Behdad Esfahbod [Mon, 16 May 2011 19:08:31 +0000 (15:08 -0400)]
[hb-view] No need to allocate an extra glyph item at the end
Behdad Esfahbod [Mon, 16 May 2011 19:07:48 +0000 (15:07 -0400)]
Minor
Behdad Esfahbod [Mon, 16 May 2011 16:24:56 +0000 (12:24 -0400)]
[test/buffer] Add more tests for nil buffer
Behdad Esfahbod [Sat, 14 May 2011 03:04:46 +0000 (23:04 -0400)]
[test/ot-tag] More tests
Behdad Esfahbod [Sat, 14 May 2011 02:55:32 +0000 (22:55 -0400)]
[API] Add HB_LANGUAGE_INVALID
Behdad Esfahbod [Sat, 14 May 2011 02:46:36 +0000 (22:46 -0400)]
[test/test-ot-tag] Test hb-ot-tag.h, fix many bugs
I'm in awe with how many bugs this test revealed. All fixed.
Behdad Esfahbod [Sat, 14 May 2011 00:25:38 +0000 (20:25 -0400)]
[TODO] Add items
Behdad Esfahbod [Thu, 12 May 2011 21:48:20 +0000 (17:48 -0400)]
[API] Add hb_face_make/is_immutable()
Behdad Esfahbod [Thu, 12 May 2011 19:19:33 +0000 (15:19 -0400)]
Finish off previous change
Behdad Esfahbod [Thu, 12 May 2011 19:14:13 +0000 (15:14 -0400)]
[API] Make get_glyph() callback return a boolean
We need to know whether the glyph exists, so we can fallback to
composing / decomposing. Assuming that glyph==0 means "doesn't exist"
wouldn't work for applications like Pango that want to use different
"doesn't exist" glyph codes for different characters. An explicit
return value fixes that.
Behdad Esfahbod [Thu, 12 May 2011 18:27:44 +0000 (14:27 -0400)]
[test/shape] Check shape output
Behdad Esfahbod [Thu, 12 May 2011 16:39:40 +0000 (12:39 -0400)]
Rename get_kernings() arguments from first/second_glyph to left/right_glyph
Makes it clear that kerning is in visual order.
Behdad Esfahbod [Thu, 12 May 2011 14:53:57 +0000 (10:53 -0400)]
[test] Add test-shape.c. Oops
Behdad Esfahbod [Thu, 12 May 2011 14:49:30 +0000 (10:49 -0400)]
[test/blob] Use MAP_ANON instead of MAP_ANONYMOUS
More portable.
Behdad Esfahbod [Thu, 12 May 2011 05:39:17 +0000 (01:39 -0400)]
[test/shape] Add simplest test for hb_shape()
Behdad Esfahbod [Thu, 12 May 2011 05:19:39 +0000 (01:19 -0400)]
[object] Make object inert during destruction
Such that user_data and other finalizers cannot resurrect object
Behdad Esfahbod [Thu, 12 May 2011 05:02:03 +0000 (01:02 -0400)]
[TODO] Update
Behdad Esfahbod [Thu, 12 May 2011 04:46:57 +0000 (00:46 -0400)]
Fix font subclass chainup
Test passing now.
Behdad Esfahbod [Thu, 12 May 2011 04:35:12 +0000 (00:35 -0400)]
[test/font] Test font_funcs subclassing
Behdad Esfahbod [Thu, 12 May 2011 04:18:28 +0000 (00:18 -0400)]
[test/font] Test empty funcs
Behdad Esfahbod [Thu, 12 May 2011 03:57:36 +0000 (23:57 -0400)]
[test/font] Test more
Behdad Esfahbod [Thu, 12 May 2011 03:31:15 +0000 (23:31 -0400)]
[API] Pass face to get_table()
Behdad Esfahbod [Thu, 12 May 2011 03:25:28 +0000 (23:25 -0400)]
[API] Remove const from font user_data
Behdad Esfahbod [Thu, 12 May 2011 03:22:55 +0000 (23:22 -0400)]
[test/font] More tests
Behdad Esfahbod [Thu, 12 May 2011 03:12:58 +0000 (23:12 -0400)]
[test/font] More tests
Behdad Esfahbod [Thu, 12 May 2011 03:07:47 +0000 (23:07 -0400)]
[test/font] More get_empty() tests
Behdad Esfahbod [Thu, 12 May 2011 03:06:02 +0000 (23:06 -0400)]
[test/font] Test get_face() / get_parent()
Behdad Esfahbod [Thu, 12 May 2011 03:05:02 +0000 (23:05 -0400)]
[test/font] Add test_font_properties()
Behdad Esfahbod [Thu, 12 May 2011 02:52:35 +0000 (22:52 -0400)]
[test/font] Start adding tests for hb-font.h
Behdad Esfahbod [Thu, 12 May 2011 02:49:29 +0000 (22:49 -0400)]
[API] Rename hb_face_create_for_data() to hb_face_create()
Behdad Esfahbod [Thu, 12 May 2011 02:33:13 +0000 (22:33 -0400)]
[API] Add hb_ot_layout_substitute_start/finish()
Behdad Esfahbod [Thu, 12 May 2011 02:23:15 +0000 (22:23 -0400)]
Remove unused hb_set_t
Behdad Esfahbod [Thu, 12 May 2011 02:21:38 +0000 (22:21 -0400)]
Remove hb_static_threadsafe_set_t
Behdad Esfahbod [Thu, 12 May 2011 02:18:31 +0000 (22:18 -0400)]
Use constructor/destructor for hb_ot_shape_plan_t
Behdad Esfahbod [Thu, 12 May 2011 02:00:56 +0000 (22:00 -0400)]
Minor
Behdad Esfahbod [Thu, 12 May 2011 01:31:25 +0000 (21:31 -0400)]
Fix compile with no mutex available
Behdad Esfahbod [Thu, 12 May 2011 01:28:01 +0000 (21:28 -0400)]
Free static mutex'es
Behdad Esfahbod [Thu, 12 May 2011 01:27:52 +0000 (21:27 -0400)]
Streamline mutex stuff
Behdad Esfahbod [Thu, 12 May 2011 01:14:34 +0000 (21:14 -0400)]
Remove unused hb_mutex_trylock()
Behdad Esfahbod [Thu, 12 May 2011 01:12:44 +0000 (21:12 -0400)]
[test] Add tests for _get_empty() funcs
Behdad Esfahbod [Thu, 12 May 2011 01:08:31 +0000 (21:08 -0400)]
[TODO] Update
Behdad Esfahbod [Wed, 11 May 2011 22:14:44 +0000 (18:14 -0400)]
[API] Add hb_*_get_empty() for all objects
Behdad Esfahbod [Wed, 11 May 2011 22:06:12 +0000 (18:06 -0400)]
[API] Add HB_UNTAG()
Useful in C API only.
Behdad Esfahbod [Wed, 11 May 2011 22:02:48 +0000 (18:02 -0400)]
Minor
Behdad Esfahbod [Wed, 11 May 2011 18:30:56 +0000 (14:30 -0400)]
Plug more leaks
All good now.
Behdad Esfahbod [Wed, 11 May 2011 18:19:18 +0000 (14:19 -0400)]
Plug leaks
Behdad Esfahbod [Wed, 11 May 2011 15:55:11 +0000 (11:55 -0400)]
[unicode] Simplify method setting
Behdad Esfahbod [Wed, 11 May 2011 15:48:28 +0000 (11:48 -0400)]
Minor
Behdad Esfahbod [Wed, 11 May 2011 14:31:33 +0000 (10:31 -0400)]
Add maxp table
Not used for anything right now. Will use to get num_glyphs in the future.
Behdad Esfahbod [Wed, 11 May 2011 12:58:21 +0000 (08:58 -0400)]
Minor
Behdad Esfahbod [Wed, 11 May 2011 04:24:34 +0000 (00:24 -0400)]
[API] Change signature of get_contour_point and get_kerning ffuncs
get_contour_point now takes glyph id before point_index.
get_kerning now takes a vector to fill-in.
Behdad Esfahbod [Wed, 11 May 2011 04:15:37 +0000 (00:15 -0400)]
[font] Do user-space conversion when chaining up to parent font
Behdad Esfahbod [Wed, 11 May 2011 04:04:15 +0000 (00:04 -0400)]
Minor
Behdad Esfahbod [Wed, 11 May 2011 04:02:02 +0000 (00:02 -0400)]
Remove hb_ot_layout_context_t, simplify code
Behdad Esfahbod [Wed, 11 May 2011 00:49:02 +0000 (20:49 -0400)]
Make default font-funcs chain-up to the parent
Behdad Esfahbod [Wed, 11 May 2011 00:41:13 +0000 (20:41 -0400)]
[API] Pass down closure user_data to font funcs
Behdad Esfahbod [Wed, 11 May 2011 00:14:44 +0000 (20:14 -0400)]
Whitespace
Behdad Esfahbod [Wed, 11 May 2011 00:04:26 +0000 (20:04 -0400)]
[API] Remove font_funcs func getter functions
Behdad Esfahbod [Wed, 11 May 2011 00:02:49 +0000 (20:02 -0400)]
[API] Add hb_font_create_sub_font() and hb_font_get_parent()
Not quite useful just yet.
Behdad Esfahbod [Tue, 10 May 2011 23:57:00 +0000 (19:57 -0400)]
[font] Fix internal sign of x/y_scale
Should have been done as part of
da975419884a535281745f30f4b32fee0bc8a7a1
Behdad Esfahbod [Tue, 10 May 2011 23:40:44 +0000 (19:40 -0400)]
[API] Remove broken-by-design hb_font_unset_funcs()
Behdad Esfahbod [Tue, 10 May 2011 23:39:32 +0000 (19:39 -0400)]
Add hb_font_make/is_immutable()
Behdad Esfahbod [Tue, 10 May 2011 23:21:07 +0000 (19:21 -0400)]
Fix build without mutex
Behdad Esfahbod [Tue, 10 May 2011 23:18:12 +0000 (19:18 -0400)]
Remove duplicate atomic_int implementation
Behdad Esfahbod [Tue, 10 May 2011 23:12:49 +0000 (19:12 -0400)]
Fix set implementation to be truly threadsafe even with destroy() callbacks
The test/object test is passing again, instead of deadlocking.
Behdad Esfahbod [Tue, 10 May 2011 23:11:27 +0000 (19:11 -0400)]
[test] Always initialize gthread such that our mutex() stuff is tested
Now the test/object test deadlocks as expected. Fix coming.
Behdad Esfahbod [Tue, 10 May 2011 23:00:07 +0000 (19:00 -0400)]
[test/object] Test user_data with destroy() callback that calls user_data
Exposes the non-atomicity of user_data opertaions at this time because
we call finish() while still locked and modifying the object. In fact,
I'm surprised that it doesn't deadlock. It should.
Behdad Esfahbod [Tue, 10 May 2011 21:55:40 +0000 (17:55 -0400)]
Add DOAP file
Behdad Esfahbod [Tue, 10 May 2011 21:48:34 +0000 (17:48 -0400)]
[test/blob] Fix bug in test
Behdad Esfahbod [Tue, 10 May 2011 21:41:44 +0000 (17:41 -0400)]
[test/blob] Add create_sub_blob()
Behdad Esfahbod [Tue, 10 May 2011 21:37:08 +0000 (17:37 -0400)]
[test] Test blob API
Behdad Esfahbod [Sat, 7 May 2011 02:28:26 +0000 (22:28 -0400)]
[API] Simplify blob API, remove lock
Behdad Esfahbod [Fri, 6 May 2011 23:30:59 +0000 (19:30 -0400)]
Add -Bsymbolic-functions to linker flags
Behdad Esfahbod [Fri, 6 May 2011 23:30:46 +0000 (19:30 -0400)]
[TODO] Update
Behdad Esfahbod [Fri, 6 May 2011 23:20:52 +0000 (19:20 -0400)]
Remove unused hb_blob_try_writable_inplace()
Behdad Esfahbod [Fri, 6 May 2011 20:28:10 +0000 (16:28 -0400)]
Add note re deadlocks
Behdad Esfahbod [Fri, 6 May 2011 04:04:28 +0000 (00:04 -0400)]
[API] Add hb_language_get_default()
It uses locale information to detect default language. It's used by
hb_shape() whenever language is not set on the buffer.
Not sure how to properly test it in the test suite. Tested by observing
that with DejaVu Sans we select the proper local glyph version for U+431
under Serbian locale. See http://www.pango.org/ScriptGallery
Behdad Esfahbod [Fri, 6 May 2011 01:31:04 +0000 (21:31 -0400)]
[API] Add version macros and functions
Step version up to 0.5.0.
Also, fix to pass "make distcheck"
Behdad Esfahbod [Thu, 5 May 2011 23:47:59 +0000 (19:47 -0400)]
Add disable-static libtool flag
No one who builds harfbuzz static uses the autotools build system to do it.
Behdad Esfahbod [Thu, 5 May 2011 23:47:12 +0000 (19:47 -0400)]
Remove win32-dll libtool flag
Since we're not win32-dll clean the way libtool docs define it.
Behdad Esfahbod [Thu, 5 May 2011 23:37:53 +0000 (19:37 -0400)]
Add AC_CANONICAL_HOST
Behdad Esfahbod [Thu, 5 May 2011 20:24:42 +0000 (16:24 -0400)]
Fix build with older glib
Behdad Esfahbod [Thu, 5 May 2011 20:09:45 +0000 (16:09 -0400)]
[buffer] Remove wrong optimization
While the cluster fields of the glyph string are usually sorted, they
wouldn't be in special cases (for example for non-native direction).
Blindly using bsearch is plain wrong. If we want to reintroduce this
optimization we have to make sure we know the buffer clusters are
monotonic and in which direction. Not sure it's worth it though.
Behdad Esfahbod [Thu, 5 May 2011 19:33:19 +0000 (15:33 -0400)]
Make user_data access threadsafe
For now, by taking a global user_data mutex.
Behdad Esfahbod [Thu, 5 May 2011 19:28:37 +0000 (15:28 -0400)]
Shrink code
Behdad Esfahbod [Thu, 5 May 2011 19:14:04 +0000 (15:14 -0400)]
Use threadsafe set implementation for hb_language lookups
Note that the static variable has to be a global static, as gcc
implements local statics differently and that would require linking
to libstdc++, which we don't want.
Behdad Esfahbod [Thu, 5 May 2011 19:07:54 +0000 (15:07 -0400)]
Add hb_threadsafe_set_t
Behdad Esfahbod [Thu, 5 May 2011 19:00:43 +0000 (15:00 -0400)]
Use hb_array_t for hb_language_t mapping
Behdad Esfahbod [Thu, 5 May 2011 18:47:53 +0000 (14:47 -0400)]
Move code around
Behdad Esfahbod [Thu, 5 May 2011 18:39:24 +0000 (14:39 -0400)]
[TODO] Remove done items
Behdad Esfahbod [Thu, 5 May 2011 18:38:16 +0000 (14:38 -0400)]
Replace fixed-size lookup_maps array with hb_array_t
Behdad Esfahbod [Thu, 5 May 2011 18:12:37 +0000 (14:12 -0400)]
Replace fixed-size feature_maps array with hb_array_t
Behdad Esfahbod [Thu, 5 May 2011 17:42:19 +0000 (13:42 -0400)]
Replace fixed-size feature_infos array with hb_array_t
Behdad Esfahbod [Thu, 5 May 2011 17:24:07 +0000 (13:24 -0400)]
Minor
Behdad Esfahbod [Thu, 5 May 2011 17:21:04 +0000 (13:21 -0400)]
Replace hb_map_t with hb_set_t which is more intuitive and flexible
Behdad Esfahbod [Thu, 5 May 2011 16:39:51 +0000 (12:39 -0400)]
Make array/map implementation more generic
Behdad Esfahbod [Thu, 5 May 2011 04:21:16 +0000 (00:21 -0400)]
Make hb_mutex_*() macros take a pointer
More intuitive.
Behdad Esfahbod [Thu, 5 May 2011 04:17:43 +0000 (00:17 -0400)]
Add hb_static_mutex_t
Behdad Esfahbod [Wed, 4 May 2011 23:27:37 +0000 (19:27 -0400)]
Move code around
Mutex (and Windows.h by extension) are fairly isolated now.
Behdad Esfahbod [Tue, 3 May 2011 05:03:53 +0000 (01:03 -0400)]
[ft] Fix font->face handling
Don't use _cached()