Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
gtk2's XIM does NOT support OverTheSpot, but gtk2's GTK immodule SUPPORTS OverTheSpot. It is why #290864 was blocked by #290469, because gtk2 XIM has no OverTheSpot, even in the latest versions. Major inputmethod frameworks all support GTK immodule and OverTheSpot, and they work great in all gtk2-based applications, besides mozilla family. Upstream's #271815 was fixed for GTK immodule with OverTheSpot. It still does not support XIM with OverTheSpot, because gtk2 builds depend on gtk2's inputmethod framework, and gtk2 XIM has no OverTheSpot. #379935 (my bug report) is NOT blocked by #290469. It is NOT related with gtk2's XIM. #290864 was blocked by #290469, obviously. My report should not be mixed with it. Not all patchs will accept by maintainers, I konw. But my report should not be blocked by a funny reason. It is mozilla's official patch, not mine. I think it is worth to be merged with Debian's packages. Mike Hommey wrote: #290864 points to the upstream bug you pointed to. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
On Fri, Jul 28, 2006 at 08:26:21PM +0800, caleb [EMAIL PROTECTED] wrote: gtk2's XIM does NOT support OverTheSpot, but gtk2's GTK immodule SUPPORTS OverTheSpot. It is why #290864 was blocked by #290469, because gtk2 XIM has no OverTheSpot, even in the latest versions. Major inputmethod frameworks all support GTK immodule and OverTheSpot, and they work great in all gtk2-based applications, besides mozilla family. Upstream's #271815 was fixed for GTK immodule with OverTheSpot. It still does not support XIM with OverTheSpot, because gtk2 builds depend on gtk2's inputmethod framework, and gtk2 XIM has no OverTheSpot. #379935 (my bug report) is NOT blocked by #290469. It is NOT related with gtk2's XIM. #290864 was blocked by #290469, obviously. My report should not be mixed with it. Not all patchs will accept by maintainers, I konw. But my report should not be blocked by a funny reason. It is mozilla's official patch, not mine. I think it is worth to be merged with Debian's packages. That the previous bug is blocked for a bad reason is one thing. But the fact remains: the upstream bug pointed by #290864 is the same as yours. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
Debian's bug is not always equal to upstream's. The problems coverage is: Debian's #290864 Mozilla's #271815 Debian's #379935 (my report). #290864 is gnome's #158678 plus part of mozilla's #271815. Mozilla is impossible to fix XIM. That is gtk2's fault, not mozilla's. Gtk1 supports XIM OverTheSpot, and mozilla's gtk1 builds works correctly. #379935 is part of mozilla's #271815. It talks only about GTK immodule, not XIM. Mozilla has its own internal structure for inputmethods, that is why it has the strange bug. No other gtk2 applications has this odd behavior. Mike Hommey wrote: That the previous bug is blocked for a bad reason is one thing. But the fact remains: the upstream bug pointed by #290864 is the same as yours. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
No, this patch needs NOT a patch to gtk, and it works great with Debian official gtk packages. This bug #379935 is different from #290864. #379935 is about GTK immodule with OverTheSpot, while #290864 is about XIM mode with OverTheSpot. #379935 needs NOT a patch to gtk, while #290864 needs gtk to fix XIM behavior. (gtk 2.6.x and lator ones had fixed XIM problem partially, but it is NOT related with my bug report #379935.) #379935's patch have been merged into mozilla cvs trunk, while #290864's one was never adopted by upstream. #379935's patch solved the OverTheSpot issue cleanly, while #290864 was a dirty hack, and #290864's patche slowdown firefox a bit. #379935's patch is for firefox 1.5 and later ones, while #290864's patch only works for firefox 1.0 series. Mike Hommey wrote: *sigh* why don't people look at already opened bugs before filing one ? This has already been reported and sounds like it needs a patch to gtk to work. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
On Thu, Jul 27, 2006 at 07:32:43PM +0800, caleb [EMAIL PROTECTED] wrote: No, this patch needs NOT a patch to gtk, and it works great with Debian official gtk packages. This bug #379935 is different from #290864. #379935 is about GTK immodule with OverTheSpot, while #290864 is about XIM mode with OverTheSpot. #379935 needs NOT a patch to gtk, while #290864 needs gtk to fix XIM behavior. (gtk 2.6.x and lator ones had fixed XIM problem partially, but it is NOT related with my bug report #379935.) #379935's patch have been merged into mozilla cvs trunk, while #290864's one was never adopted by upstream. #379935's patch solved the OverTheSpot issue cleanly, while #290864 was a dirty hack, and #290864's patche slowdown firefox a bit. #379935's patch is for firefox 1.5 and later ones, while #290864's patch only works for firefox 1.0 series. #290864 points to the upstream bug you pointed to. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
Package: firefox
Version: 1.5.dfsg+1.5.0.4-3
Severity: important
Tags: patch l10n
It is bug #271815 of mozilla.org, and is fixed in cvs trunk.
( https://bugzilla.mozilla.org/show_bug.cgi?id=271815 )
The file I attached is mozilla.org's official patch, which works great for
firefox 1.5 and later ones. The patch is clean, and does not slowdown firefox.
The bug #271815 is: Text cursor always points to a wrong place under
OverTheSpot mode. It is very annoying when typing CJKV characters.
(OnTheSpot mode has not this bug.)
Debian's official mozilla suite have applied an old version patch, which only
worked for firefox 1.0 and older. Firefox 1.5 and newer ones need this new
OverTheSpot patch.
(Mozilla.org's nightly builds, aka firefox 3.0a1, has applied this patch, and
OverTheSpot works great.)
-- System Information:
Debian Release: testing/unstable
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: i386 (i686)
Shell: /bin/sh linked to /bin/bash
Kernel: Linux 2.6.17-1-k7
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968)
Versions of packages firefox depends on:
ii debianutils 2.17 Miscellaneous utilities specific t
ii fontconfig 2.3.2-7 generic font configuration library
ii libatk1.0-0 1.12.1-1 The ATK accessibility toolkit
ii libc6 2.3.6-15 GNU C Library: Shared libraries
ii libcairo2 1.2.0-3 The Cairo 2D vector graphics libra
ii libfontconfig1 2.3.2-7 generic font configuration library
ii libfreetype62.2.1-2 FreeType 2 font engine, shared lib
ii libgcc1 1:4.1.1-9GCC support library
ii libglib2.0-02.10.3-3 The GLib library of C routines
ii libgtk2.0-0 2.8.18-1 The GTK+ graphical user interface
ii libidl0 0.8.6-1 library for parsing CORBA IDL file
ii libjpeg62 6b-13The Independent JPEG Group's JPEG
ii libpango1.0-0 1.12.3-1+b1 Layout and rendering of internatio
ii libpng12-0 1.2.8rel-5.2 PNG library - runtime
ii libstdc++6 4.1.1-9 The GNU Standard C++ Library v3
ii libx11-62:1.0.0-7X11 client-side library
ii libxft2 2.1.8.2-8FreeType-based font drawing librar
ii libxinerama11:1.0.1-4X11 Xinerama extension library
ii libxp6 1:1.0.0-1X Printing Extension (Xprint) clie
ii libxt6 1:1.0.0-5X11 toolkit intrinsics library
ii psmisc 22.2-1 Utilities that use the proc filesy
ii zlib1g 1:1.2.3-13 compression library - runtime
firefox recommends no packages.
-- no debconf information
Index: editor/libeditor/text/nsPlaintextEditor.cpp
===
RCS file: /cvsroot/mozilla/editor/libeditor/text/nsPlaintextEditor.cpp,v
retrieving revision 1.95
diff -u -8 -p -w -r1.95 nsPlaintextEditor.cpp
--- editor/libeditor/text/nsPlaintextEditor.cpp 1 May 2006 05:25:49 - 1.95
+++ editor/libeditor/text/nsPlaintextEditor.cpp 19 Jun 2006 08:32:58 -
@@ -1538,35 +1538,52 @@ nsPlaintextEditor::GetEmbeddedObjects(ns
#pragma mark -
#pragma mark nsIEditorIMESupport overrides
#pragma mark -
#endif
NS_IMETHODIMP
nsPlaintextEditor::SetCompositionString(const nsAString aCompositionString, nsIPrivateTextRangeList* aTextRangeList,nsTextEventReply* aReply)
{
- NS_ASSERTION(aTextRangeList, null ptr);
- if (!aTextRangeList)
-return NS_ERROR_NULL_POINTER;
-
- // workaround for windows ime bug 23558: we get every ime event twice.
- // for escape keypress, this causes an empty string to be passed
- // twice, which freaks out the editor. This is to detect and avoid that
- // situation:
- if (aCompositionString.IsEmpty() !mIMETextNode)
+ if (!aTextRangeList !aCompositionString.IsEmpty())
{
-return NS_OK;
+NS_ERROR(aTextRangeList is null but the composition string is not null);
+return NS_ERROR_NULL_POINTER;
}
- mIMETextRangeList = aTextRangeList;
-
nsCOMPtrnsIPresShell ps = do_QueryReferent(mPresShellWeak);
if (!ps)
return NS_ERROR_NOT_INITIALIZED;
+ nsCOMPtrnsISelection selection;
+ nsresult result = GetSelection(getter_AddRefs(selection));
+ if (NS_FAILED(result)) return result;
+
+ nsCOMPtrnsICaret caretP;
+ ps-GetCaret(getter_AddRefs(caretP));
+
+ // We should return caret position if it is possible. Because this event
+ // dispatcher always expects to be returned the correct caret position.
+ // But in following cases, we don't need to process the composition string,
+ // so, we only need to return the caret position.
+
+ // aCompositionString.IsEmpty() !mIMETextNode:
+ // Workaround for Windows IME bug 23558: We get every IME event twice.
+ // For escape keypress, this causes
Bug#379935: Wrong cursor place in OverTheSpot mode of inputmethods
block 379935 by 290469 severity 379935 wishlist tag 379935 + l10n merge 290864 379935 forwarded 379935 https://bugzilla.mozilla.org/show_bug.cgi?id=271815 thanks *sigh* why don't people look at already opened bugs before filing one ? On Wed, Jul 26, 2006 at 09:11:04PM +0800, Wen-Yen Chuang [EMAIL PROTECTED] wrote: Package: firefox Version: 1.5.dfsg+1.5.0.4-3 Severity: important Tags: patch l10n It is bug #271815 of mozilla.org, and is fixed in cvs trunk. ( https://bugzilla.mozilla.org/show_bug.cgi?id=271815 ) The file I attached is mozilla.org's official patch, which works great for firefox 1.5 and later ones. The patch is clean, and does not slowdown firefox. The bug #271815 is: Text cursor always points to a wrong place under OverTheSpot mode. It is very annoying when typing CJKV characters. (OnTheSpot mode has not this bug.) Debian's official mozilla suite have applied an old version patch, which only worked for firefox 1.0 and older. Firefox 1.5 and newer ones need this new OverTheSpot patch. (Mozilla.org's nightly builds, aka firefox 3.0a1, has applied this patch, and OverTheSpot works great.) This has already been reported and sounds like it needs a patch to gtk to work. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
