Behdad Esfahbod [Wed, 1 Oct 2014 22:48:40 +0000 (18:48 -0400)]
[travis] Only run coveralls under gcc
Behdad Esfahbod [Wed, 1 Oct 2014 22:18:25 +0000 (18:18 -0400)]
Help clang with gcov
Behdad Esfahbod [Wed, 1 Oct 2014 22:07:27 +0000 (18:07 -0400)]
[travis] Try to make coverage work with clang
Behdad Esfahbod [Wed, 1 Oct 2014 21:57:43 +0000 (17:57 -0400)]
[travis] Minor
Behdad Esfahbod [Wed, 1 Oct 2014 21:45:17 +0000 (17:45 -0400)]
Add README.md symlink to make github happy
Behdad Esfahbod [Wed, 1 Oct 2014 21:44:30 +0000 (17:44 -0400)]
[travis] Re-enable clang
Behdad Esfahbod [Wed, 1 Oct 2014 21:41:41 +0000 (17:41 -0400)]
Add build and coverage status links
Behdad Esfahbod [Wed, 1 Oct 2014 21:35:03 +0000 (17:35 -0400)]
[travis] Report matrix settings to help debugging
Behdad Esfahbod [Wed, 1 Oct 2014 21:20:31 +0000 (17:20 -0400)]
[travis] More coveralls
Behdad Esfahbod [Wed, 1 Oct 2014 21:05:58 +0000 (17:05 -0400)]
[travis] Give coveralls a hand
Behdad Esfahbod [Wed, 1 Oct 2014 21:03:26 +0000 (17:03 -0400)]
[travis] Disable clang for now
Behdad Esfahbod [Wed, 1 Oct 2014 20:53:00 +0000 (16:53 -0400)]
[test] Fixup test
Behdad Esfahbod [Wed, 1 Oct 2014 20:49:22 +0000 (16:49 -0400)]
[travis] Another try at coveralls.io
Behdad Esfahbod [Wed, 1 Oct 2014 20:38:47 +0000 (16:38 -0400)]
[travis] First try to hookup coveralls.io code coverage tracking
Behdad Esfahbod [Wed, 1 Oct 2014 20:14:59 +0000 (16:14 -0400)]
[travis] Unbreak
Behdad Esfahbod [Wed, 1 Oct 2014 20:09:08 +0000 (16:09 -0400)]
[travis] Minor
Behdad Esfahbod [Wed, 1 Oct 2014 18:24:07 +0000 (14:24 -0400)]
Create ragel-generated files in srcdir
Behdad Esfahbod [Wed, 1 Oct 2014 17:33:21 +0000 (13:33 -0400)]
Clean ragel-generated headers in maintainercleanfiles
Behdad Esfahbod [Wed, 1 Oct 2014 16:41:03 +0000 (12:41 -0400)]
Don't use tmp files in Makefile
Useful for code-coverage generation of the ragel-generated files
as they will now contain the correct .hh filename in them.
Behdad Esfahbod [Wed, 1 Oct 2014 15:56:07 +0000 (11:56 -0400)]
check-static-inits: Filter out zero-length sections
When compiling with -O0, zero-sized constructors were showing up
and confusing the test.
Behdad Esfahbod [Wed, 1 Oct 2014 15:55:14 +0000 (11:55 -0400)]
check-static-inits: check for static finalizers
Behdad Esfahbod [Wed, 1 Oct 2014 15:28:01 +0000 (11:28 -0400)]
Hide other bubble-sort
Behdad Esfahbod [Wed, 1 Oct 2014 15:07:08 +0000 (11:07 -0400)]
Hide bubble-sort!
Behdad Esfahbod [Thu, 25 Sep 2014 14:59:46 +0000 (17:59 +0300)]
[util] Fix ansi output when surface is empty
Behdad Esfahbod [Thu, 25 Sep 2014 14:45:49 +0000 (17:45 +0300)]
[ot-font] Add hb_ot_face_cmap_accelerator_t
Behdad Esfahbod [Thu, 25 Sep 2014 14:26:56 +0000 (17:26 +0300)]
[ot-font] Minor
Behdad Esfahbod [Thu, 25 Sep 2014 14:15:35 +0000 (17:15 +0300)]
[ot-font] Handle missing vertical metrics tables
Behdad Esfahbod [Thu, 25 Sep 2014 13:53:24 +0000 (16:53 +0300)]
[ot-font] Add metrics_accel_t
Behdad Esfahbod [Thu, 25 Sep 2014 10:04:08 +0000 (13:04 +0300)]
[ot-font] Start adding vertical support
Behdad Esfahbod [Mon, 25 Aug 2014 16:54:57 +0000 (12:54 -0400)]
Fix build
Behdad Esfahbod [Mon, 25 Aug 2014 16:31:19 +0000 (12:31 -0400)]
[arabic] Don't #include __FILE__
Fine!
https://code.google.com/p/chromium/issues/detail?id=406957
Behdad Esfahbod [Mon, 25 Aug 2014 15:48:12 +0000 (11:48 -0400)]
[arabic] Allow disabling win1256 fallback code
By defining HB_NO_WIN1256.
Behdad Esfahbod [Fri, 22 Aug 2014 16:06:57 +0000 (12:06 -0400)]
[ot] Prefer Windows platform cmap tables / accept all Unicode encodingIDs
Some fonts on the Mac ship with (0,1).
Behdad Esfahbod [Fri, 15 Aug 2014 18:28:04 +0000 (14:28 -0400)]
[win32] Avoid preprocessor warnings re macros
Behdad Esfahbod [Thu, 14 Aug 2014 17:33:37 +0000 (13:33 -0400)]
Fix check-symbols on OS X
Behdad Esfahbod [Thu, 14 Aug 2014 17:29:30 +0000 (13:29 -0400)]
[coretext] Hide feature_mappings
Behdad Esfahbod [Thu, 14 Aug 2014 17:25:55 +0000 (13:25 -0400)]
Simplify hb-object
Behdad Esfahbod [Thu, 14 Aug 2014 17:15:21 +0000 (13:15 -0400)]
Really fix clang build this time
Behdad Esfahbod [Thu, 14 Aug 2014 17:11:33 +0000 (13:11 -0400)]
Revert "Fix build on clang after previous commit"
This reverts commit
9c80cbc87f23040a334e30c750f64523d9ef10c8.
Behdad Esfahbod [Thu, 14 Aug 2014 17:09:52 +0000 (13:09 -0400)]
Fix build on clang after previous commit
Behdad Esfahbod [Thu, 14 Aug 2014 17:05:36 +0000 (13:05 -0400)]
Rewrite this==NULL checks to avoid undefined behavior
Fixes https://code.google.com/p/chromium/issues/detail?id=403594
Behdad Esfahbod [Thu, 14 Aug 2014 16:59:16 +0000 (12:59 -0400)]
Remove this==NULL check from face->reference_table()
Not supposed to happen, and apparently this is undefined in C++.
https://code.google.com/p/chromium/issues/detail?id=403594
Behdad Esfahbod [Thu, 14 Aug 2014 16:57:02 +0000 (12:57 -0400)]
Make hb_object_t members private
In preparation for fixing:
https://code.google.com/p/chromium/issues/detail?id=403594
Behdad Esfahbod [Wed, 13 Aug 2014 23:42:01 +0000 (19:42 -0400)]
Minor
Behdad Esfahbod [Wed, 13 Aug 2014 16:39:34 +0000 (12:39 -0400)]
0.9.35
Behdad Esfahbod [Tue, 12 Aug 2014 23:26:35 +0000 (19:26 -0400)]
[coretext] Add version guards for kCTLanguageAttributeName
Behdad Esfahbod [Tue, 12 Aug 2014 23:17:19 +0000 (19:17 -0400)]
[coretext] Pass buffer language to CoreText
Behdad Esfahbod [Tue, 12 Aug 2014 23:10:33 +0000 (19:10 -0400)]
[coretext] Minor
Behdad Esfahbod [Tue, 12 Aug 2014 22:57:08 +0000 (18:57 -0400)]
[hb-coretext] Fix cluster order of notdef runs in RTL text
Behdad Esfahbod [Tue, 12 Aug 2014 21:15:09 +0000 (17:15 -0400)]
[shape-plan] Fix typo!
The only effect is, if shaper_list was not NULL and no shaper was found,
we now don't insert anything into cache and return earlier.
Behdad Esfahbod [Tue, 12 Aug 2014 21:14:36 +0000 (17:14 -0400)]
[shape-plan] Better debug messages
Behdad Esfahbod [Tue, 12 Aug 2014 21:03:27 +0000 (17:03 -0400)]
[shape-plan] Simplify macro
Behdad Esfahbod [Tue, 12 Aug 2014 21:02:59 +0000 (17:02 -0400)]
Fix gcc warning
Behdad Esfahbod [Tue, 12 Aug 2014 20:50:22 +0000 (16:50 -0400)]
[shape-plan] Fix shape-plan caching with more than one requested shaper
Wasn't breaking out of loop, ouch!
http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1246370.html
Behdad Esfahbod [Tue, 12 Aug 2014 20:49:18 +0000 (16:49 -0400)]
[shape-plan] Add debug tracing
Behdad Esfahbod [Tue, 12 Aug 2014 19:49:47 +0000 (15:49 -0400)]
[coretext] Fix buffer resizing
Was very broken. Now fixed and tested.
Behdad Esfahbod [Tue, 12 Aug 2014 18:25:11 +0000 (14:25 -0400)]
[coretext] Rewind scratch-allocated arrays when not needed anymore
Behdad Esfahbod [Tue, 12 Aug 2014 17:12:31 +0000 (13:12 -0400)]
[wince] Two more Windows CE fixes
Report has it that it builds (and works) now:
https://codereview.qt-project.org/#/c/92087/
Behdad Esfahbod [Tue, 12 Aug 2014 14:32:41 +0000 (10:32 -0400)]
[coretext] Compare CGFont and PS name, if CTFont didn't match
See comments.
Fixes vertical text. CoreText backend is in very good shape now!
Also see:
5a0eed3b50629be4826e4e9428f2c3255195395d
25f4fb9b56bb3f8bec821571c78f8829e40daa54
Fixes http://github.com/behdad/harfbuzz/pull/36
Behdad Esfahbod [Tue, 12 Aug 2014 03:47:16 +0000 (23:47 -0400)]
[coretext] Implement vertical shaping
Currently doesn't work though, we detect font fallback. Apparently
matching on ct_font is not safe for this. Looks like commit
25f4fb9b56bb3f8bec821571c78f8829e40daa54 wasn't enough after all.
Behdad Esfahbod [Tue, 12 Aug 2014 00:45:12 +0000 (20:45 -0400)]
[coretext] Remove unnecessary alt_size
Wasn't needed after
a6b8dc87421de33746b0b14d86d2d1532aec02af.
Behdad Esfahbod [Tue, 12 Aug 2014 00:02:45 +0000 (20:02 -0400)]
[coretext] Use CFRunStatus
Assert that all runs had expected direction, and take hint for
non-monotone clusters.
Behdad Esfahbod [Tue, 12 Aug 2014 00:01:37 +0000 (20:01 -0400)]
[coretext] Minor
Behdad Esfahbod [Mon, 11 Aug 2014 23:16:26 +0000 (19:16 -0400)]
[coretext] Remove debug printf!
Behdad Esfahbod [Mon, 11 Aug 2014 22:40:01 +0000 (18:40 -0400)]
API: Do not clear buffer-flags in hb_buffer_clear_contents()
After
763e5466c0a03a7c27020e1e2598e488612529a7, one doesn't
need to set flags for different pieces of text. The flags now
are something the client sets up once, depending on how it
actually uses the buffer. As such, don't clear it in
clear_contents().
Tests updated.
Behdad Esfahbod [Mon, 11 Aug 2014 22:23:43 +0000 (18:23 -0400)]
Minor
Behdad Esfahbod [Mon, 11 Aug 2014 21:46:50 +0000 (17:46 -0400)]
[coretext] Pass buffer direction to CoreText
Have to use a CTTypesetter for this.
Behdad Esfahbod [Mon, 11 Aug 2014 21:46:12 +0000 (17:46 -0400)]
[coretext] Minor
It's hard to handle all possible NULL returns from CoreText. Add one
more...
Behdad Esfahbod [Mon, 11 Aug 2014 19:29:18 +0000 (15:29 -0400)]
[coretext] Attach marks to base clusters
Fixes https://githu.com/behdad/harfbuzz/issues/49
to the extent that it can be fixed.
Behdad Esfahbod [Mon, 11 Aug 2014 19:11:59 +0000 (15:11 -0400)]
[coretext] Use input clusters
Before, this shaper was returning UTF-16 cluster indices instead of
returning whatever cluster values the user had had set up in the buffer.
Ouch!
Behdad Esfahbod [Mon, 11 Aug 2014 19:08:19 +0000 (15:08 -0400)]
[coretext] Fix buffer resize handling
We can't really resize buffer and continue in this shaper as we are
using the scratch buffer for string_ref and log_cluster. Restructure
shaper to retry from (almost) scratch.
Behdad Esfahbod [Mon, 11 Aug 2014 17:25:43 +0000 (13:25 -0400)]
[coretext] Always compute log_clusters
To be used soon.
Behdad Esfahbod [Mon, 11 Aug 2014 17:42:42 +0000 (13:42 -0400)]
Minor
Behdad Esfahbod [Mon, 11 Aug 2014 06:04:38 +0000 (02:04 -0400)]
[coretext] Simplify cluster mapping
Behdad Esfahbod [Sun, 10 Aug 2014 23:19:38 +0000 (19:19 -0400)]
[coretext] Minor
Behdad Esfahbod [Sun, 10 Aug 2014 23:05:25 +0000 (19:05 -0400)]
[coretext] Fix fallback detection
Fixes http://github.com/behdad/harfbuzz/pull/36
Behdad Esfahbod [Sun, 10 Aug 2014 22:59:47 +0000 (18:59 -0400)]
[coretext] Fix last range
Test with:
hb-view /Library/Fonts/Zapfino.ttf ZapfinoZapfino --features=-dlig[7:] --shaper=coretext
Behdad Esfahbod [Sun, 10 Aug 2014 22:52:07 +0000 (18:52 -0400)]
Fix hb_in_range() for types smaller than int
As exercised by hb-coretext .notdef code.
Behdad Esfahbod [Sun, 10 Aug 2014 22:04:50 +0000 (18:04 -0400)]
[wince] Try to fix some stuff on Windows CE
Based on errors seen here:
http://testresults.qt-project.org/ci/QtBase_dev_Integration/build_03837/wince70embedded-armv4i-msvc2008_Windows_7/log.txt.gz
Fully untested.
Behdad Esfahbod [Sun, 10 Aug 2014 21:42:19 +0000 (17:42 -0400)]
Minor
Behdad Esfahbod [Sun, 10 Aug 2014 21:40:24 +0000 (17:40 -0400)]
[coretext] Fix assert on Yosemite
Apparently those functions documented as sometimes returning NULL
actually exercise that right in OS X 10.10 Yosemite. The scratch
was too small for that. I *think* I fixed it, but haven't tested
as I don't have Yosemite.
Behdad Esfahbod [Fri, 18 Apr 2014 18:21:08 +0000 (11:21 -0700)]
[coretext] Use CGFont as face_data
Behdad Esfahbod [Thu, 7 Aug 2014 19:38:52 +0000 (15:38 -0400)]
[arabic] Bug 82306 - Mandaic had errors in its Unicode Joining_Type
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=82306
Behdad Esfahbod [Wed, 6 Aug 2014 17:34:49 +0000 (13:34 -0400)]
Use atexit() only if it's safe to call from shared library
Apparently they are not (advertised as?) safe on BSD systems.
We ignore the case of static libraries.
Whitelisted on glibc, Android, and MSVC / mingw.
https://bugs.freedesktop.org/show_bug.cgi?id=82246
Behdad Esfahbod [Tue, 5 Aug 2014 18:19:01 +0000 (14:19 -0400)]
[arabic] Fix fallback shaping regression
Was broken in
615d00ea252739da57edbd980ff27e573.
Fixes https://github.com/behdad/harfbuzz/pull/48
Micro-test added.
Behdad Esfahbod [Sat, 2 Aug 2014 23:24:55 +0000 (19:24 -0400)]
0.9.34
Behdad Esfahbod [Sat, 2 Aug 2014 21:46:38 +0000 (17:46 -0400)]
[uniscribe] Minor
Behdad Esfahbod [Sat, 2 Aug 2014 21:18:46 +0000 (17:18 -0400)]
Make sure gsubgpos buffer vars are available during fallback_position
Add buffer var allocation asserts to a few key places.
Behdad Esfahbod [Sat, 2 Aug 2014 20:31:16 +0000 (16:31 -0400)]
Disable 'liga' for vertical text
The reason we turned it on is because Kazuraki uses it. But that's
not reason enough. Until the OpenType spec gets its act together re
adding design-direction to lookups, this is better user experience.
Behdad Esfahbod [Sat, 2 Aug 2014 20:17:44 +0000 (16:17 -0400)]
Make it easier to use HB_BUFFER_FLAG_BOT/EOT
Previously, we expected users to provide BOT/EOT flags when the
text *segment* was at paragraph boundaries. This meant that for
clients that provide full paragraph to HarfBuzz (eg. Pango), they
had code like this:
hb_buffer_set_flags (hb_buffer,
(item_offset == 0 ? HB_BUFFER_FLAG_BOT : 0) |
(item_offset + item_length == paragraph_length ?
HB_BUFFER_FLAG_EOT : 0));
hb_buffer_add_utf8 (hb_buffer,
paragraph_text, paragraph_length,
item_offset, item_length);
After this change such clients can simply say:
hb_buffer_set_flags (hb_buffer,
HB_BUFFER_FLAG_BOT | HB_BUFFER_FLAG_EOT);
hb_buffer_add_utf8 (hb_buffer,
paragraph_text, paragraph_length,
item_offset, item_length);
Ie, HarfBuzz itself checks whether the segment is at the beginning/end
of the paragraph. Clients that only pass item-at-a-time to HarfBuzz
continue not setting any flags whatsoever.
Another way to put it is: if there's pre-context text in the buffer,
HarfBuzz ignores the BOT flag. If there's post-context, it ignores
EOT flag.
Behdad Esfahbod [Sat, 2 Aug 2014 19:00:34 +0000 (15:00 -0400)]
Merge branch 'win1256'
Behdad Esfahbod [Sat, 2 Aug 2014 18:58:38 +0000 (14:58 -0400)]
[arabic/win1256] Remove unused MultipleSubst macros
Behdad Esfahbod [Fri, 1 Aug 2014 23:05:33 +0000 (19:05 -0400)]
[arabic/win1256] Really fix lam-alef this time
https://bugzilla.mozilla.org/show_bug.cgi?id=1045139#c26
https://github.com/behdad/harfbuzz/commit/
b276e897d17519a2c28f79d024904ac2bdd86bcf#commitcomment-7243499
Behdad Esfahbod [Fri, 1 Aug 2014 20:14:33 +0000 (16:14 -0400)]
[arabic/win1256] Hook up lamMedi lookup
Restructure lookup array to accommodate.
https://bugzilla.mozilla.org/show_bug.cgi?id=1045139#c18
Behdad Esfahbod [Fri, 1 Aug 2014 20:09:31 +0000 (16:09 -0400)]
[arabic/win1256] Fix TEH MARBUTA final form
https://bugzilla.mozilla.org/show_bug.cgi?id=1045139#c18
Behdad Esfahbod [Thu, 31 Jul 2014 22:21:05 +0000 (18:21 -0400)]
[arabic/win1256] Fix shaping of JEEM, HAH, and KHAH
https://bugzilla.mozilla.org/show_bug.cgi?id=1045139#c16
Behdad Esfahbod [Thu, 31 Jul 2014 15:04:00 +0000 (11:04 -0400)]
[arabic/win1256] Remove GCC'ism
The table can now compile independently too. If we cannot make it work
on MSVC, we can always generate the data and distribute it.
The code now compiles cleanly with:
gcc -c -xc -std=c99 -Werror -pedantic hb-ot-shape-complex-arabic-win1256.hh
g++ -c -xc -std=c++1x -Werror -pedantic hb-ot-shape-complex-arabic-win1256.hh
See:
https://github.com/behdad/harfbuzz/commit/
a97f537cec209649302899975d76ca2b2661da7a#commitcomment-7218736
Behdad Esfahbod [Wed, 30 Jul 2014 06:15:44 +0000 (02:15 -0400)]
[arabic] Implement Windows-1256 private shaping
Bug 1045139 - The Arabic text with "MS Sans Serif" font is rendered bad
https://bugzilla.mozilla.org/show_bug.cgi?id=1045139
This is only enabled on Windows platforms, and requires support from
Uniscribe to work. But for clients that do hook up to Uniscribe, this
fixes shaping of Windows-1256-encoded bitmap fonts like "MS Sans Serif".
The code and table together have just less than a 1kb footprint when
enabled.
UNTESTED. I might even have broken regular Arabic fallback shaping.
Behdad Esfahbod [Thu, 31 Jul 2014 22:51:37 +0000 (18:51 -0400)]
[hangul] Don't apply 'calt'
See comments.
Micro-test added.
Behdad Esfahbod [Thu, 31 Jul 2014 22:33:11 +0000 (18:33 -0400)]
[test] Fix record-test.sh