platform/upstream/libHarfBuzzSharp.git
5 years agoMerge pull request #1441 from harfbuzz/cff-doublenum
Behdad Esfahbod [Tue, 4 Dec 2018 09:14:13 +0000 (04:14 -0500)]
Merge pull request #1441 from harfbuzz/cff-doublenum

use double as CFF Number implementation

5 years agomore double changes
Michiharu Ariza [Tue, 4 Dec 2018 00:59:41 +0000 (16:59 -0800)]
more double changes

5 years agoadded minimized test cases
Michiharu Ariza [Tue, 4 Dec 2018 00:18:10 +0000 (16:18 -0800)]
added minimized test cases

5 years agoadditional precision made a difference in extents test
Michiharu Ariza [Tue, 4 Dec 2018 00:06:58 +0000 (16:06 -0800)]
additional precision made a difference in extents test

5 years agomore double changes
Michiharu Ariza [Mon, 3 Dec 2018 23:32:42 +0000 (15:32 -0800)]
more double changes

5 years agoNumber to use double for all types
Michiharu Ariza [Mon, 3 Dec 2018 22:48:46 +0000 (14:48 -0800)]
Number to use double for all types

5 years agoMerge pull request #1439 from ebraminio/tracking
Behdad Esfahbod [Mon, 3 Dec 2018 19:30:51 +0000 (14:30 -0500)]
Merge pull request #1439 from ebraminio/tracking

[aat] Expose hb_aat_layout_has_tracking API

5 years ago[aat] Expose hb_aat_layout_has_tracking API
Ebrahim Byagowi [Mon, 3 Dec 2018 18:34:48 +0000 (22:04 +0330)]
[aat] Expose hb_aat_layout_has_tracking API

5 years ago[aat] Expose a couple APIs
Behdad Esfahbod [Mon, 3 Dec 2018 15:41:37 +0000 (10:41 -0500)]
[aat] Expose a couple APIs

New API:
+hb_aat_layout_has_substitution()
+hb_aat_layout_has_positioning()

5 years ago[aat] Fix division sign fallout
Behdad Esfahbod [Sun, 2 Dec 2018 17:38:53 +0000 (12:38 -0500)]
[aat] Fix division sign fallout

Happened after 11d2f49af8f53340134c844173f4d8655b00dea3
since now nClasses is unsigned int...

5 years agoAdd casts to hb_array_t<>
Behdad Esfahbod [Sun, 2 Dec 2018 01:04:45 +0000 (20:04 -0500)]
Add casts to hb_array_t<>

5 years agoSimplify build source list
Behdad Esfahbod [Sun, 2 Dec 2018 00:34:18 +0000 (19:34 -0500)]
Simplify build source list

5 years agoMinor
Behdad Esfahbod [Sun, 2 Dec 2018 00:28:17 +0000 (19:28 -0500)]
Minor

5 years agoNew approach to change BigEndian casts to be int-sized
Behdad Esfahbod [Sat, 1 Dec 2018 18:12:21 +0000 (13:12 -0500)]
New approach to change BigEndian casts to be int-sized

Fixes spurious warnings like:
Fixes https://github.com/harfbuzz/harfbuzz/issues/1436

5 years agoChange hb_assert_unsigned_t<> to hb_is_signed<>
Behdad Esfahbod [Sat, 1 Dec 2018 18:03:52 +0000 (13:03 -0500)]
Change hb_assert_unsigned_t<> to hb_is_signed<>

5 years agoRevert "Change BigEndian casts to be int-sized"
Behdad Esfahbod [Sat, 1 Dec 2018 05:14:48 +0000 (00:14 -0500)]
Revert "Change BigEndian casts to be int-sized"

This reverts commit eb5ddd32bf4e458ca0af8d5784f8fd46485ad225.

Broke tests, badly.  To be investigated and reenabled.

5 years agoMinor
Behdad Esfahbod [Sat, 1 Dec 2018 05:04:29 +0000 (00:04 -0500)]
Minor

5 years agoChange BigEndian casts to be int-sized
Behdad Esfahbod [Sat, 1 Dec 2018 05:03:01 +0000 (00:03 -0500)]
Change BigEndian casts to be int-sized

Fixes spurious warnings like:
Fixes https://github.com/harfbuzz/harfbuzz/issues/1436

5 years agoFix more warning
Behdad Esfahbod [Sat, 1 Dec 2018 04:38:24 +0000 (23:38 -0500)]
Fix more warning

5 years agoFix warning
Behdad Esfahbod [Sat, 1 Dec 2018 04:20:50 +0000 (23:20 -0500)]
Fix warning

5 years ago[CFF] Change spaces to tabs
Behdad Esfahbod [Sat, 1 Dec 2018 04:04:59 +0000 (23:04 -0500)]
[CFF] Change spaces to tabs

5 years ago[CFF] Whitespace
Behdad Esfahbod [Sat, 1 Dec 2018 03:54:57 +0000 (22:54 -0500)]
[CFF] Whitespace

5 years agoMinor
Behdad Esfahbod [Sat, 1 Dec 2018 03:54:20 +0000 (22:54 -0500)]
Minor

5 years agoRemove generated files that were accidentally added by CFF2 branch
Behdad Esfahbod [Sat, 1 Dec 2018 03:43:22 +0000 (22:43 -0500)]
Remove generated files that were accidentally added by CFF2 branch

5 years agoMerge pull request #1113 from harfbuzz/cff-subset
Behdad Esfahbod [Sat, 1 Dec 2018 03:40:54 +0000 (22:40 -0500)]
Merge pull request #1113 from harfbuzz/cff-subset

CFF/CFF2 subsetter

5 years agoUse nullptr instead of 0/NULL. (#1435)
Bruce Mitchener [Sat, 1 Dec 2018 03:22:49 +0000 (10:22 +0700)]
Use nullptr instead of 0/NULL. (#1435)

* Use nullptr instead of 0/NULL.

* Update test-name-table.cc

5 years agoundo 0u
Michiharu Ariza [Sat, 1 Dec 2018 02:58:14 +0000 (18:58 -0800)]
undo 0u

5 years agoMerge branch 'master' into cff-subset
Michiharu Ariza [Sat, 1 Dec 2018 02:47:34 +0000 (18:47 -0800)]
Merge branch 'master' into cff-subset

5 years agoMerge pull request #1433 from harfbuzz/overload-overload
Behdad Esfahbod [Sat, 1 Dec 2018 02:36:35 +0000 (21:36 -0500)]
Merge pull request #1433 from harfbuzz/overload-overload

Fix ambiguous overload errors with old compilers

5 years agoRevert ugly fixes
Behdad Esfahbod [Sat, 1 Dec 2018 01:45:40 +0000 (20:45 -0500)]
Revert ugly fixes

Now that we have 6daf45e0, revert cryptic hacks...

This reverts commit abd81ed4f5cbc5a94171747909bc6b77551cb929.
This reverts commit 9c6921c08c905a0cf45ba0182134e6ff910fac51.
This reverts commit d39760cabfe4007cefdfc45231e85e93fababac2.
This reverts commit fedd8e6c176dea85194693399e50243eb1c117c4.

Fixes https://github.com/harfbuzz/harfbuzz/issues/1374

5 years agoMake operator [] take signed int
Behdad Esfahbod [Sat, 1 Dec 2018 00:57:12 +0000 (19:57 -0500)]
Make operator [] take signed int

The built-in operator takes signed int.  So, match it, such that
the built-in is never a better or equally-good match to our operator.
Fixes "ambiguous overload" errors from gcc 4.2 and VS 2008.

See https://github.com/harfbuzz/harfbuzz/issues/1374

5 years agoredo fixes
Michiharu Ariza [Sat, 1 Dec 2018 00:59:41 +0000 (16:59 -0800)]
redo fixes

5 years agoyet another
Michiharu Ariza [Sat, 1 Dec 2018 00:50:46 +0000 (16:50 -0800)]
yet another

5 years agosome more
Michiharu Ariza [Sat, 1 Dec 2018 00:31:01 +0000 (16:31 -0800)]
some more

5 years agofix another pesky gcc error
Michiharu Ariza [Sat, 1 Dec 2018 00:07:08 +0000 (16:07 -0800)]
fix another pesky gcc error

5 years agosilence picky gcc errors
Michiharu Ariza [Fri, 30 Nov 2018 23:15:31 +0000 (15:15 -0800)]
silence picky gcc errors

5 years agoMerge branch 'master' into cff-subset
Michiharu Ariza [Fri, 30 Nov 2018 23:00:52 +0000 (15:00 -0800)]
Merge branch 'master' into cff-subset

5 years agoOne more.........
Behdad Esfahbod [Fri, 30 Nov 2018 21:50:30 +0000 (16:50 -0500)]
One more.........

I wonder if there's something better to do about these :(.

In file included from hb-ot-color.cc:31:
hb-ot-color-cpal-table.hh: In member function 'unsigned int OT::CPAL::get_size() const':
hb-ot-color-cpal-table.hh:118: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:

5 years agoOne more....
Behdad Esfahbod [Fri, 30 Nov 2018 21:04:52 +0000 (16:04 -0500)]
One more....

hb-ot-vorg-table.hh:96: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:87: note: candidate 1: const Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) const [with Type = OT::VertOriginMetric, unsigned int PreallocedCount = 8u]
hb-ot-vorg-table.hh:96: note: candidate 2: operator[](const T*, int) <built-in>

5 years agoOne more...
Behdad Esfahbod [Fri, 30 Nov 2018 20:55:30 +0000 (15:55 -0500)]
One more...

Sigh.

hb-ot-kern-table.hh: In member function 'int OT::KernSubTableFormat3<KernSubTableHeader>::get_kerning(hb_codepoint_t, hb_codepoint_t) const':
hb-ot-kern-table.hh:59: error: ambiguous overload for 'operator[]' in 'kernValue[kernIndex[i]]'
hb-ot-kern-table.hh:59: note: candidates are: operator[](T*, int) <built-in>
hb-dsalgs.hh:574: note:                 Type& hb_array_t<Type>::operator[](unsigned int) const [with Type = const OT::IntType<short int, 2u>]

5 years agoMore...
Behdad Esfahbod [Fri, 30 Nov 2018 20:16:57 +0000 (15:16 -0500)]
More...

hb-ot-layout-gsubgpos.hh:1707: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
...

5 years agoMore weird fixes
Behdad Esfahbod [Fri, 30 Nov 2018 17:52:21 +0000 (12:52 -0500)]
More weird fixes

In file included from hb-ot-name.cc:29:
hb-ot-name-table.hh: In member function 'unsigned int OT::name::get_size() const':
hb-ot-name-table.hh:157: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-open-type.hh:354: note: candidate 1: const Type& OT::UnsizedArrayOf<Type>::operator[](unsigned int) const [with Type = OT::NameRecord]
hb-ot-name-table.hh:157: note: candidate 2: operator[](const T*, int) <built-in>
hb-ot-name-table.hh: In member function 'void OT::name::accelerator_t::init(hb_face_t*)':
hb-ot-name-table.hh:196: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-dsalgs.hh:574: note: candidate 1: Type& hb_array_t<Type>::operator[](unsigned int) const [with Type = const OT::NameRecord]
hb-ot-name-table.hh:196: note: candidate 2: operator[](T*, int) <built-in>
hb-ot-name-table.hh:197: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-dsalgs.hh:574: note: candidate 1: Type& hb_array_t<Type>::operator[](unsigned int) const [with Type = const OT::NameRecord]
hb-ot-name-table.hh:197: note: candidate 2: operator[](T*, int) <built-in>
hb-ot-name-table.hh:198: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-dsalgs.hh:574: note: candidate 1: Type& hb_array_t<Type>::operator[](unsigned int) const [with Type = const OT::NameRecord]
hb-ot-name-table.hh:198: note: candidate 2: operator[](T*, int) <built-in>
make[4]: *** [libharfbuzz_la-hb-ot-name.lo] Error 1
make[3]: *** [all-recursive] Error 1

5 years agoMerge commit 'ae79fdaa7774d3f886a8f03926577c3bd2010b03'
Behdad Esfahbod [Fri, 30 Nov 2018 16:53:30 +0000 (11:53 -0500)]
Merge commit 'ae79fdaa7774d3f886a8f03926577c3bd2010b03'

5 years agoUmm. Cryptic, yes
Behdad Esfahbod [Fri, 30 Nov 2018 16:51:26 +0000 (11:51 -0500)]
Umm.  Cryptic, yes

In file included from hb-face.cc:35:
hb-ot-cmap-table.hh: In member function 'void OT::CmapSubtableFormat4::_compiles_assertion_on_line_388() const':
hb-ot-cmap-table.hh:388: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-open-type.hh:354: note: candidate 1: const Type& OT::UnsizedArrayOf<Type>::operator[](unsigned int) const [with Type = OT::IntType<short unsigned int, 2u>]
hb-ot-cmap-table.hh:388: note: candidate 2: operator[](const T*, int) <built-in>
hb-ot-cmap-table.hh: In member function 'void OT::CmapSubtableFormat4::_instance_assertion_on_line_388() const':
hb-ot-cmap-table.hh:388: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-open-type.hh:354: note: candidate 1: const Type& OT::UnsizedArrayOf<Type>::operator[](unsigned int) const [with Type = OT::IntType<short unsigned int, 2u>]
hb-ot-cmap-table.hh:388: note: candidate 2: operator[](const T*, int) <built-in>
hb-face.cc: In function 'hb_blob_t* _hb_face_builder_data_reference_blob(hb_face_builder_data_t*)':
hb-face.cc:650: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:81: note: candidate 1: Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) [with Type = hb_face_builder_data_t::table_entry_t, unsigned int PreallocedCount = 32u]
hb-face.cc:650: note: candidate 2: operator[](T*, int) <built-in>
hb-face.cc:650: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:81: note: candidate 1: Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) [with Type = hb_face_builder_data_t::table_entry_t, unsigned int PreallocedCount = 32u]
hb-face.cc:650: note: candidate 2: operator[](const T*, int) <built-in>
hb-face.cc:651: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:81: note: candidate 1: Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) [with Type = hb_face_builder_data_t::table_entry_t, unsigned int PreallocedCount = 32u]
hb-face.cc:651: note: candidate 2: operator[](T*, int) <built-in>
hb-face.cc:651: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:81: note: candidate 1: Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) [with Type = hb_face_builder_data_t::table_entry_t, unsigned int PreallocedCount = 32u]
hb-face.cc:651: note: candidate 2: operator[](const T*, int) <built-in>

5 years agoUmm. Cryptic, yes
Behdad Esfahbod [Fri, 30 Nov 2018 16:51:26 +0000 (11:51 -0500)]
Umm.  Cryptic, yes

hb-face.cc:650: error: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
hb-vector.hh:81: note: candidate 1: Type& hb_vector_t<Type, PreallocedCount>::operator[](unsigned int) [with Type = hb_face_builder_data_t::table_entry_t, unsigned int PreallocedCount = 32u]
hb-face.cc:650: note: candidate 2: operator[](T*, int) <built-in>

5 years agoFix a few more sizeof(vector[0]) errors with weird compilers
Behdad Esfahbod [Fri, 30 Nov 2018 16:31:39 +0000 (11:31 -0500)]
Fix a few more sizeof(vector[0]) errors with weird compilers

5 years agoFix build for realz
Behdad Esfahbod [Fri, 30 Nov 2018 06:11:04 +0000 (01:11 -0500)]
Fix build for realz

5 years agoFix build
Behdad Esfahbod [Fri, 30 Nov 2018 05:44:40 +0000 (00:44 -0500)]
Fix build

5 years ago[post] Rename v2 to v2X
Behdad Esfahbod [Fri, 30 Nov 2018 05:38:08 +0000 (00:38 -0500)]
[post] Rename v2 to v2X

5 years agoMove things
Behdad Esfahbod [Fri, 30 Nov 2018 05:32:12 +0000 (00:32 -0500)]
Move things

5 years agoFix build for realz
Behdad Esfahbod [Thu, 29 Nov 2018 21:29:30 +0000 (16:29 -0500)]
Fix build for realz

5 years ago[uniscribe] Fix build
Behdad Esfahbod [Thu, 29 Nov 2018 21:03:15 +0000 (16:03 -0500)]
[uniscribe] Fix build

5 years ago[test][aat] Remove extra --shaper ot
Ebrahim Byagowi [Thu, 29 Nov 2018 20:36:40 +0000 (00:06 +0330)]
[test][aat] Remove extra --shaper ot

As run-tests.py already adds it

5 years ago[ci] Re-enable llvm-gcc-4.2 bots (#1429)
Ebrahim Byagowi [Thu, 29 Nov 2018 20:27:50 +0000 (23:57 +0330)]
[ci] Re-enable llvm-gcc-4.2 bots (#1429)

5 years agoLimit __builtin_bswap16 to GCC >= 5 as it was implemented on 4.8
Ebrahim Byagowi [Thu, 29 Nov 2018 20:17:20 +0000 (23:47 +0330)]
Limit __builtin_bswap16 to GCC >= 5 as it was implemented on 4.8

5 years ago[coretext] Fix compile
Behdad Esfahbod [Thu, 29 Nov 2018 20:07:44 +0000 (15:07 -0500)]
[coretext] Fix compile

Fingers crossed.

5 years agoConvert "static const hb_tag_t" constants to enum
Behdad Esfahbod [Thu, 29 Nov 2018 20:04:34 +0000 (15:04 -0500)]
Convert "static const hb_tag_t" constants to enum

5 years agoConvert misc "static const" constants to enum
Behdad Esfahbod [Thu, 29 Nov 2018 20:01:10 +0000 (15:01 -0500)]
Convert misc "static const" constants to enum

5 years agoConvert "static const bool" constants to anonymous enum
Behdad Esfahbod [Thu, 29 Nov 2018 19:53:43 +0000 (14:53 -0500)]
Convert "static const bool" constants to anonymous enum

5 years ago[vector] Make pointer cast explicit
Behdad Esfahbod [Thu, 29 Nov 2018 19:34:44 +0000 (14:34 -0500)]
[vector] Make pointer cast explicit

Too bad this doesn't help MSVC 2008 build, as explicit operators are
C++11.

5 years agoHand-hold older compilers
Behdad Esfahbod [Thu, 29 Nov 2018 19:28:44 +0000 (14:28 -0500)]
Hand-hold older compilers

5 years ago[test/text-rendering-tests] Update from upstream
Behdad Esfahbod [Thu, 29 Nov 2018 17:39:52 +0000 (12:39 -0500)]
[test/text-rendering-tests] Update from upstream

5 years agoMake shaper's override_features() override user features as well
Behdad Esfahbod [Thu, 29 Nov 2018 17:32:47 +0000 (12:32 -0500)]
Make shaper's override_features() override user features as well

The override_features is used to override features that are normally
discretionary features, but in a specific shaper are for various
reasons desired to be bolted on or off, because they've been used
for inherent shaping.  As such, it makes sense that they also
override user features.  Ie. if user turned 'liga' on, we don't
want Khmer shaping to become broken...  Or turn 'clig' off...

Fixes https://github.com/harfbuzz/harfbuzz/issues/1310

5 years ago[khmer] Move 'clig' to overrides
Behdad Esfahbod [Thu, 29 Nov 2018 17:30:14 +0000 (12:30 -0500)]
[khmer] Move 'clig' to overrides

Prerequisite for https://github.com/harfbuzz/harfbuzz/issues/1310

5 years agoFix "typename outside template" issues
Behdad Esfahbod [Thu, 29 Nov 2018 17:18:14 +0000 (12:18 -0500)]
Fix "typename outside template" issues

Nothing an extra template class wouldn't fix...

Fixes https://github.com/harfbuzz/harfbuzz/issues/1419

5 years ago2.2.0
Behdad Esfahbod [Thu, 29 Nov 2018 16:53:53 +0000 (11:53 -0500)]
2.2.0

5 years ago[icu] Minor
Behdad Esfahbod [Thu, 29 Nov 2018 16:34:22 +0000 (11:34 -0500)]
[icu] Minor

5 years agoMerge pull request #1418 from gvictor/replace_icu_deprecated
Behdad Esfahbod [Thu, 29 Nov 2018 15:39:18 +0000 (10:39 -0500)]
Merge pull request #1418 from gvictor/replace_icu_deprecated

Replace @Deprecated ICU API - USCRIPT_CODE_LIMIT

5 years ago[test][aat.kern] More (#1427)
Ebrahim Byagowi [Thu, 29 Nov 2018 08:06:05 +0000 (11:36 +0330)]
[test][aat.kern] More (#1427)

5 years agoworkaround for issue #1417
Michiharu Ariza [Thu, 29 Nov 2018 01:06:11 +0000 (17:06 -0800)]
workaround for issue #1417

5 years agoMerge branch 'master' into cff-subset
Michiharu Ariza [Thu, 29 Nov 2018 00:27:45 +0000 (16:27 -0800)]
Merge branch 'master' into cff-subset

5 years agoMerge pull request #1425 from mbutterick/patch-1
Behdad Esfahbod [Wed, 28 Nov 2018 22:21:57 +0000 (17:21 -0500)]
Merge pull request #1425 from mbutterick/patch-1

update simple shaping example (closes #298)

5 years agoupdate simple shaping example (closes #298)
Matthew Butterick [Wed, 28 Nov 2018 21:46:12 +0000 (13:46 -0800)]
update simple shaping example (closes #298)

add call to `FT_Set_Char_Size`, otherwise default size remains at `0`, and glyph positions come back as `0` too

5 years ago[test][aat] Update expectency
Ebrahim Byagowi [Wed, 28 Nov 2018 21:25:05 +0000 (00:55 +0330)]
[test][aat] Update expectency

It is not visually noticeable but apparently affected by kern format2 correct implementation.
I should've checked CoreText result which can't as CircleCI outage.

5 years ago[test] Add test for format2 kern (#1423)
Ebrahim Byagowi [Wed, 28 Nov 2018 20:37:00 +0000 (00:07 +0330)]
[test] Add test for format2 kern (#1423)

5 years ago[kerx] Fix Format2 index calc again
Behdad Esfahbod [Wed, 28 Nov 2018 20:24:30 +0000 (15:24 -0500)]
[kerx] Fix Format2 index calc again

5 years agoMerge remote-tracking branch 'fdo/master'
Behdad Esfahbod [Wed, 28 Nov 2018 20:08:01 +0000 (15:08 -0500)]
Merge remote-tracking branch 'fdo/master'

5 years ago[kern] Fix kern table Format2 offsetToIndex
Behdad Esfahbod [Wed, 28 Nov 2018 20:06:01 +0000 (15:06 -0500)]
[kern] Fix kern table Format2 offsetToIndex

Fixes https://github.com/harfbuzz/harfbuzz/issues/1421

5 years ago[kerx] Fix ClassTable implementation for 'kern' table Format 2
Behdad Esfahbod [Wed, 28 Nov 2018 19:51:56 +0000 (14:51 -0500)]
[kerx] Fix ClassTable implementation for 'kern' table Format 2

5 years ago[aat] Towards adding two ClassTable's
Behdad Esfahbod [Wed, 28 Nov 2018 19:46:26 +0000 (14:46 -0500)]
[aat] Towards adding two ClassTable's

5 years ago[test][aat] Add a test and make macOS runners faster (#1422)
Ebrahim Byagowi [Wed, 28 Nov 2018 16:58:42 +0000 (20:28 +0330)]
[test][aat] Add a test and make macOS runners faster (#1422)

5 years agoMerge pull request #1398 from Adenilson/bigInt01
Behdad Esfahbod [Tue, 27 Nov 2018 22:40:32 +0000 (17:40 -0500)]
Merge pull request #1398 from Adenilson/bigInt01

Optimize harfbuzz big integer conversions

5 years agoComment
Behdad Esfahbod [Tue, 27 Nov 2018 22:40:09 +0000 (17:40 -0500)]
Comment

5 years agoOptimize harfbuzz big integer conversions
Adenilson Cavalcanti [Tue, 20 Nov 2018 22:41:19 +0000 (14:41 -0800)]
Optimize harfbuzz big integer conversions

Profiling showed that type conversions were adding considerable cycles in time
spent doing text shaping.

The idea is to optimize it using native processor instructions to help Blink
layout performance.

Doing further investigation revelead that compilers may not use the
proper instruction on ARM 32bits builds (i.e. REV16).

One way to insure that the generated ASM was ideal for both gcc/clang
was using __builtin_bswap16.

Added bonus is that we no longer need to test for CPU architecture.

5 years agoReplace @Deprecated ICU API - USCRIPT_CODE_LIMIT
Victor Chang [Tue, 27 Nov 2018 11:34:56 +0000 (11:34 +0000)]
Replace @Deprecated ICU API - USCRIPT_CODE_LIMIT

Use of the deprecated API USCRIPT_CODE_LIMIT prevents harfbuzz
using the ICU4C as a shared library.

The API has been replaced by u_getIntPropertyMaxValue(UCHAR_SCRIPT)

5 years ago[ci] Add .codecov.yml
Ebrahim Byagowi [Mon, 26 Nov 2018 15:28:39 +0000 (18:58 +0330)]
[ci] Add .codecov.yml

Similar to https://github.com/GoogleChrome/lighthouse/blob/master/.codecov.yml

No strong preference on commenting, feel free to enable it again

5 years ago[test][aat] Enable Tamil MN test (#1414)
Ebrahim Byagowi [Mon, 26 Nov 2018 13:28:58 +0000 (16:58 +0330)]
[test][aat] Enable Tamil MN test (#1414)

5 years ago[aat] Ignore GSUB table of Muthu Foundry if they have morx table
Behdad Esfahbod [Sun, 25 Nov 2018 21:51:22 +0000 (16:51 -0500)]
[aat] Ignore GSUB table of Muthu Foundry if they have morx table

Fixes https://github.com/harfbuzz/harfbuzz/issues/1410

5 years ago[GDEF] Move more code
Behdad Esfahbod [Sun, 25 Nov 2018 21:38:36 +0000 (16:38 -0500)]
[GDEF] Move more code

5 years ago[GDEF] Move code around
Behdad Esfahbod [Sun, 25 Nov 2018 20:59:18 +0000 (15:59 -0500)]
[GDEF] Move code around

5 years ago[GDEF] Minor
Behdad Esfahbod [Sun, 25 Nov 2018 20:51:01 +0000 (15:51 -0500)]
[GDEF] Minor

5 years ago[test][aat] fix 10.13.6 Helvetica expectation
Ebrahim Byagowi [Sun, 25 Nov 2018 10:07:23 +0000 (13:37 +0330)]
[test][aat] fix 10.13.6 Helvetica expectation

5 years ago[aat] Add m grave test (#1412)
Ebrahim Byagowi [Sun, 25 Nov 2018 09:20:30 +0000 (12:50 +0330)]
[aat] Add m grave test (#1412)

5 years ago[aat] Fix macos expectation
Ebrahim Byagowi [Sun, 25 Nov 2018 07:55:17 +0000 (11:25 +0330)]
[aat] Fix macos expectation

5 years ago[aat.feat] Rework API and implementation
Behdad Esfahbod [Sun, 25 Nov 2018 06:14:40 +0000 (01:14 -0500)]
[aat.feat] Rework API and implementation

Fixes https://github.com/harfbuzz/harfbuzz/pull/1346

5 years ago[aat.feat] Add _MAX_VALUE to enums
Behdad Esfahbod [Sat, 24 Nov 2018 17:08:36 +0000 (12:08 -0500)]
[aat.feat] Add _MAX_VALUE to enums

5 years ago[aat.feat] Minor
Behdad Esfahbod [Sat, 24 Nov 2018 16:02:04 +0000 (11:02 -0500)]
[aat.feat] Minor

5 years ago[aat.feat] Whitespace
Behdad Esfahbod [Sat, 24 Nov 2018 16:00:42 +0000 (11:00 -0500)]
[aat.feat] Whitespace

5 years ago[aat.feat] Port to SortedUnsizedArrayOf.bsearch()
Behdad Esfahbod [Sat, 24 Nov 2018 15:54:40 +0000 (10:54 -0500)]
[aat.feat] Port to SortedUnsizedArrayOf.bsearch()

5 years ago[AAT.feat] Use lsearch for looking up SettingName's
Behdad Esfahbod [Sat, 24 Nov 2018 15:46:56 +0000 (10:46 -0500)]
[AAT.feat] Use lsearch for looking up SettingName's

They are not sorted.