This makes sure that we don't have cursors disappearing on Windows upon
scrolling because we can't find a cursor that exists on the system during
a scroll, and unlike GTK-3.x, we do not default to the arrow pointer on GTK4.
Just mimic what we have on X11 and Wayland: the trusty standard arrow pointer.
Fixes issue #3581.
Instead of now-unused GdkWin32Cursor class (a subclass of GdkCursor),
add a stand-alone GdkWin32HCursor class that is a wrapper around
HCURSOR handle.
On creation it's given a display instance, a HCURSOR handle and a boolean
that indicates whether the HCURSOR handle can or cannot be destroyed
(this depends on how the handle was obtained).
That information is stored in a hash table inside the GdkWin32Display
singleton, each entry of that table has reference count.
When the GdkWin32HCursor object is finalized, it reduces the reference
count on the table entry in the GdkWin32Display. When it's created,
it either adds such an entry or refs an existing one.
This way two pieces of code (or the same piece of code called
multiple times) that independently obtain the same HCURSOR from the OS
will get to different GdkWin32HCursor instances, but GdkWin32Display
will know that both use the same handle.
Once the reference count reaches 0 on the table entry, it is freed
and the handle (if destroyable) is put on the destruction list,
and an idle destruction function is queued.
If the same handle is once again registered for use before the
idle destructior is invoked (this happens, for example, when
an old cursor is destroyed and then replaced with a new one),
the handle gets removed from the destruction list.
The destructor just calls DestroyCursor() on each handle, calling
SetCursor(NULL) before doing that when the handle is in use.
This ensures that SetCursor(NULL) (which will cause cursor to disappear,
which is bad by itself, and which will also cause flickering if the
cursor is set to a non-NULL again shortly afterward)
is almost never called, unless GTK messes up and keeps using a cursor
beyond its lifetime.
This scheme also ensures that non-destructable cursors are not destroyed.
It's also possible to call _gdk_win32_display_hcursor_ref()
and _gdk_win32_display_hcursor_unref() manually instead of creating
GdkWin32HCursor objects, but that is not recommended.
Rename all *window.[ch] source files.
This is an automatic operation, done by the following commands:
for i in $(git ls-files gdk | grep window); do
git mv $i $(echo $i | sed s/window/surface/);
git sed -f g $(basename $i) $(basename $i | sed s/window/surface/) ;
done
git checkout NEWS* po-properties po
Like the X11 and Wayland backends, re-work how the cursors are being
handled. So, we use a hash table to cache up the HCURSORS that we
create along the way.
We still need to cache up the icon/cursor themes since this is something
that is not part of Windows but was added on to support icon/cursor themes
such as Adwaita on Windows, but should be in-line with what is going on in
GdkCursor.
Also, remove the _gdk_grab_cursor global variable in gdkprivate-win32.h,
and replace it with another variable in the GdkWin32Display structure,
to make things cleaner in the process.
https://bugzilla.gnome.org/show_bug.cgi?id=773299
Load themed cursors from the same places they are loaded on freedesktop systems,
but use W32 API functions to do so (works for .cur/.ani cursors instead of X
cursors).
Refactor the code for cursor handling. Prefer loading cursors by name.
Do not load actual cursors when loading the theme. Find the files and remember
the arguments/calls for loading them instead. Keeping HCURSOR instance in the
hashmap would result in multiple GdkCursors using the same HCURSOR. Given that
we use DestroyCursor() to off them, this would cause problems (at the very
least - DestroyCursor() would fail).
Store GdkCursor instances in a cache. Update cached cursors when theme changes.
Recognize "system" theme as a special (and default) case. When it is set,
prefer system cursors and fall back to Adwaita cursors and (as a last resort)
built-in X cursors. Otherwise prefer theme cursors and fall back to system and
X cursors.
Force GTK to use "left_ptr" cursor when no cursor is set. Using NULL makes
it use the system default "arrow", which is not the intended behaviour when
a non-system theme is selected.
Ignore cursor size setting and query the OS for the required cursor size, as
Windows (almost) does not allow setting cursors of arbitrary size.
https://bugzilla.gnome.org/show_bug.cgi?id=749287
This is purely to support gdk_cursor_new_from_name().
In particular, its counterpart, gdk_cursor_new_for_display(), will not
be affected, because there's no GDK_LEFT_PTR_WATCH cursor type,
and because i don't have a fallback cursor bitmask for gdk/win32/xcursors.h
This reverts commit 24d3f3fcb2.
Sorry, I am going to re-commit this very shortly with a new
commit message, as I found the commit message to be quite
wrong and misleading.
Due to the work on gdk_cursor_new_from_surface (commit b2113b73),
get_cursor_for_pixbuf() in GdkDisplayClass was converted to
get_cursor_for_surface(), which means the GDK Win32 backend needs to be
updated for the code to build and run on Windows, plus some function
prototypes and declarations/calls need to be updated as well.
https://bugzilla.gnome.org/show_bug.cgi?id=705980
We want a surface so we can properly represent the scale factor for it.
All backends are converted to use surfaces and we reimplement the
backwards compat code in the generic code.
Include gdkwin32.h (which includes gdkprivate-win32.h and gdkwin32cursor.h
during the build of GDK-Win32) so that
gdk_win32_icon_to_pixbuf_libgtk_only() and
gdk_win32_pixbuf_to_hicon_libgtk_only() get exported, so that the GTK
DLL can link correctly.
There are sure regressions but basic stuff seems to be working
again after all the API breakage done with comments like
"Win32 and Quartz need to be ported still."
2009-01-21 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c: Rename static local functions and
variables to not have any unnecessary _ or _gdk_win32 prefix.
* gdk/win32/gdkcursor-win32.c (hcursor_from_type): Implement
creating a GDK_BLANK_CURSOR.
svn path=/trunk/; revision=22165
2008-08-05 Tor Lillqvist <tml@novell.com>
Bug 544684 - Win64 issue, window handles are assumed to be 32-bit
* gdk/win32/gdkcursor-win32.c
* gdk/win32/xcursors.h: Change some gchar* to guchar* and vice
versa to avoid gcc 4.4 signedness warnings.
* gdk/win32/gdkevents-win32.c: Add some guchar and char pointer
casts to get rid of gcc 4.4 signedness warnings. Print GdkAtom
values in debugging output using the %p format.
* gdk/win32/gdkkeys-win32.c
* gdk/win32/gdkfont-win32.c
* gdk/win32/gdkmain-win32.c: Add some casts to avoid gcc warnings.
* gdk/win32/gdkwindow-win32.c: Use SetWindowLongPtr() instead of
SetWindowLong().
* gdk/win32/gdkwin32id.c (gdk_handle_hash): Use all 64 bits of a
HANDLE on Win64.
svn path=/trunk/; revision=20994
2007-02-04 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c: Don't define BITMAPV5HEADER on
mingw with w32api >= 3.8, which has it in wingdi.h. (#403896)
svn path=/trunk/; revision=17256
2006-10-29 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c
(gdk_win32_icon_to_pixbuf_libgtk_only): Implement for B&W cursors,
for instance the built-in GDK ones.
2006-08-29 Tor Lillqvist <tml@novell.com>
Remove support for Windows 9x/ME. GTK+ hasn't worked on Win9x
since 2.6 or 2.8. It's pointless to keep the Win9x code in here as
it isn't being maintained anyway. If somebody is interested, it
can always be found in older GTK+ versions, and in CVS.
* gdk/win32/gdkcursor-win32.c
* gdk/win32/gdkdnd-win32.c
* gdk/win32/gdkdrawable-win32.c
* gdk/win32/gdkgc-win32.c
* gdk/win32/gdkglobals-win32.c
* gdk/win32/gdkkeys-win32.c
* gdk/win32/gdkmain-win32.c
* gdk/win32/gdkproperty-win32.c
* gdk/win32/gdkselection-win32.c: Remove the G_WIN32_IS_NT_BASED()
and G_WIN32_HAVE_WIDECHAR_API() tests and their false (Win9x)
branches, and any variables or static functions used only by the
Win9x branches.
* gdk/win32/gdkprivate-win32.h: Remove backup definitions for
constants that aren't missing from current mingw and MSVC6
headers.
* gdk/win32/gdkmain-win32.c
* gdk/win32/gdkprivate-win32.h: Remove the _gdk_win32_gdi_failed()
function. On NT-based Windows GetLastError() returns error codes
also for failed GDI calls, so we can use _gdk_win32_api_failed()
always.
2006-04-06 Tor Lillqvist <tml@novell.com>
Correct the appearance of non-square icons and cursors: pad them
to square so that Windows won't stretch them.
* gdk/win32/gdkcursor-win32.c (create_alpha_bitmap)
(create_color_bitmap): Always create square bitmaps, take only
side length as argument.
(pixbuf_to_hbitmaps_alpha_winxp, pixbuf_to_hbitmaps_normal):
Corresponding changes.
2005-11-07 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c (pixbuf_to_hbitmaps_normal):
Correct the calculation of maskstride. (#320152, Peter Zelezny)
2005-11-06 Tor Lillqvist <tml@novell.com>
(pixbuf_to_hbitmaps_alpha_winxp): Use an 1-bit mask bitmap,
like pixbuf_to_hbitmaps_normal().
(_gdk_win32_pixbuf_to_hicon_supports_alpha): Check
G_WIN32_IS_NT_BASED() first, so we can pretend being on Win9x by
setting the G_WIN32_PRETEND_WIN9X environment variable.
2005-11-06 Tor Lillqvist <tml@novell.com>
Make icon masks work on Win98 (#320152, Peter Zelezny)
* gdk/win32/gdkcursor-win32.c (create_color_bitmap): Take also a
parameter for the depth of the bitmap, so that this function can
be used to create 1-bit bitmaps, too.
(pixbuf_to_hbitmaps_normal): Create an 1-bit bitmap for the mask,
and initialize it properly.
2005-11-03 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c: As there is only one GdkDisplay in
the Win32 backend, check that GdkDisplay* parameters are equal to
_gdk_display instead of using the unnecessarily general
GDK_IS_DISPLAY().
2005-11-03 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c (pixbuf_to_hbitmaps_alpha_winxp):
Rename the variables for the color bitmap to have "color" in their
name, for similarity with pixbuf_to_hbitmaps_normal(). Create a
color bitmap for the mask, too, instead of creating a b&w bitmap
with CreateBitmap(). Set up the mask bitmap's contents, ones for
those pixels in the color bitmap where the alpha is zero, zero for
other pixels. We used to use an unitialized mask bitmap! This
meant that icons and cursors created presumably worked more or
less by accident. Totally blank icons with zero alpha everywhere
(as used by gtktrayicon.c) definitely did not work as expected.
2005-09-18 Hans Breuer <hans@breuer.org>
* gdk/gdk.symbols gdk/win32/gdkwin32.h gdk/win32/gdkcursorwin32.c : export
_gdk_win32_pixbuf_to_hicon() as gdk_win32_pixbuf_to_hicon_libgtk_only()
* gtk/gtktrayicon-win32.c : a lot of debug spew but it has started to show
the icon and is blinking with teststatusicon. More to come ...
2005-07-06 Tor Lillqvist <tml@novell.com>
* gdk/win32/gdkcursor-win32.c
(gdk_win32_icon_to_pixbuf_libgtk_only): Don't create the GdkPixbuf
until we know that we have something to put in it.
