css_parse_url() from header No code changes, just moving the function in the c file so it doesn't need to be forward-declared.
2011-05-21 01:18:00 +00:00
_gtk_css_parser_take_error (parser, error);
}
else
{
_gtk_css_parser_error (parser, "Expected '(' after 'url'");
return NULL;
}
}
path = _gtk_css_parser_read_string (parser);
if (path == NULL)
return NULL;
if (!_gtk_css_parser_try (parser, ")", TRUE))
{
_gtk_css_parser_error (parser, "No closing ')' found for 'url'");
g_free (path);
return NULL;
}
}
else
{
path = _gtk_css_parser_try_name (parser, TRUE);
if (path == NULL)
{
_gtk_css_parser_error (parser, "Not a valid url");
return NULL;
}
}
file = g_file_resolve_relative_path (base, path);
g_free (path);
return file;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
pattern_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (_gtk_css_parser_begins_with (parser, '-'))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
Add initial cut at win32 theme support for CSS We now support -gtk-win32-theme-part(class,part,state) in background and border-image CSS properties. This renders the corresponding theme part using DrawThemeBackground() and acts as a base for a CSS based windows theme. Note that we build the parsing code even on non-win32 so that all themese will parse the same on all arches. We draw pink instead of the actual theme parts on non-win32 though.
2011-11-15 19:52:45 +00:00
int res;
res = _gtk_win32_theme_part_parse (parser, base, value);
if (res >= 0)
return res > 0;
/* < 0 => continue */
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_value_unset (value);
g_value_init (value, GTK_TYPE_GRADIENT);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return gradient_value_parse (parser, base, value);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
else
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
GError *error = NULL;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
gchar *path;
GdkPixbuf *pixbuf;
GFile *file;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
cairo_surface_t *surface;
cairo_pattern_t *pattern;
cairo_t *cr;
cairo_matrix_t matrix;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
stringfuncs: Remove _gtk_css_parse_url() from header No code changes, just moving the function in the c file so it doesn't need to be forward-declared.
2011-05-21 01:18:00 +00:00
file = gtk_css_parse_url (parser, base);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (file == NULL)
return FALSE;
path = g_file_get_path (file);
g_object_unref (file);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
pixbuf = gdk_pixbuf_new_from_file (path, &error);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_free (path);
if (pixbuf == NULL)
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
{
_gtk_css_parser_take_error (parser, error);
return FALSE;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
surface = cairo_image_surface_create (CAIRO_FORMAT_ARGB32,
gdk_pixbuf_get_width (pixbuf),
gdk_pixbuf_get_height (pixbuf));
cr = cairo_create (surface);
gdk_cairo_set_source_pixbuf (cr, pixbuf, 0, 0);
cairo_paint (cr);
pattern = cairo_pattern_create_for_surface (surface);
cairo_matrix_init_scale (&matrix,
gdk_pixbuf_get_width (pixbuf),
gdk_pixbuf_get_height (pixbuf));
cairo_pattern_set_matrix (pattern, &matrix);
cairo_surface_destroy (surface);
cairo_destroy (cr);
g_object_unref (pixbuf);
g_value_take_boxed (value, pattern);
}
return TRUE;
}
styleproperty: dump image surfaces as PNG data URLs We're not able to load that (yet?), but it's useful for producing diffable test dumps.
2011-07-20 22:43:00 +00:00
static cairo_status_t
surface_write (void *closure,
const unsigned char *data,
unsigned int length)
{
g_byte_array_append (closure, data, length);
return CAIRO_STATUS_SUCCESS;
}
static void
surface_print (cairo_surface_t *surface,
GString * string)
{
#if CAIRO_HAS_PNG_FUNCTIONS
GByteArray *array;
char *base64;
array = g_byte_array_new ();
cairo_surface_write_to_png_stream (surface, surface_write, array);
base64 = g_base64_encode (array->data, array->len);
g_byte_array_free (array, TRUE);
g_string_append (string, "url(\"data:image/png;base64,");
g_string_append (string, base64);
g_string_append (string, "\")");
g_free (base64);
#else
g_string_append (string, "none /* you need cairo png functions enabled to make this work */");
#endif
}
static void
pattern_value_print (const GValue *value,
GString *string)
{
cairo_pattern_t *pattern;
cairo_surface_t *surface;
pattern = g_value_get_boxed (value);
if (pattern == NULL)
{
g_string_append (string, "none");
return;
}
switch (cairo_pattern_get_type (pattern))
{
case CAIRO_PATTERN_TYPE_SURFACE:
if (cairo_pattern_get_surface (pattern, &surface) != CAIRO_STATUS_SUCCESS)
{
g_assert_not_reached ();
}
surface_print (surface, string);
break;
case CAIRO_PATTERN_TYPE_SOLID:
case CAIRO_PATTERN_TYPE_LINEAR:
case CAIRO_PATTERN_TYPE_RADIAL:
default:
g_assert_not_reached ();
break;
}
}
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
static gboolean
shadow_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
gboolean have_inset, have_color, have_lengths;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
gdouble hoffset, voffset, blur, spread;
GtkSymbolicColor *color;
GtkShadow *shadow;
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
guint i;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
shadow = _gtk_shadow_new ();
do
{
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
have_inset = have_lengths = have_color = FALSE;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
for (i = 0; i < 3; i++)
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
{
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
if (!have_inset &&
_gtk_css_parser_try (parser, "inset", TRUE))
{
have_inset = TRUE;
continue;
}
if (!have_lengths &&
_gtk_css_parser_try_double (parser, &hoffset))
{
have_lengths = TRUE;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
if (!_gtk_css_parser_try_double (parser, &voffset))
{
_gtk_css_parser_error (parser, "Horizontal and vertical offsets are required");
_gtk_shadow_unref (shadow);
return FALSE;
}
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
if (!_gtk_css_parser_try_double (parser, &blur))
blur = 0;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
if (!_gtk_css_parser_try_double (parser, &spread))
spread = 0;
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
continue;
}
if (!have_color)
{
have_color = TRUE;
/* XXX: the color is optional and UA-defined if it's missing,
* but it doesn't really make sense for us...
*/
color = _gtk_css_parser_read_symbolic_color (parser);
if (color == NULL)
{
_gtk_shadow_unref (shadow);
return FALSE;
}
}
}
if (!have_color || !have_lengths)
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
{
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
_gtk_css_parser_error (parser, "Must specify at least color and offsets");
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
_gtk_shadow_unref (shadow);
return FALSE;
}
_gtk_shadow_append (shadow,
hoffset, voffset,
blur, spread,
styleproperty: Parse shadow options in any order The CSS spec allows random ordering of various properties, so we should, too.
2011-06-11 03:24:54 +00:00
have_inset, color);
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
gtk_symbolic_color_unref (color);
}
while (_gtk_css_parser_try (parser, ",", TRUE));
g_value_take_boxed (value, shadow);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
shadow_value_print (const GValue *value,
GString *string)
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
{
GtkShadow *shadow;
shadow = g_value_get_boxed (value);
if (shadow == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, "none");
else
_gtk_shadow_print (shadow, string);
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
}
css: start background-repeat By default, a background image is stretched. Instead, it is worth to have a tiled background. This patch allows background surfaces to be repeated or not, and should be compatible with future extensions and CSS. https://bugzilla.gnome.org/show_bug.cgi?id=663522
2011-11-06 18:33:05 +00:00
static gboolean
background_repeat_value_parse (GtkCssParser *parser,
GFile *file,
GValue *value)
{
GtkCssBackgroundRepeat repeat;
GtkCssBackgroundRepeatStyle style;
if (_gtk_css_parser_try (parser, "repeat", TRUE))
style = GTK_CSS_BACKGROUND_REPEAT_STYLE_REPEAT;
else if (_gtk_css_parser_try (parser, "no-repeat", TRUE))
style = GTK_CSS_BACKGROUND_REPEAT_STYLE_NO_REPEAT;
else
style = GTK_CSS_BACKGROUND_REPEAT_STYLE_NONE;
repeat.repeat = style;
g_value_set_boxed (value, &repeat);
return TRUE;
}
static const gchar *
background_repeat_style_to_string (GtkCssBackgroundRepeatStyle repeat)
{
switch (repeat)
{
case GTK_CSS_BACKGROUND_REPEAT_STYLE_REPEAT:
return "repeat";
case GTK_CSS_BACKGROUND_REPEAT_STYLE_NO_REPEAT:
return "no-repeat";
default:
return NULL;
}
}
static void
background_repeat_value_print (const GValue *value,
GString *string)
{
GtkCssBackgroundRepeat *repeat;
repeat = g_value_get_boxed (value);
g_string_append (string, background_repeat_style_to_string (repeat->repeat));
}
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
static gboolean
border_image_repeat_value_parse (GtkCssParser *parser,
GFile *file,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
GtkCssBorderImageRepeat image_repeat;
Rename s/GtkCssRepeatStyle/GtkCssBorderRepeatStyle/g https://bugzilla.gnome.org/show_bug.cgi?id=663522
2011-11-07 23:10:13 +00:00
GtkCssBorderRepeatStyle styles[2];
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
gint i;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
for (i = 0; i < 2; i++)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (_gtk_css_parser_try (parser, "stretch", TRUE))
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
styles[i] = GTK_CSS_REPEAT_STYLE_NONE;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
else if (_gtk_css_parser_try (parser, "repeat", TRUE))
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
styles[i] = GTK_CSS_REPEAT_STYLE_REPEAT;
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
else if (_gtk_css_parser_try (parser, "round", TRUE))
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
styles[i] = GTK_CSS_REPEAT_STYLE_ROUND;
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
else if (_gtk_css_parser_try (parser, "space", TRUE))
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
styles[i] = GTK_CSS_REPEAT_STYLE_SPACE;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
else if (i == 0)
{
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
styles[1] = styles[0] = GTK_CSS_REPEAT_STYLE_NONE;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
break;
}
else
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
styles[i] = styles[0];
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
border-image-repeat: Fix order of hrepeat and vrepeat
2011-06-08 05:04:38 +00:00
image_repeat.hrepeat = styles[0];
image_repeat.vrepeat = styles[1];
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
g_value_set_boxed (value, &image_repeat);
return TRUE;
}
static const gchar *
Rename s/GtkCssRepeatStyle/GtkCssBorderRepeatStyle/g https://bugzilla.gnome.org/show_bug.cgi?id=663522
2011-11-07 23:10:13 +00:00
border_image_repeat_style_to_string (GtkCssBorderRepeatStyle repeat)
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
{
switch (repeat)
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
{
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
case GTK_CSS_REPEAT_STYLE_NONE:
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
return "stretch";
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
case GTK_CSS_REPEAT_STYLE_REPEAT:
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
return "repeat";
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
case GTK_CSS_REPEAT_STYLE_ROUND:
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
return "round";
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
case GTK_CSS_REPEAT_STYLE_SPACE:
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
return "space";
default:
return NULL;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
}
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
static void
border_image_repeat_value_print (const GValue *value,
GString *string)
{
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
GtkCssBorderImageRepeat *image_repeat;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
image_repeat = g_value_get_boxed (value);
css: Don't print both border-image-repeat values when they're identical
2011-07-21 00:40:34 +00:00
g_string_append (string, border_image_repeat_style_to_string (image_repeat->hrepeat));
if (image_repeat->hrepeat != image_repeat->vrepeat)
{
g_string_append (string, " ");
g_string_append (string, border_image_repeat_style_to_string (image_repeat->vrepeat));
}
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
}
static gboolean
border_image_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
Use G_VALUE_INIT Instead of an explicit { 0, } when declaring the variable.
2011-09-30 15:31:04 +00:00
GValue temp = G_VALUE_INIT;
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
cairo_pattern_t *pattern = NULL;
Save a generic boxes source in GtkImageBorder We used to special handle gradient, but we want to be able to store other sources (that will eventually resolve to cairo patterns). For instance, this is needed to handle win32 theme part sources.
2011-11-15 15:26:46 +00:00
gconstpointer *boxed = NULL;
GType boxed_type;
border-image: add support for border-image-width too It's useful to set a slice size != border-width, as backgrounds are clipped to border-width too. As slices can be half-transparent and overlap the background, this would not fill the border box properly if we only use a single property for specifying the width. Also, this brings us even closer to CSS3.
2011-06-09 23:59:14 +00:00
GtkBorder slice, *width = NULL, *parsed_slice;
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
GtkCssBorderImageRepeat repeat, *parsed_repeat;
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
gboolean retval = FALSE;
GtkBorderImage *image = NULL;
g_value_init (&temp, CAIRO_GOBJECT_TYPE_PATTERN);
if (!pattern_value_parse (parser, base, &temp))
return FALSE;
Save a generic boxes source in GtkImageBorder We used to special handle gradient, but we want to be able to store other sources (that will eventually resolve to cairo patterns). For instance, this is needed to handle win32 theme part sources.
2011-11-15 15:26:46 +00:00
boxed_type = G_VALUE_TYPE (&temp);
if (boxed_type != CAIRO_GOBJECT_TYPE_PATTERN)
boxed = g_value_dup_boxed (&temp);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
else
pattern = g_value_dup_boxed (&temp);
g_value_unset (&temp);
g_value_init (&temp, GTK_TYPE_BORDER);
if (!border_value_parse (parser, base, &temp))
goto out;
border-image: add support for border-image-width too It's useful to set a slice size != border-width, as backgrounds are clipped to border-width too. As slices can be half-transparent and overlap the background, this would not fill the border box properly if we only use a single property for specifying the width. Also, this brings us even closer to CSS3.
2011-06-09 23:59:14 +00:00
parsed_slice = g_value_get_boxed (&temp);
slice = *parsed_slice;
if (_gtk_css_parser_try (parser, "/", TRUE))
{
g_value_unset (&temp);
g_value_init (&temp, GTK_TYPE_BORDER);
if (!border_value_parse (parser, base, &temp))
goto out;
width = g_value_dup_boxed (&temp);
}
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
g_value_unset (&temp);
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
g_value_init (&temp, GTK_TYPE_CSS_BORDER_IMAGE_REPEAT);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
if (!border_image_repeat_value_parse (parser, base, &temp))
goto out;
parsed_repeat = g_value_get_boxed (&temp);
repeat = *parsed_repeat;
g_value_unset (&temp);
Save a generic boxes source in GtkImageBorder We used to special handle gradient, but we want to be able to store other sources (that will eventually resolve to cairo patterns). For instance, this is needed to handle win32 theme part sources.
2011-11-15 15:26:46 +00:00
if (boxed != NULL)
image = _gtk_border_image_new_for_boxed (boxed_type, boxed, &slice, width, &repeat);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
else if (pattern != NULL)
border-image: add support for border-image-width too It's useful to set a slice size != border-width, as backgrounds are clipped to border-width too. As slices can be half-transparent and overlap the background, this would not fill the border box properly if we only use a single property for specifying the width. Also, this brings us even closer to CSS3.
2011-06-09 23:59:14 +00:00
image = _gtk_border_image_new (pattern, &slice, width, &repeat);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
if (image != NULL)
{
retval = TRUE;
g_value_take_boxed (value, image);
}
out:
if (pattern != NULL)
cairo_pattern_destroy (pattern);
Save a generic boxes source in GtkImageBorder We used to special handle gradient, but we want to be able to store other sources (that will eventually resolve to cairo patterns). For instance, this is needed to handle win32 theme part sources.
2011-11-15 15:26:46 +00:00
if (boxed != NULL)
g_boxed_free (boxed_type, boxed);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
border-image: add support for border-image-width too It's useful to set a slice size != border-width, as backgrounds are clipped to border-width too. As slices can be half-transparent and overlap the background, this would not fill the border box properly if we only use a single property for specifying the width. Also, this brings us even closer to CSS3.
2011-06-09 23:59:14 +00:00
if (width != NULL)
gtk_border_free (width);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
return retval;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
enum_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GEnumClass *enum_class;
GEnumValue *enum_value;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
str = _gtk_css_parser_try_ident (parser, TRUE);
if (str == NULL)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser, "Expected an identifier");
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
enum_class = g_type_class_ref (G_VALUE_TYPE (value));
enum_value = g_enum_get_value_by_nick (enum_class, str);
if (enum_value)
g_value_set_enum (value, enum_value->value);
else
_gtk_css_parser_error (parser,
"Unknown value '%s' for enum type '%s'",
str, g_type_name (G_VALUE_TYPE (value)));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_type_class_unref (enum_class);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (str);
return enum_value != NULL;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
enum_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GEnumClass *enum_class;
GEnumValue *enum_value;
enum_class = g_type_class_ref (G_VALUE_TYPE (value));
enum_value = g_enum_get_value (enum_class, g_value_get_enum (value));
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, enum_value->value_nick);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_type_class_unref (enum_class);
}
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
flags_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GFlagsClass *flags_class;
GFlagsValue *flag_value;
guint flags = 0;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
char *str;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
flags_class = g_type_class_ref (G_VALUE_TYPE (value));
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
do {
str = _gtk_css_parser_try_ident (parser, TRUE);
if (str == NULL)
{
_gtk_css_parser_error (parser, "Expected an identifier");
g_type_class_unref (flags_class);
return FALSE;
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
flag_value = g_flags_get_value_by_nick (flags_class, str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (!flag_value)
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser,
"Unknown flag value '%s' for type '%s'",
str, g_type_name (G_VALUE_TYPE (value)));
/* XXX Do we want to return FALSE here? We can get
* forward-compatibility for new values this way
*/
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_type_class_unref (flags_class);
return FALSE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (str);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
while (_gtk_css_parser_try (parser, ",", FALSE));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
g_type_class_unref (flags_class);
g_value_set_enum (value, flags);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
flags_value_print (const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
GFlagsClass *flags_class;
guint i, flags;
flags_class = g_type_class_ref (G_VALUE_TYPE (value));
flags = g_value_get_flags (value);
for (i = 0; i < flags_class->n_values; i++)
{
GFlagsValue *flags_value = &flags_class->values[i];
if (flags & flags_value->value)
{
if (string->len != 0)
g_string_append (string, ", ");
g_string_append (string, flags_value->value_nick);
}
}
g_type_class_unref (flags_class);
}
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
static gboolean
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
bindings_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
{
GPtrArray *array;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
GtkBindingSet *binding_set;
char *name;
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
array = g_ptr_array_new ();
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
do {
name = _gtk_css_parser_try_ident (parser, TRUE);
if (name == NULL)
{
_gtk_css_parser_error (parser, "Not a valid binding name");
g_ptr_array_free (array, TRUE);
return FALSE;
}
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
binding_set = gtk_binding_set_find (name);
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
if (!binding_set)
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
{
_gtk_css_parser_error (parser, "No binding set named '%s'", name);
g_free (name);
continue;
}
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
g_ptr_array_add (array, binding_set);
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_free (name);
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
while (_gtk_css_parser_try (parser, ",", TRUE));
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
g_value_take_boxed (value, array);
return TRUE;
}
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
static void
bindings_value_print (const GValue *value,
GString *string)
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
{
GPtrArray *array;
guint i;
array = g_value_get_boxed (value);
for (i = 0; i < array->len; i++)
{
GtkBindingSet *binding_set = g_ptr_array_index (array, i);
if (i > 0)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
g_string_append (string, ", ");
g_string_append (string, binding_set->set_name);
css: Move binding set parsing to strfuncs
2011-05-18 16:03:23 +00:00
}
}
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
static gboolean
border_corner_radius_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
GtkCssBorderCornerRadius corner;
if (!_gtk_css_parser_try_double (parser, &corner.horizontal))
{
_gtk_css_parser_error (parser, "Expected a number");
return FALSE;
}
else if (corner.horizontal < 0)
goto negative;
if (!_gtk_css_parser_try_double (parser, &corner.vertical))
corner.vertical = corner.horizontal;
else if (corner.vertical < 0)
goto negative;
g_value_set_boxed (value, &corner);
return TRUE;
negative:
_gtk_css_parser_error (parser, "Border radius values cannot be negative");
return FALSE;
}
static void
border_corner_radius_value_print (const GValue *value,
GString *string)
{
GtkCssBorderCornerRadius *corner;
corner = g_value_get_boxed (value);
if (corner == NULL)
{
g_string_append (string, "none");
return;
}
string_append_double (string, corner->horizontal);
if (corner->horizontal != corner->vertical)
{
g_string_append_c (string, ' ');
string_append_double (string, corner->vertical);
}
}
static gboolean
border_radius_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
GtkCssBorderRadius border;
if (!_gtk_css_parser_try_double (parser, &border.top_left.horizontal))
{
_gtk_css_parser_error (parser, "Expected a number");
return FALSE;
}
else if (border.top_left.horizontal < 0)
goto negative;
if (_gtk_css_parser_try_double (parser, &border.top_right.horizontal))
{
if (border.top_right.horizontal < 0)
goto negative;
if (_gtk_css_parser_try_double (parser, &border.bottom_right.horizontal))
{
if (border.bottom_right.horizontal < 0)
goto negative;
if (!_gtk_css_parser_try_double (parser, &border.bottom_left.horizontal))
border.bottom_left.horizontal = border.top_right.horizontal;
else if (border.bottom_left.horizontal < 0)
goto negative;
}
else
{
border.bottom_right.horizontal = border.top_left.horizontal;
border.bottom_left.horizontal = border.top_right.horizontal;
}
}
else
{
border.top_right.horizontal = border.top_left.horizontal;
border.bottom_right.horizontal = border.top_left.horizontal;
border.bottom_left.horizontal = border.top_left.horizontal;
}
if (_gtk_css_parser_try (parser, "/", TRUE))
{
if (!_gtk_css_parser_try_double (parser, &border.top_left.vertical))
{
_gtk_css_parser_error (parser, "Expected a number");
return FALSE;
}
else if (border.top_left.vertical < 0)
goto negative;
if (_gtk_css_parser_try_double (parser, &border.top_right.vertical))
{
if (border.top_right.vertical < 0)
goto negative;
if (_gtk_css_parser_try_double (parser, &border.bottom_right.vertical))
{
if (border.bottom_right.vertical < 0)
goto negative;
if (!_gtk_css_parser_try_double (parser, &border.bottom_left.vertical))
border.bottom_left.vertical = border.top_right.vertical;
else if (border.bottom_left.vertical < 0)
goto negative;
}
else
{
border.bottom_right.vertical = border.top_left.vertical;
border.bottom_left.vertical = border.top_right.vertical;
}
}
else
{
border.top_right.vertical = border.top_left.vertical;
border.bottom_right.vertical = border.top_left.vertical;
border.bottom_left.vertical = border.top_left.vertical;
}
}
else
{
border.top_left.vertical = border.top_left.horizontal;
border.top_right.vertical = border.top_right.horizontal;
border.bottom_right.vertical = border.bottom_right.horizontal;
border.bottom_left.vertical = border.bottom_left.horizontal;
}
/* border-radius is an int property for backwards-compat reasons */
g_value_unset (value);
g_value_init (value, GTK_TYPE_CSS_BORDER_RADIUS);
g_value_set_boxed (value, &border);
return TRUE;
negative:
_gtk_css_parser_error (parser, "Border radius values cannot be negative");
return FALSE;
}
static void
border_radius_value_print (const GValue *value,
GString *string)
{
GtkCssBorderRadius *border;
border = g_value_get_boxed (value);
if (border == NULL)
{
g_string_append (string, "none");
return;
}
string_append_double (string, border->top_left.horizontal);
if (border->top_left.horizontal != border->top_right.horizontal ||
border->top_left.horizontal != border->bottom_right.horizontal ||
border->top_left.horizontal != border->bottom_left.horizontal)
{
g_string_append_c (string, ' ');
string_append_double (string, border->top_right.horizontal);
if (border->top_left.horizontal != border->bottom_right.horizontal ||
border->top_right.horizontal != border->bottom_left.horizontal)
{
g_string_append_c (string, ' ');
string_append_double (string, border->bottom_right.horizontal);
if (border->top_right.horizontal != border->bottom_left.horizontal)
{
g_string_append_c (string, ' ');
string_append_double (string, border->bottom_left.horizontal);
}
}
}
if (border->top_left.horizontal != border->top_left.vertical ||
border->top_right.horizontal != border->top_right.vertical ||
border->bottom_right.horizontal != border->bottom_right.vertical ||
border->bottom_left.horizontal != border->bottom_left.vertical)
{
g_string_append (string, " / ");
string_append_double (string, border->top_left.vertical);
if (border->top_left.vertical != border->top_right.vertical ||
border->top_left.vertical != border->bottom_right.vertical ||
border->top_left.vertical != border->bottom_left.vertical)
{
g_string_append_c (string, ' ');
string_append_double (string, border->top_right.vertical);
if (border->top_left.vertical != border->bottom_right.vertical ||
border->top_right.vertical != border->bottom_left.vertical)
{
g_string_append_c (string, ' ');
string_append_double (string, border->bottom_right.vertical);
if (border->top_right.vertical != border->bottom_left.vertical)
{
g_string_append_c (string, ' ');
string_append_double (string, border->bottom_left.vertical);
}
}
}
}
}
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
static gboolean
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
{
if (_gtk_css_parser_try (parser, "transparent", TRUE))
{
GdkRGBA transparent = { 0, 0, 0, 0 };
g_value_set_boxed (value, &transparent);
return TRUE;
}
return rgba_value_parse (parser, base, value);
}
static gboolean
border_color_shorthand_value_parse (GtkCssParser *parser,
GFile *base,
GValue *value)
{
GtkSymbolicColor *symbolic;
GPtrArray *array;
array = g_ptr_array_new_with_free_func ((GDestroyNotify) gtk_symbolic_color_unref);
do
{
if (_gtk_css_parser_try (parser, "transparent", TRUE))
{
GdkRGBA transparent = { 0, 0, 0, 0 };
symbolic = gtk_symbolic_color_new_literal (&transparent);
}
else
{
symbolic = _gtk_css_parser_read_symbolic_color (parser);
if (symbolic == NULL)
return FALSE;
}
g_ptr_array_add (array, symbolic);
}
while (array->len < 4 &&
!_gtk_css_parser_is_eof (parser) &&
!_gtk_css_parser_begins_with (parser, ';') &&
!_gtk_css_parser_begins_with (parser, '}'));
switch (array->len)
{
default:
g_assert_not_reached ();
break;
case 1:
g_ptr_array_add (array, gtk_symbolic_color_ref (g_ptr_array_index (array, 0)));
/* fall through */
case 2:
g_ptr_array_add (array, gtk_symbolic_color_ref (g_ptr_array_index (array, 0)));
/* fall through */
case 3:
g_ptr_array_add (array, gtk_symbolic_color_ref (g_ptr_array_index (array, 1)));
/* fall through */
case 4:
break;
}
g_value_unset (value);
g_value_init (value, G_TYPE_PTR_ARRAY);
styleproperty: Fix leak
2011-07-06 18:41:31 +00:00
g_value_take_boxed (value, array);
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
return TRUE;
}
style: Convert border-width to a shorthand Add the corresponding properties it refers to, too: - border-top-width - border-left-width - border-bottom-width - border-right-width
2011-05-21 22:10:43 +00:00
/*** PACKING ***/
static GParameter *
unpack_border (const GValue *value,
guint *n_params,
const char *top,
const char *left,
const char *bottom,
const char *right)
{
GParameter *parameter = g_new0 (GParameter, 4);
GtkBorder *border = g_value_get_boxed (value);
parameter[0].name = top;
g_value_init (&parameter[0].value, G_TYPE_INT);
g_value_set_int (&parameter[0].value, border->top);
parameter[1].name = left;
g_value_init (&parameter[1].value, G_TYPE_INT);
g_value_set_int (&parameter[1].value, border->left);
parameter[2].name = bottom;
g_value_init (&parameter[2].value, G_TYPE_INT);
g_value_set_int (&parameter[2].value, border->bottom);
parameter[3].name = right;
g_value_init (&parameter[3].value, G_TYPE_INT);
g_value_set_int (&parameter[3].value, border->right);
*n_params = 4;
return parameter;
}
static void
pack_border (GValue *value,
GtkStyleProperties *props,
GtkStateFlags state,
const char *top,
const char *left,
const char *bottom,
const char *right)
{
GtkBorder border;
int t, l, b, r;
gtk_style_properties_get (props,
state,
top, &t,
left, &l,
bottom, &b,
right, &r,
NULL);
border.top = t;
border.left = l;
border.bottom = b;
border.right = r;
g_value_set_boxed (value, &border);
}
static GParameter *
unpack_border_width (const GValue *value,
guint *n_params)
{
return unpack_border (value, n_params,
"border-top-width", "border-left-width",
"border-bottom-width", "border-right-width");
}
static void
pack_border_width (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
style: Convert border-width to a shorthand Add the corresponding properties it refers to, too: - border-top-width - border-left-width - border-bottom-width - border-right-width
2011-05-21 22:10:43 +00:00
{
pack_border (value, props, state,
"border-top-width", "border-left-width",
"border-bottom-width", "border-right-width");
}
style: Make padding a shorthand ... and add the real properties: - padding-top - padding-left - padding-bottom - padding-right
2011-05-22 03:42:17 +00:00
static GParameter *
unpack_padding (const GValue *value,
guint *n_params)
{
return unpack_border (value, n_params,
"padding-top", "padding-left",
"padding-bottom", "padding-right");
}
static void
pack_padding (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
style: Make padding a shorthand ... and add the real properties: - padding-top - padding-left - padding-bottom - padding-right
2011-05-22 03:42:17 +00:00
{
pack_border (value, props, state,
"padding-top", "padding-left",
"padding-bottom", "padding-right");
}
style: Make margin a shorthand ... and add the real properties: - margin-top - margin-left - margin-bottom - margin-right
2011-05-22 03:55:12 +00:00
static GParameter *
unpack_margin (const GValue *value,
guint *n_params)
{
return unpack_border (value, n_params,
"margin-top", "margin-left",
"margin-bottom", "margin-right");
}
static void
pack_margin (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
style: Make margin a shorthand ... and add the real properties: - margin-top - margin-left - margin-bottom - margin-right
2011-05-22 03:55:12 +00:00
{
pack_border (value, props, state,
"margin-top", "margin-left",
"margin-bottom", "margin-right");
}
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
static GParameter *
unpack_border_radius (const GValue *value,
guint *n_params)
{
GParameter *parameter = g_new0 (GParameter, 4);
GtkCssBorderRadius *border;
if (G_VALUE_HOLDS_BOXED (value))
border = g_value_get_boxed (value);
else
border = NULL;
parameter[0].name = "border-top-left-radius";
g_value_init (&parameter[0].value, GTK_TYPE_CSS_BORDER_CORNER_RADIUS);
parameter[1].name = "border-top-right-radius";
g_value_init (&parameter[1].value, GTK_TYPE_CSS_BORDER_CORNER_RADIUS);
parameter[2].name = "border-bottom-right-radius";
g_value_init (&parameter[2].value, GTK_TYPE_CSS_BORDER_CORNER_RADIUS);
parameter[3].name = "border-bottom-left-radius";
g_value_init (&parameter[3].value, GTK_TYPE_CSS_BORDER_CORNER_RADIUS);
if (border)
{
g_value_set_boxed (&parameter[0].value, &border->top_left);
g_value_set_boxed (&parameter[1].value, &border->top_right);
g_value_set_boxed (&parameter[2].value, &border->bottom_right);
g_value_set_boxed (&parameter[3].value, &border->bottom_left);
}
*n_params = 4;
return parameter;
}
static void
pack_border_radius (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
{
GtkCssBorderCornerRadius *top_left;
/* NB: We are an int property, so we have to resolve to an int here.
* So we just resolve to an int. We pick one and stick to it.
* Lesson learned: Don't query border-radius shorthand, query the
* real properties instead. */
gtk_style_properties_get (props,
state,
"border-top-left-radius", &top_left,
NULL);
if (top_left)
g_value_set_int (value, top_left->horizontal);
g_free (top_left);
}
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
static GParameter *
unpack_font_description (const GValue *value,
guint *n_params)
{
GParameter *parameter = g_new0 (GParameter, 5);
PangoFontDescription *description;
PangoFontMask mask;
guint n;
/* For backwards compat, we only unpack values that are indeed set.
* For strict CSS conformance we need to unpack all of them.
* Note that we do set all of them in the parse function, so it
* will not have effects when parsing CSS files. It will though
* for custom style providers.
*/
description = g_value_get_boxed (value);
n = 0;
if (description)
mask = pango_font_description_get_set_fields (description);
else
mask = 0;
if (mask & PANGO_FONT_MASK_FAMILY)
{
GPtrArray *strv = g_ptr_array_new ();
g_ptr_array_add (strv, g_strdup (pango_font_description_get_family (description)));
g_ptr_array_add (strv, NULL);
parameter[n].name = "font-family";
g_value_init (&parameter[n].value, G_TYPE_STRV);
g_value_take_boxed (&parameter[n].value,
g_ptr_array_free (strv, FALSE));
n++;
}
if (mask & PANGO_FONT_MASK_STYLE)
{
parameter[n].name = "font-style";
g_value_init (&parameter[n].value, PANGO_TYPE_STYLE);
g_value_set_enum (&parameter[n].value,
pango_font_description_get_style (description));
n++;
}
if (mask & PANGO_FONT_MASK_VARIANT)
{
parameter[n].name = "font-variant";
g_value_init (&parameter[n].value, PANGO_TYPE_VARIANT);
g_value_set_enum (&parameter[n].value,
pango_font_description_get_variant (description));
n++;
}
if (mask & PANGO_FONT_MASK_WEIGHT)
{
parameter[n].name = "font-weight";
g_value_init (&parameter[n].value, PANGO_TYPE_WEIGHT);
g_value_set_enum (&parameter[n].value,
pango_font_description_get_weight (description));
n++;
}
if (mask & PANGO_FONT_MASK_SIZE)
{
parameter[n].name = "font-size";
g_value_init (&parameter[n].value, G_TYPE_DOUBLE);
g_value_set_double (&parameter[n].value,
(double) pango_font_description_get_size (description) / PANGO_SCALE);
n++;
}
*n_params = n;
return parameter;
}
static void
pack_font_description (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
{
PangoFontDescription *description;
char **families;
PangoStyle style;
PangoVariant variant;
PangoWeight weight;
double size;
gtk_style_properties_get (props,
state,
"font-family", &families,
"font-style", &style,
"font-variant", &variant,
"font-weight", &weight,
"font-size", &size,
NULL);
description = pango_font_description_new ();
/* xxx: Can we set all the families here somehow? */
if (families)
pango_font_description_set_family (description, families[0]);
pango_font_description_set_size (description, round (size * PANGO_SCALE));
pango_font_description_set_style (description, style);
pango_font_description_set_variant (description, variant);
pango_font_description_set_weight (description, weight);
gtkstyleproperty: fix small memory leak The released value is a string array and thus g_free is not sufficient. https://bugzilla.gnome.org/show_bug.cgi?id=655173
2011-07-23 13:50:05 +00:00
g_strfreev (families);
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
g_value_take_boxed (value, description);
}
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
static GParameter *
unpack_border_color (const GValue *value,
guint *n_params)
{
GParameter *parameter = g_new0 (GParameter, 4);
GType type;
type = G_VALUE_TYPE (value);
if (type == G_TYPE_PTR_ARRAY)
type = GTK_TYPE_SYMBOLIC_COLOR;
parameter[0].name = "border-top-color";
g_value_init (&parameter[0].value, type);
parameter[1].name = "border-right-color";
g_value_init (&parameter[1].value, type);
parameter[2].name = "border-bottom-color";
g_value_init (&parameter[2].value, type);
parameter[3].name = "border-left-color";
g_value_init (&parameter[3].value, type);
if (G_VALUE_TYPE (value) == G_TYPE_PTR_ARRAY)
{
GPtrArray *array = g_value_get_boxed (value);
guint i;
for (i = 0; i < 4; i++)
g_value_set_boxed (&parameter[i].value, g_ptr_array_index (array, i));
}
else
{
/* can be RGBA or symbolic color */
gpointer p = g_value_get_boxed (value);
g_value_set_boxed (&parameter[0].value, p);
g_value_set_boxed (&parameter[1].value, p);
g_value_set_boxed (&parameter[2].value, p);
g_value_set_boxed (&parameter[3].value, p);
}
*n_params = 4;
return parameter;
}
static void
pack_border_color (GValue *value,
GtkStyleProperties *props,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStateFlags state,
GtkStylePropertyContext *context)
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
{
/* NB: We are a color property, so we have to resolve to a color here.
* So we just resolve to a color. We pick one and stick to it.
* Lesson learned: Don't query border-color shorthand, query the
* real properties instead. */
g_value_unset (value);
gtk_style_properties_get_property (props, "border-top-color", state, value);
}
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
/*** UNSET FUNCS ***/
static void
unset_font_description (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "font-family", state);
gtk_style_properties_unset_property (props, "font-style", state);
gtk_style_properties_unset_property (props, "font-variant", state);
gtk_style_properties_unset_property (props, "font-weight", state);
gtk_style_properties_unset_property (props, "font-size", state);
}
static void
unset_margin (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "margin-top", state);
gtk_style_properties_unset_property (props, "margin-right", state);
gtk_style_properties_unset_property (props, "margin-bottom", state);
gtk_style_properties_unset_property (props, "margin-left", state);
}
static void
unset_padding (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "padding-top", state);
gtk_style_properties_unset_property (props, "padding-right", state);
gtk_style_properties_unset_property (props, "padding-bottom", state);
gtk_style_properties_unset_property (props, "padding-left", state);
}
static void
unset_border_width (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "border-top-width", state);
gtk_style_properties_unset_property (props, "border-right-width", state);
gtk_style_properties_unset_property (props, "border-bottom-width", state);
gtk_style_properties_unset_property (props, "border-left-width", state);
}
static void
unset_border_radius (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "border-top-right-radius", state);
gtk_style_properties_unset_property (props, "border-bottom-right-radius", state);
gtk_style_properties_unset_property (props, "border-bottom-left-radius", state);
gtk_style_properties_unset_property (props, "border-top-left-radius", state);
}
static void
unset_border_color (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "border-top-color", state);
gtk_style_properties_unset_property (props, "border-right-color", state);
gtk_style_properties_unset_property (props, "border-bottom-color", state);
gtk_style_properties_unset_property (props, "border-left-color", state);
}
static void
unset_border_image (GtkStyleProperties *props,
GtkStateFlags state)
{
gtk_style_properties_unset_property (props, "border-image-source", state);
gtk_style_properties_unset_property (props, "border-image-slice", state);
gtk_style_properties_unset_property (props, "border-image-repeat", state);
gtk_style_properties_unset_property (props, "border-image-width", state);
}
styleproperty: Add a default value func for border-image-width This is necessary because we want to use NULL as the default value. But the default value for borders is { 0, 0, 0, 0 } and not NULL. Fixes border-image-gradient and border-image-repeat reftests.
2011-06-11 07:23:34 +00:00
/*** default values ***/
static void
border_image_width_default_value (GtkStyleProperties *props,
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
GtkStateFlags state,
styleproperty: Add a default value func for border-image-width This is necessary because we want to use NULL as the default value. But the default value for borders is { 0, 0, 0, 0 } and not NULL. Fixes border-image-gradient and border-image-repeat reftests.
2011-06-11 07:23:34 +00:00
GValue *value)
{
}
styleproperty: Default to transparent background-color
2011-09-26 13:53:48 +00:00
static void
background_color_default_value (GtkStyleProperties *props,
GtkStateFlags state,
GValue *value)
{
GdkRGBA transparent_black = { 0, 0, 0, 0 };
g_value_set_boxed (value, &transparent_black);
}
styleproperty: The default border color is the normal color
2011-06-12 23:50:06 +00:00
static void
border_color_default_value (GtkStyleProperties *props,
GtkStateFlags state,
GValue *value)
{
g_value_unset (value);
gtk_style_properties_get_property (props, "color", state, value);
}
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
/*** API ***/
static void
css_string_funcs_init (void)
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
if (G_LIKELY (parse_funcs != NULL))
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return;
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
parse_funcs = g_hash_table_new (NULL, NULL);
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
print_funcs = g_hash_table_new (NULL, NULL);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GDK_TYPE_RGBA,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
rgba_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
rgba_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GDK_TYPE_COLOR,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
color_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
color_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GTK_TYPE_SYMBOLIC_COLOR,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
symbolic_color_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
symbolic_color_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (PANGO_TYPE_FONT_DESCRIPTION,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
font_description_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
font_description_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_BOOLEAN,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
boolean_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
boolean_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_INT,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
int_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
int_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_UINT,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
uint_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
uint_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_DOUBLE,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
double_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
double_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_FLOAT,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
float_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
float_value_print);
cssprovider: Add support for strings
2011-04-12 02:26:10 +00:00
register_conversion_function (G_TYPE_STRING,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
string_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
string_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GTK_TYPE_THEMING_ENGINE,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
theming_engine_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
theming_engine_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GTK_TYPE_ANIMATION_DESCRIPTION,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
animation_description_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
animation_description_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GTK_TYPE_BORDER,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
border_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
border_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (GTK_TYPE_GRADIENT,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
gradient_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
gradient_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (CAIRO_GOBJECT_TYPE_PATTERN,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
pattern_value_parse,
styleproperty: dump image surfaces as PNG data URLs We're not able to load that (yet?), but it's useful for producing diffable test dumps.
2011-07-20 22:43:00 +00:00
pattern_value_print);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
register_conversion_function (GTK_TYPE_BORDER_IMAGE,
border_image_value_parse,
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
NULL);
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
register_conversion_function (GTK_TYPE_CSS_BORDER_IMAGE_REPEAT,
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
border_image_repeat_value_parse,
border_image_repeat_value_print);
css: add a parser for GTK_TYPE_SHADOW properties https://bugzilla.gnome.org/show_bug.cgi?id=649314
2011-05-19 18:48:22 +00:00
register_conversion_function (GTK_TYPE_SHADOW,
shadow_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
shadow_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_ENUM,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
enum_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
enum_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
register_conversion_function (G_TYPE_FLAGS,
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
flags_value_parse,
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
flags_value_print);
css: start background-repeat By default, a background image is stretched. Instead, it is worth to have a tiled background. This patch allows background surfaces to be repeated or not, and should be compatible with future extensions and CSS. https://bugzilla.gnome.org/show_bug.cgi?id=663522
2011-11-06 18:33:05 +00:00
register_conversion_function (GTK_TYPE_CSS_BACKGROUND_REPEAT,
background_repeat_value_parse,
background_repeat_value_print);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
gboolean
css: Rename function to _gtk_style_property_parse_value() ... and take an optional style property as argument. This way, we can allow custom parse functions for properties. The style property needs to be optional so that we can use it for widget style properties, too.
2011-05-25 23:01:44 +00:00
_gtk_style_property_parse_value (const GtkStyleProperty *property,
GValue *value,
GtkCssParser *parser,
GFile *base)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
styleproperty: Rename ParseFunc and PrintFunc typedefs Rename them to GtkStyleParseFunc and GtkStylePrintFunc, so we can export them.
2011-05-25 23:32:01 +00:00
GtkStyleParseFunc func;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
g_return_val_if_fail (value != NULL, FALSE);
g_return_val_if_fail (parser != NULL, FALSE);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css_string_funcs_init ();
css: Move property parsing into styleproperty file Just shuffles code around for parsing properties.
2011-05-25 23:25:00 +00:00
if (property)
{
if (_gtk_css_parser_try (parser, "none", TRUE))
{
/* Insert the default value, so it has an opportunity
* to override other style providers when merged
*/
g_param_value_set_default (property->pspec, value);
return TRUE;
}
else if (property->property_parse_func)
{
GError *error = NULL;
char *value_str;
gboolean success;
value_str = _gtk_css_parser_read_value (parser);
if (value_str == NULL)
return FALSE;
success = (*property->property_parse_func) (value_str, value, &error);
g_free (value_str);
return success;
}
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
func = property->parse_func;
css: Move property parsing into styleproperty file Just shuffles code around for parsing properties.
2011-05-25 23:25:00 +00:00
}
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
else
func = NULL;
css: Move property parsing into styleproperty file Just shuffles code around for parsing properties.
2011-05-25 23:25:00 +00:00
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
if (func == NULL)
func = g_hash_table_lookup (parse_funcs,
GSIZE_TO_POINTER (G_VALUE_TYPE (value)));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (func == NULL)
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
func = g_hash_table_lookup (parse_funcs,
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
GSIZE_TO_POINTER (g_type_fundamental (G_VALUE_TYPE (value))));
if (func == NULL)
{
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
_gtk_css_parser_error (parser,
"Cannot convert to type '%s'",
g_type_name (G_VALUE_TYPE (value)));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
return FALSE;
}
css: Make property parsing functions take a css parser Instead of reading a string and then passing that in, let the parse functions use the full power of the parser.
2011-05-18 16:32:22 +00:00
return (*func) (parser, base, value);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
styleproperty: Change _css_value_to_string() Name it _gtk_style_property_print_value() and actually pass it the style property. This way, we can later change it to use custom print functions for different style properties.
2011-05-23 05:16:17 +00:00
void
_gtk_style_property_print_value (const GtkStyleProperty *property,
const GValue *value,
GString *string)
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
{
styleproperty: Rename ParseFunc and PrintFunc typedefs Rename them to GtkStyleParseFunc and GtkStylePrintFunc, so we can export them.
2011-05-25 23:32:01 +00:00
GtkStylePrintFunc func;
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
css_string_funcs_init ();
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
if (property)
func = property->print_func;
else
func = NULL;
if (func == NULL)
func = g_hash_table_lookup (print_funcs,
GSIZE_TO_POINTER (G_VALUE_TYPE (value)));
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
if (func == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
func = g_hash_table_lookup (print_funcs,
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
GSIZE_TO_POINTER (g_type_fundamental (G_VALUE_TYPE (value))));
styleproperty: Change _css_value_to_string() Name it _gtk_style_property_print_value() and actually pass it the style property. This way, we can later change it to use custom print functions for different style properties.
2011-05-23 05:16:17 +00:00
if (func == NULL)
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
{
styleproperty: Change _css_value_to_string() Name it _gtk_style_property_print_value() and actually pass it the style property. This way, we can later change it to use custom print functions for different style properties.
2011-05-23 05:16:17 +00:00
char *s = g_strdup_value_contents (value);
g_string_append (string, s);
g_free (s);
return;
css: Change to_string() to print() It's preferrable to print to an existing GString instead of returning a char* everywhere that needs to be freed later.
2011-05-23 05:02:36 +00:00
}
styleproperty: Change _css_value_to_string() Name it _gtk_style_property_print_value() and actually pass it the style property. This way, we can later change it to use custom print functions for different style properties.
2011-05-23 05:16:17 +00:00
func (value, string);
css: Overhaul value parsing Value parsing only sometimes emitted errors. Sometimes it didn't emit errors but ignored the value, sometimes it took a default, sometimes it converted it to something it deemed suitable. While refactoring, I moved the whole GValue <=> char * conversion routines to a separate file, to make navigating the core css provider easier.
2011-04-08 14:08:28 +00:00
}
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
void
_gtk_style_property_default_value (const GtkStyleProperty *property,
GtkStyleProperties *properties,
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
GtkStateFlags state,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
GValue *value)
{
if (property->default_value_func)
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
property->default_value_func (properties, state, value);
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
else if (property->pspec->value_type == GTK_TYPE_THEMING_ENGINE)
g_value_set_object (value, gtk_theming_engine_load (NULL));
else if (property->pspec->value_type == PANGO_TYPE_FONT_DESCRIPTION)
g_value_take_boxed (value, pango_font_description_from_string ("Sans 10"));
else if (property->pspec->value_type == GDK_TYPE_RGBA)
{
GdkRGBA color;
gdk_rgba_parse (&color, "pink");
g_value_set_boxed (value, &color);
}
else if (property->pspec->value_type == GTK_TYPE_BORDER)
{
g_value_take_boxed (value, gtk_border_new ());
}
else
g_param_value_set_default (property->pspec, value);
}
styleproperty: Add _gtk_style_property_is_inherit() And use it instead of gtk_style_param_get_inherit()
2011-05-31 10:05:43 +00:00
gboolean
_gtk_style_property_is_inherit (const GtkStyleProperty *property)
{
g_return_val_if_fail (property != NULL, FALSE);
Revert "API: Add API to set style properties to be inherit" Inheritance is now done using style property flags, so thereis no ned anymore to clobber the pspec flags namespace. Also, there is no need to expose this functionality in the public API without exposing more of the styleproperty API. This essentially reverts commit 9bfd3d2eec96cba146fc9f5e1ab972584672e3d5.
2011-05-31 10:42:30 +00:00
return property->flags & GTK_STYLE_PROPERTY_INHERIT ? TRUE : FALSE;
styleproperty: Add _gtk_style_property_is_inherit() And use it instead of gtk_style_param_get_inherit()
2011-05-31 10:05:43 +00:00
}
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
static gboolean
resolve_color (GtkStyleProperties *props,
GValue *value)
{
GdkRGBA color;
/* Resolve symbolic color to GdkRGBA */
if (!gtk_symbolic_color_resolve (g_value_get_boxed (value), props, &color))
return FALSE;
/* Store it back, this is where GdkRGBA caching happens */
g_value_unset (value);
g_value_init (value, GDK_TYPE_RGBA);
g_value_set_boxed (value, &color);
return TRUE;
}
static gboolean
resolve_color_rgb (GtkStyleProperties *props,
GValue *value)
{
GdkColor color = { 0 };
GdkRGBA rgba;
if (!gtk_symbolic_color_resolve (g_value_get_boxed (value), props, &rgba))
return FALSE;
color.red = rgba.red * 65535. + 0.5;
color.green = rgba.green * 65535. + 0.5;
color.blue = rgba.blue * 65535. + 0.5;
g_value_unset (value);
g_value_init (value, GDK_TYPE_COLOR);
g_value_set_boxed (value, &color);
return TRUE;
}
Add initial cut at win32 theme support for CSS We now support -gtk-win32-theme-part(class,part,state) in background and border-image CSS properties. This renders the corresponding theme part using DrawThemeBackground() and acts as a base for a CSS based windows theme. Note that we build the parsing code even on non-win32 so that all themese will parse the same on all arches. We draw pink instead of the actual theme parts on non-win32 though.
2011-11-15 19:52:45 +00:00
static gboolean
resolve_win32_theme_part (GtkStyleProperties *props,
GValue *value,
GValue *value_out,
GtkStylePropertyContext *context)
{
GtkWin32ThemePart *part;
cairo_pattern_t *pattern;
part = g_value_get_boxed (value);
if (part == NULL)
return FALSE;
pattern = _gtk_win32_theme_part_render (part, context->width, context->height);
g_value_take_boxed (value_out, pattern);
return TRUE;
}
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
static gboolean
resolve_gradient (GtkStyleProperties *props,
GValue *value)
{
cairo_pattern_t *gradient;
if (!gtk_gradient_resolve (g_value_get_boxed (value), props, &gradient))
return FALSE;
/* Store it back, this is where cairo_pattern_t caching happens */
g_value_unset (value);
g_value_init (value, CAIRO_GOBJECT_TYPE_PATTERN);
g_value_take_boxed (value, gradient);
return TRUE;
}
static gboolean
resolve_shadow (GtkStyleProperties *props,
GValue *value)
{
GtkShadow *resolved, *base;
base = g_value_get_boxed (value);
if (base == NULL)
return TRUE;
if (_gtk_shadow_get_resolved (base))
return TRUE;
resolved = _gtk_shadow_resolve (base, props);
if (resolved == NULL)
return FALSE;
g_value_take_boxed (value, resolved);
return TRUE;
}
void
_gtk_style_property_resolve (const GtkStyleProperty *property,
GtkStyleProperties *props,
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
GtkStateFlags state,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStylePropertyContext *context,
make _gtk_style_property_resolve copy the result to an output value This is preparation for allowing it to return a newly created value, rather than just copying one.
2011-11-15 16:15:49 +00:00
GValue *val,
GValue *val_out)
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
{
if (G_VALUE_TYPE (val) == GTK_TYPE_SYMBOLIC_COLOR)
{
if (property->pspec->value_type == GDK_TYPE_RGBA)
{
styleproperty: Fall back to default values, instead of recursing Copy/paste error. The original commit message even explains this.
2011-06-11 08:02:58 +00:00
if (resolve_color (props, val))
make _gtk_style_property_resolve copy the result to an output value This is preparation for allowing it to return a newly created value, rather than just copying one.
2011-11-15 16:15:49 +00:00
goto out;
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
else if (property->pspec->value_type == GDK_TYPE_COLOR)
{
styleproperty: Fall back to default values, instead of recursing Copy/paste error. The original commit message even explains this.
2011-06-11 08:02:58 +00:00
if (resolve_color_rgb (props, val))
make _gtk_style_property_resolve copy the result to an output value This is preparation for allowing it to return a newly created value, rather than just copying one.
2011-11-15 16:15:49 +00:00
goto out;
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
styleproperty: Fall back to default values, instead of recursing Copy/paste error. The original commit message even explains this.
2011-06-11 08:02:58 +00:00
g_value_unset (val);
g_value_init (val, property->pspec->value_type);
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
_gtk_style_property_default_value (property, props, state, val);
styleproperty: Resolve NULL RGBAs to pink .. instead of returning NULL and crashing. Fixes no-colors reftest
2011-06-11 08:44:16 +00:00
}
else if (G_VALUE_TYPE (val) == GDK_TYPE_RGBA)
{
if (g_value_get_boxed (val) == NULL)
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
_gtk_style_property_default_value (property, props, state, val);
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
else if (G_VALUE_TYPE (val) == GTK_TYPE_GRADIENT)
{
g_return_if_fail (property->pspec->value_type == CAIRO_GOBJECT_TYPE_PATTERN);
if (!resolve_gradient (props, val))
styleproperty: Fall back to default values, instead of recursing Copy/paste error. The original commit message even explains this.
2011-06-11 08:02:58 +00:00
{
g_value_unset (val);
g_value_init (val, CAIRO_GOBJECT_TYPE_PATTERN);
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
_gtk_style_property_default_value (property, props, state, val);
styleproperty: Fall back to default values, instead of recursing Copy/paste error. The original commit message even explains this.
2011-06-11 08:02:58 +00:00
}
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
else if (G_VALUE_TYPE (val) == GTK_TYPE_SHADOW)
{
if (!resolve_shadow (props, val))
styleproperty: Make default_value() take a state Because of this, resolve() needs to take a state, too. This is so that we can have default values depend on other values (like border color).
2011-06-12 23:26:11 +00:00
_gtk_style_property_default_value (property, props, state, val);
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
Add initial cut at win32 theme support for CSS We now support -gtk-win32-theme-part(class,part,state) in background and border-image CSS properties. This renders the corresponding theme part using DrawThemeBackground() and acts as a base for a CSS based windows theme. Note that we build the parsing code even on non-win32 so that all themese will parse the same on all arches. We draw pink instead of the actual theme parts on non-win32 though.
2011-11-15 19:52:45 +00:00
else if (G_VALUE_TYPE (val) == GTK_TYPE_WIN32_THEME_PART)
{
if (resolve_win32_theme_part (props, val, val_out, context))
return; /* Don't copy val, this sets val_out */
_gtk_style_property_default_value (property, props, state, val);
}
make _gtk_style_property_resolve copy the result to an output value This is preparation for allowing it to return a newly created value, rather than just copying one.
2011-11-15 16:15:49 +00:00
out:
g_value_copy (val, val_out);
styleproperty: Move resolving code Also, resolving now always succeeds - if it doesn't we fall back to the default value right here, instead of later.
2011-05-31 14:51:38 +00:00
}
style: Add support for shorthand properties Shorthand properties are basically the same a in CSS. For storage in style properties or the CSS provider, they are unpacked into the real values, so it is possible to partially override them. No properties are yet converted to the new world yet, this is just the code for supporting them.
2011-05-21 21:47:19 +00:00
gboolean
_gtk_style_property_is_shorthand (const GtkStyleProperty *property)
{
g_return_val_if_fail (property != NULL, FALSE);
return property->pack_func != NULL;
}
GParameter *
_gtk_style_property_unpack (const GtkStyleProperty *property,
const GValue *value,
guint *n_params)
{
g_return_val_if_fail (property != NULL, NULL);
g_return_val_if_fail (property->unpack_func != NULL, NULL);
g_return_val_if_fail (value != NULL, NULL);
g_return_val_if_fail (n_params != NULL, NULL);
return property->unpack_func (value, n_params);
}
void
_gtk_style_property_pack (const GtkStyleProperty *property,
GtkStyleProperties *props,
GtkStateFlags state,
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
GtkStylePropertyContext *context,
style: Add support for shorthand properties Shorthand properties are basically the same a in CSS. For storage in style properties or the CSS provider, they are unpacked into the real values, so it is possible to partially override them. No properties are yet converted to the new world yet, this is just the code for supporting them.
2011-05-21 21:47:19 +00:00
GValue *value)
{
g_return_if_fail (property != NULL);
g_return_if_fail (property->pack_func != NULL);
g_return_if_fail (GTK_IS_STYLE_PROPERTIES (props));
g_return_if_fail (G_IS_VALUE (value));
Add GtkStylePropertyContext and use it in when getting style properties At the toplevel we have _gtk_theming_engine_get, which lets us pass in a property context with (atm) the size to get the property for. Then there is a lot of plumbing to push this down into the lower layers of the style property code until finally hitting the property resolvers. I need this because I will be adding a property resolver for win32 theme parts, and they render differently depending on the size (i.e. they don't scale linearly). The idea is that the code to get the background properties will pass in the final size and we will resolve the theme part specification to that particular size. If the old non-context calls are used we just hardcode a size of 100x100.
2011-11-15 15:56:59 +00:00
property->pack_func (value, props, state, context);
style: Add support for shorthand properties Shorthand properties are basically the same a in CSS. For storage in style properties or the CSS provider, they are unpacked into the real values, so it is possible to partially override them. No properties are yet converted to the new world yet, this is just the code for supporting them.
2011-05-21 21:47:19 +00:00
}
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
static void
gtk_style_property_init (void)
{
if (G_LIKELY (properties))
return;
/* stuff is never freed, so no need for free functions */
properties = g_hash_table_new (g_str_hash, g_str_equal);
/* note that gtk_style_properties_register_property() calls this function,
* so make sure we're sanely inited to avoid infloops */
styleproperty: Use the new flags for inheritance Instead of gtk_style_param_set_inherit()
2011-05-31 10:35:42 +00:00
_gtk_style_property_register (g_param_spec_boxed ("color",
"Foreground color",
"Foreground color",
GDK_TYPE_RGBA, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Use the new flags for inheritance Instead of gtk_style_param_set_inherit()
2011-05-31 10:35:42 +00:00
NULL);
styleproperty: Use the long form to register background-color We want to set a default value.
2011-09-26 13:48:53 +00:00
_gtk_style_property_register (g_param_spec_boxed ("background-color",
"Background color",
"Background color",
GDK_TYPE_RGBA, 0),
0,
NULL,
NULL,
NULL,
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse,
styleproperty: Use the long form to register background-color We want to set a default value.
2011-09-26 13:48:53 +00:00
NULL,
styleproperty: Default to transparent background-color
2011-09-26 13:53:48 +00:00
background_color_default_value,
styleproperty: Use the long form to register background-color We want to set a default value.
2011-09-26 13:48:53 +00:00
NULL);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
_gtk_style_property_register (g_param_spec_boxed ("font-family",
"Font family",
"Font family",
G_TYPE_STRV, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
font_family_parse,
font_family_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
NULL);
_gtk_style_property_register (g_param_spec_enum ("font-style",
"Font style",
"Font style",
PANGO_TYPE_STYLE,
PANGO_STYLE_NORMAL, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
NULL);
_gtk_style_property_register (g_param_spec_enum ("font-variant",
"Font variant",
"Font variant",
PANGO_TYPE_VARIANT,
PANGO_VARIANT_NORMAL, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
NULL);
/* xxx: need to parse this properly, ie parse the numbers */
_gtk_style_property_register (g_param_spec_enum ("font-weight",
"Font weight",
"Font weight",
PANGO_TYPE_WEIGHT,
PANGO_WEIGHT_NORMAL, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
NULL);
_gtk_style_property_register (g_param_spec_double ("font-size",
"Font size",
"Font size",
0, G_MAXDOUBLE, 0, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
NULL);
styleproperty: Use the new flags for inheritance Instead of gtk_style_param_set_inherit()
2011-05-31 10:35:42 +00:00
_gtk_style_property_register (g_param_spec_boxed ("font",
"Font Description",
"Font Description",
PANGO_TYPE_FONT_DESCRIPTION, 0),
GTK_STYLE_PROPERTY_INHERIT,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant
2011-05-29 05:03:59 +00:00
NULL,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant This is the second try at this. The first was backed out previously due to bugginess. Let's hope this one survives a bit longer. Also makes the font-family CSS test work again.
2011-05-29 05:03:59 +00:00
unpack_font_description,
pack_font_description,
css: Make font property a shorthand ... and implement the CSS font properties: - font-size - font-style - font-family - font-weight - font-variant
2011-05-29 05:03:59 +00:00
font_description_value_parse,
font_description_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_font_description);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
styleproperty: Use the new flags for inheritance Instead of gtk_style_param_set_inherit()
2011-05-31 10:35:42 +00:00
_gtk_style_property_register (g_param_spec_boxed ("text-shadow",
"Text shadow",
"Text shadow",
GTK_TYPE_SHADOW, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Use the new flags for inheritance Instead of gtk_style_param_set_inherit()
2011-05-31 10:35:42 +00:00
NULL);
styleproperties: add an 'icon-shadow' style property It's a custom property with the same semantics of text-shadow, which works on icons and icon-like UI elements, such as spinners, arrows and separators.
2011-05-21 03:48:36 +00:00
_gtk_style_property_register (g_param_spec_boxed ("icon-shadow",
"Icon shadow",
"Icon shadow",
GTK_TYPE_SHADOW, 0),
GTK_STYLE_PROPERTY_INHERIT,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: add an 'icon-shadow' style property It's a custom property with the same semantics of text-shadow, which works on icons and icon-like UI elements, such as spinners, arrows and separators.
2011-05-21 03:48:36 +00:00
NULL);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
styleproperty: register the box-shadow property
2011-06-02 00:59:14 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("box-shadow",
"Box shadow",
"Box shadow",
GTK_TYPE_SHADOW, 0));
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
style: Make margin a shorthand ... and add the real properties: - margin-top - margin-left - margin-bottom - margin-right
2011-05-22 03:55:12 +00:00
g_param_spec_int ("margin-top",
"margin top",
"Margin at top",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("margin-left",
"margin left",
"Margin at left",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("margin-bottom",
"margin bottom",
"Margin at bottom",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("margin-right",
"margin right",
"Margin at right",
0, G_MAXINT, 0, 0));
_gtk_style_property_register (g_param_spec_boxed ("margin",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Margin",
"Margin",
style: Make margin a shorthand ... and add the real properties: - margin-top - margin-left - margin-bottom - margin-right
2011-05-22 03:55:12 +00:00
GTK_TYPE_BORDER, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
style: Make margin a shorthand ... and add the real properties: - margin-top - margin-left - margin-bottom - margin-right
2011-05-22 03:55:12 +00:00
NULL,
unpack_margin,
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
pack_margin,
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_margin);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
style: Make padding a shorthand ... and add the real properties: - padding-top - padding-left - padding-bottom - padding-right
2011-05-22 03:42:17 +00:00
g_param_spec_int ("padding-top",
"padding top",
"Padding at top",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("padding-left",
"padding left",
"Padding at left",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("padding-bottom",
"padding bottom",
"Padding at bottom",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("padding-right",
"padding right",
"Padding at right",
0, G_MAXINT, 0, 0));
_gtk_style_property_register (g_param_spec_boxed ("padding",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Padding",
"Padding",
style: Make padding a shorthand ... and add the real properties: - padding-top - padding-left - padding-bottom - padding-right
2011-05-22 03:42:17 +00:00
GTK_TYPE_BORDER, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
style: Make padding a shorthand ... and add the real properties: - padding-top - padding-left - padding-bottom - padding-right
2011-05-22 03:42:17 +00:00
NULL,
unpack_padding,
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
pack_padding,
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_padding);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
style: Convert border-width to a shorthand Add the corresponding properties it refers to, too: - border-top-width - border-left-width - border-bottom-width - border-right-width
2011-05-21 22:10:43 +00:00
g_param_spec_int ("border-top-width",
"border top width",
"Border width at top",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("border-left-width",
"border left width",
"Border width at left",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("border-bottom-width",
"border bottom width",
"Border width at bottom",
0, G_MAXINT, 0, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_int ("border-right-width",
"border right width",
"Border width at right",
0, G_MAXINT, 0, 0));
_gtk_style_property_register (g_param_spec_boxed ("border-width",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Border width",
"Border width, in pixels",
style: Convert border-width to a shorthand Add the corresponding properties it refers to, too: - border-top-width - border-left-width - border-bottom-width - border-right-width
2011-05-21 22:10:43 +00:00
GTK_TYPE_BORDER, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
style: Convert border-width to a shorthand Add the corresponding properties it refers to, too: - border-top-width - border-left-width - border-bottom-width - border-right-width
2011-05-21 22:10:43 +00:00
NULL,
unpack_border_width,
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
pack_border_width,
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_border_width);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-top-left-radius",
"Border top left radius",
"Border radius of top left corner, in pixels",
GTK_TYPE_CSS_BORDER_CORNER_RADIUS, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
NULL,
NULL,
NULL,
border_corner_radius_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
border_corner_radius_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-top-right-radius",
"Border top right radius",
"Border radius of top right corner, in pixels",
GTK_TYPE_CSS_BORDER_CORNER_RADIUS, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
NULL,
NULL,
NULL,
border_corner_radius_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
border_corner_radius_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-bottom-right-radius",
"Border bottom right radius",
"Border radius of bottom right corner, in pixels",
GTK_TYPE_CSS_BORDER_CORNER_RADIUS, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
NULL,
NULL,
NULL,
border_corner_radius_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
border_corner_radius_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-bottom-left-radius",
"Border bottom left radius",
"Border radius of bottom left corner, in pixels",
GTK_TYPE_CSS_BORDER_CORNER_RADIUS, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
NULL,
NULL,
NULL,
border_corner_radius_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
border_corner_radius_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
_gtk_style_property_register (g_param_spec_int ("border-radius",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Border radius",
"Border radius, in pixels",
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
0, G_MAXINT, 0, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
NULL,
unpack_border_radius,
pack_border_radius,
border_radius_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
border_radius_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_border_radius);
css: Add all border-radius properties Implement all border-radisu properties as specified by http://www.w3.org/TR/css3-background/#the-border-radius The end goal here is to get joined buttons to really look joined.
2011-05-26 01:50:53 +00:00
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_enum ("border-style",
"Border style",
"Border style",
GTK_TYPE_BORDER_STYLE,
GTK_BORDER_STYLE_NONE, 0));
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-top-color",
"Border top color",
"Border top color",
GDK_TYPE_RGBA, 0),
0,
NULL,
NULL,
NULL,
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse,
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
border_color_default_value,
NULL);
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-right-color",
"Border right color",
"Border right color",
GDK_TYPE_RGBA, 0),
0,
NULL,
NULL,
NULL,
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse,
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
border_color_default_value,
NULL);
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-bottom-color",
"Border bottom color",
"Border bottom color",
styleproperty: The default border color is the normal color
2011-06-12 23:50:06 +00:00
GDK_TYPE_RGBA, 0),
0,
NULL,
NULL,
NULL,
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse,
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
border_color_default_value,
NULL);
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-left-color",
"Border left color",
"Border left color",
GDK_TYPE_RGBA, 0),
0,
NULL,
styleproperty: The default border color is the normal color
2011-06-12 23:50:06 +00:00
NULL,
NULL,
styleproperty: Allow 'transparent' keyword for background-color
2011-09-26 13:56:46 +00:00
transparent_color_value_parse,
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
border_color_default_value,
NULL);
styleproperty: Add border-*-color style properties And make border-color a shorthand
2011-06-13 01:45:18 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-color",
"Border color",
"Border color",
GDK_TYPE_RGBA, 0),
0,
NULL,
unpack_border_color,
pack_border_color,
border_color_shorthand_value_parse,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_border_color);
styleproperty: The default border color is the normal color
2011-06-12 23:50:06 +00:00
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("background-image",
"Background Image",
"Background Image",
CAIRO_GOBJECT_TYPE_PATTERN, 0));
css: start background-repeat By default, a background image is stretched. Instead, it is worth to have a tiled background. This patch allows background surfaces to be repeated or not, and should be compatible with future extensions and CSS. https://bugzilla.gnome.org/show_bug.cgi?id=663522
2011-11-06 18:33:05 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("background-repeat",
"Background repeat",
"Background repeat",
GTK_TYPE_CSS_BACKGROUND_REPEAT, 0));
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
g_param_spec_boxed ("border-image-source",
"Border image source",
"Border image source",
CAIRO_GOBJECT_TYPE_PATTERN, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("border-image-repeat",
"Border image repeat",
"Border image repeat",
border-image: move private types to gtkcsstypes.c
2011-06-05 20:49:54 +00:00
GTK_TYPE_CSS_BORDER_IMAGE_REPEAT, 0));
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("border-image-slice",
"Border image slice",
"Border image slice",
GTK_TYPE_BORDER, 0));
styleproperty: Add a default value func for border-image-width This is necessary because we want to use NULL as the default value. But the default value for borders is { 0, 0, 0, 0 } and not NULL. Fixes border-image-gradient and border-image-repeat reftests.
2011-06-11 07:23:34 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-image-width",
border-image: add support for border-image-width too It's useful to set a slice size != border-width, as backgrounds are clipped to border-width too. As slices can be half-transparent and overlap the background, this would not fill the border box properly if we only use a single property for specifying the width. Also, this brings us even closer to CSS3.
2011-06-09 23:59:14 +00:00
"Border image width",
"Border image width",
styleproperty: Add a default value func for border-image-width This is necessary because we want to use NULL as the default value. But the default value for borders is { 0, 0, 0, 0 } and not NULL. Fixes border-image-gradient and border-image-repeat reftests.
2011-06-11 07:23:34 +00:00
GTK_TYPE_BORDER, 0),
0,
NULL,
NULL,
NULL,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
border_image_width_default_value,
NULL);
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
_gtk_style_property_register (g_param_spec_boxed ("border-image",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Border Image",
"Border Image",
styleproperty: turn border-image into a shorthand property It's composed by - border-image-source: a cairo_pattern_t holding an image from file or a gradient - border-image-slice: a GtkBorder containing the slice offsets to apply on the image - border-image-repeat: a GtkRepeatType for the image We deviate from pure CSS3 in the following ways: * border-image-width is assumed to be always 1, i.e. always equal to what's specified by border-width. I don't think it's a particularly useful property to have, but we could add it later if needed. * border-image-outset is absent, as we can't render outside of the allocation yet. https://bugzilla.gnome.org/show_bug.cgi?id=651194
2011-05-26 19:54:38 +00:00
GTK_TYPE_BORDER_IMAGE, 0),
0,
NULL,
_gtk_border_image_unpack,
_gtk_border_image_pack,
NULL,
NULL,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperties: Add unset functions for shorthands This fixes calls to gtk_style_properties_unset_property() for shorthand properties. https://bugzilla.gnome.org/show_bug.cgi?id=654678
2011-07-19 09:42:31 +00:00
unset_border_image);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_properties_register_property (NULL,
g_param_spec_object ("engine",
"Theming Engine",
"Theming Engine",
GTK_TYPE_THEMING_ENGINE, 0));
gtk_style_properties_register_property (NULL,
g_param_spec_boxed ("transition",
"Transition animation description",
"Transition animation description",
GTK_TYPE_ANIMATION_DESCRIPTION, 0));
/* Private property holding the binding sets */
styleproperty: Make keybindings use custom parse/print functions We don't want to use it for GPtrArray.
2011-05-25 23:50:41 +00:00
_gtk_style_property_register (g_param_spec_boxed ("gtk-key-bindings",
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
"Key bindings",
"Key bindings",
styleproperty: Make keybindings use custom parse/print functions We don't want to use it for GPtrArray.
2011-05-25 23:50:41 +00:00
G_TYPE_PTR_ARRAY, 0),
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
0,
styleproperty: Make keybindings use custom parse/print functions We don't want to use it for GPtrArray.
2011-05-25 23:50:41 +00:00
NULL,
NULL,
NULL,
bindings_value_parse,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
bindings_value_print,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
NULL,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
NULL);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
}
const GtkStyleProperty *
_gtk_style_property_lookup (const char *name)
{
gtk_style_property_init ();
return g_hash_table_lookup (properties, name);
}
void
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
_gtk_style_property_register (GParamSpec *pspec,
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
GtkStylePropertyFlags flags,
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
GtkStylePropertyParser property_parse_func,
GtkStyleUnpackFunc unpack_func,
GtkStylePackFunc pack_func,
GtkStyleParseFunc parse_func,
GtkStylePrintFunc print_func,
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
GtkStyleDefaultValueFunc default_value_func,
GtkStyleUnsetFunc unset_func)
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
{
const GtkStyleProperty *existing;
GtkStyleProperty *node;
style: Add support for shorthand properties Shorthand properties are basically the same a in CSS. For storage in style properties or the CSS provider, they are unpacked into the real values, so it is possible to partially override them. No properties are yet converted to the new world yet, this is just the code for supporting them.
2011-05-21 21:47:19 +00:00
g_return_if_fail ((pack_func == NULL) == (unpack_func == NULL));
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
gtk_style_property_init ();
existing = _gtk_style_property_lookup (pspec->name);
if (existing != NULL)
{
g_warning ("Property \"%s\" was already registered with type %s",
pspec->name, g_type_name (existing->pspec->value_type));
return;
}
node = g_slice_new0 (GtkStyleProperty);
styleproperty: Add flags This allows to set a bunch of specific features, like inheritance in here, without having to molest the psec.
2011-05-31 10:20:47 +00:00
node->flags = flags;
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
node->pspec = pspec;
css: Rename struct variable I want to add parse_func as something different soon.
2011-05-25 22:09:26 +00:00
node->property_parse_func = property_parse_func;
style: Add support for shorthand properties Shorthand properties are basically the same a in CSS. For storage in style properties or the CSS provider, they are unpacked into the real values, so it is possible to partially override them. No properties are yet converted to the new world yet, this is just the code for supporting them.
2011-05-21 21:47:19 +00:00
node->pack_func = pack_func;
node->unpack_func = unpack_func;
styleproperty: Allow passing in a parse func and a print func It's not used yet, but it's now possible to specify a custom read or write func.
2011-05-25 23:46:49 +00:00
node->parse_func = parse_func;
node->print_func = print_func;
styleproperty: Move default value setting to styleproperty.c
2011-05-28 03:43:07 +00:00
node->default_value_func = default_value_func;
styleproperties: Add an unset vfunc to style properties This makes unsetting work for shorthands.
2011-07-19 09:11:12 +00:00
node->unset_func = unset_func;
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
Fix warning
2011-08-06 16:56:01 +00:00
/* pspec owns name */
g_hash_table_insert (properties, (gchar *)pspec->name, node);
styleproperties: Move the style property handling code ... from gtkstyleproperties.c to gtkstyleproperty.c
2011-05-21 02:17:28 +00:00
}
Reference in New Issue Copy Permalink