summaryrefslogtreecommitdiff
path: root/pango
Commit message (Collapse)AuthorAgeFilesLines
* Revert "Remove the .vcproj files before rewriting them"Behdad Esfahbod2010-03-308-205/+54
| | | | | Revert .vcproj files and accompanying commits as the build stuff needs more work.
* Remove the .vcproj files before rewriting themTor Lillqvist2010-03-301-0/+3
| | | | | In case they are write protected for some reason. Might help Behdad's problem with make dist.
* Disable use of HB Device tablesBehdad Esfahbod2010-03-291-0/+2
| | | | | There's cleanup needed in the Device code. The scale stuff is pretty bogus right now.
* [HB] Fix leak when duplicating blobBehdad Esfahbod2010-03-271-2/+4
|
* Bug 611033 - Layout using ATSUI fontmap treats absolute font sizes a...Kristian Rietveld2010-03-212-25/+30
| | | | | | | | | | pango_atsui_font_map_load_font() has been corrected to take into account whether the size retrieved from the given font description is absolute when setting this size again on best_description. _pango_cairo_atsui_font_new() has been cleaned up with regard to size conversion and usage. The code now properly takes the absolute property into account and is more clear.
* [HB] Fix the mystery bug!Behdad Esfahbod2010-03-165-31/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | A couple bugs joined forces to exhibit the mystery behavior of crashes / infinite loops on OS X / wrong kerning / invalid memory access. Pooh! The bugs were involved: - Wrong pointer math with ValueRecord in PairPosFormat1 - Fallout from avoiding flex arrays, code not correctly updated to remove sizeof() usage. We strictly never use sizeof() directly now. And the PairPos code is cleaned up. Should fix them all. Bugs are: Bug 605655 - Pango 1.26.2 introduces kerning bug Bug 611229 - Pango reads from uninitialized memory Bug 593240 - (pangoosx) Crash / infinite loop with Mac OS X We were also doing wrong math converting Device adjustments to hb_position_t. Fallout from FreeType days. Should shift 16, not 6. Fixed that too. There's still another bug: we don't sanitize Device records referenced from value records. Fixing that also.
* [HB] CosmeticBehdad Esfahbod2010-03-161-3/+3
|
* Add a couple of files accidentally left out in previous commitTor Lillqvist2010-03-151-0/+30
|
* Add solution and project files for Visual Studio buildsTor Lillqvist2010-03-157-54/+172
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | The VS files are kept in build/win32/vs9, the same relative location as in GLib, mostly for historical reasons. Update README.win32 to reflect the VS possibility. Include the VS solution and project files in the tarball when doing a release. To avoid having to list source files in several locations, generate files listing source files at "make dist" time in the corresponding source directories. Use the C preprocessor to preprocess .vcprojin files that include said list files into the actual .vcproj project files. Provide a rc file for the pangocairo DLL, too. Construct and provide in the tarball a pregenerated config.h.win32 which can be used as config.h when building with Visual C without running any configure script. Provide pregenerarated module-defs.h.win32, module-defs-lang.c.win32 files too. Define PANGO_MODULE_PREFIX appropriately in the relevant module source files if not available from the compilation command line. Update module-defs-fc.c.win32 and module-defs-win32.c.win32 to match what the configure script provides.
* Bug 609326 - Complex script shaping failed in the FT2 backend on WindowsBehdad Esfahbod2010-03-011-7/+1
| | | | Check for face->stream->read == NULL instead of face->stream->base != NULL.
* Make blob unlocking 64bit-safeBehdad Esfahbod2010-03-011-7/+7
| | | | Bug 604128 - Applications crash when displaying Hebrew characters
* Bug 609326 - Complex script shaping failed in the FT2 backend on WindowsTor Lillqvist2010-03-011-1/+7
| | | | | | I am not sure whether the root cause is a bug in FreeType, or the way Pango uses FreeType, or some more exotic mechanism. Anyway, add a simple workaround.
* Ensure proper alignment of NullPoolBehdad Esfahbod2010-02-221-1/+1
| | | | Bug 593240 - (pangoosx) Crash / infinite loop with Mac OS X
* Bug 610183 - Pango doesn't compile with gcc 3.3.3Behdad Esfahbod2010-02-171-1/+1
| | | | Disable some __attribute__s with gcc 3.
* [HB] Remove unused argumentBehdad Esfahbod2010-01-263-4/+1
|
* [HB/GDEF] Fix bug in building synthetic GDEF tableBehdad Esfahbod2010-01-261-1/+3
|
* Do not get the gir and typelib dir for g-iJohan Dahlin2009-12-311-2/+2
| | | | Since it breaks when using a system install of gobject-introspection
* Simplify cluster analysisBehdad Esfahbod2009-12-251-3/+2
|
* Always recalculate run width after breaking the runBehdad Esfahbod2009-12-251-8/+2
| | | | | Before we only did this if justifying. However, computing the width is essentially free these days, so do it always.
* Zero the space at the end of line when breaking linesBehdad Esfahbod2009-12-251-0/+10
| | | | | | | | | Previously we were not zeroing the final space in the line when breaking lines and were doing that only after lines were broken. This was broken since setting layout width to its own logical width (under width=-1) could result in differently broken lines. That's fixed now. Problem originally reported on gtk-list on 2009-12-22 by Ben Pfaff.
* CosmeticBehdad Esfahbod2009-12-251-2/+1
|
* [modules] Don't warn multiple timesBehdad Esfahbod2009-12-251-12/+32
|
* Don't use devel version numbers in Since tagsKristian Rietveld2009-12-241-1/+1
|
* Revert usage of CoreText in ATSUI backendKristian Rietveld2009-12-224-14/+44
| | | | | | | | | | | | | | | | | | | | | | Next to CGFontRef we also store an ATSFontRef now. This required the addition of the public for backends function pango_atsui_font_get_atsfont(). I have marked this call to be available starting version 1.27, as Pango is now in the midst of the 1.26 series. Since we now store the ATSFontRef next to CGFontRef, we can continue to make use of the ATS API in pango_cairo_atsui_font_create_metrics_for_context(). The CoreText calls have been removed. Both Tiger and earlier, and Leopard and later machines will now use the ATS API. It is well known that the ATS API has been deprecated in Leopard. In the near future a brand-new CoreText backend has to be written (from scratch) that should be installed and used on Macs running Leopard and later. This patch has been tested on 32-bit Tiger and 64-bit Snow Leopard and found to be functioning correctly. The build should now work again fine on Tiger machines, as it has been broken since the introduction of the CoreText calls.
* Fix FT_Face finalizer callBehdad Esfahbod2009-12-201-1/+1
|
* Use Makefile.introspectionJohan Dahlin2009-12-181-86/+59
| | | | | | | | To simplify the introspection scanner/compiler build rules. This also changes to the girs are included in the dist. https://bugzilla.gnome.org/show_bug.cgi?id=604770
* Simplify make rulesBehdad Esfahbod2009-12-141-4/+2
|
* Do not multiply return value of CTFontGetDescent with -1Kristian Rietveld2009-12-061-1/+1
| | | | | | | The return value of CTFontGetDescent is already positive, as opposed to the ats_metrics.descent value used before. This is a regression introduced by the previous patch "porting" the backend to CoreText, which also broke support for Tiger and earlier -- we will fix that soonish.
* Bug 602823 - void function cannot return valueBehdad Esfahbod2009-11-251-1/+1
|
* [GSUB] More ligature/component fixingBehdad Esfahbod2009-11-221-2/+2
| | | | | | | We can only reuse the ligid if it belongs to a previous ligature, not a component! Cherry-picked from harfbuzz-ng repo.
* Fix MarkMark issue with ligid and componentsBehdad Esfahbod2009-11-222-10/+7
| | | | Cherry-picked from harfbuzz-ng repo.
* Bug 602660 - [HB] don't use flexible arraysBehdad Esfahbod2009-11-228-137/+122
| | | | Cherry-picked from harfbuzz-ng repo.
* Another C++ strictness fixBehdad Esfahbod2009-11-192-5/+5
| | | | | | | | | | | | | | | | Pango Bug 602408 - Invalid C++ code breaks compile with Sun C++ Compiler (Error: A union member cannot have a user-defined assignment operator) According to the bug: C++ Programming Language by Bjarne Stroustrup: Chapter 10.4.12 forbids explicitly using of union members with constructors, destructors or assignment operations. So we use a set() method instead of the assignment operator. Ugly, but hey, that's life. Cherry-pick from harfbuzz-ng repo.
* Fix pangocairo-fc fontsize weirdnessBehdad Esfahbod2009-11-171-3/+2
|
* Bug 341481 - pangocairo kerning problem with nonidentity scalingBehdad Esfahbod2009-11-173-6/+113
| | | | | Finally fix this embarrassing bug. The fix is a kludge, but it will be redone for 1.28 (harfbuzz-ng) anyway.
* Bug 596614 - Pango application receives SIGSEGV on pasting text with U+202EBehdad Esfahbod2009-11-175-27/+61
| | | | | Fixed all modules to reverse glyphs if run is rtl. Except for Hangul module. Fixed pango_shape() to detect that, warn, and reverse.
* Remove REBUILD black magicBehdad Esfahbod2009-11-101-2/+2
| | | | Shouldn't be needed.
* Fix leak with non-mmapped fontsBehdad Esfahbod2009-11-031-1/+1
|
* [HB] Add glyph_metrics typedefBehdad Esfahbod2009-10-291-0/+9
|
* [HB] Add point_index to get_contour_point()Behdad Esfahbod2009-10-291-0/+1
| | | | Pointed out by Keith Stribley on HarfBuzz list.
* [HB] Fix lookup_flag ignoringBehdad Esfahbod2009-10-292-1/+2
| | | | Reported by Keith Stribley on HarfBuzz list.
* Allow using uninstalled glibBehdad Esfahbod2009-10-221-2/+2
| | | | Reported by mpsuzuki on gtk-i18n-list on 2009-10-09.
* TypoBehdad Esfahbod2009-10-051-1/+1
|
* Bug 596837 - Typo in pango_cairo_show_layout documentationBehdad Esfahbod2009-09-301-1/+1
|
* [HB] Add commentBehdad Esfahbod2009-09-211-0/+5
|
* [HB] Fix apply_lookup() loopBehdad Esfahbod2009-09-211-1/+2
| | | | | Part of Bug 595539 - Regressions in rendering certain Thai sequences with OpenType font
* [HB] Fix bug in chain_context_lookup() invocationsBehdad Esfahbod2009-09-211-2/+2
| | | | | Part of Bug 595539 - Regressions in rendering certain Thai sequences with OpenType font
* Bug 595350 - set LC_ALL=C, not LANG=CBehdad Esfahbod2009-09-161-1/+2
|
* Bug 594936 - pango_layout_set_wrapBehdad Esfahbod2009-09-111-1/+1
| | | | | Remove wrong optimization in pango_layout_set_wrap(). The assumption that if the layout was not wrapped before, it will not wrap now is simply wrong.
* Bug 594934 - pango_layout_copyBehdad Esfahbod2009-09-112-32/+22
| | | | Use memcpy() to copy value fields so we don't miss any.
View Lorry Controller Status