a9fe787a11
[sanitizer] Add reset_object(), make set_object() do bounds-check
...
Affects morx/kerx run-time only currently. Will adjust their sanitize next.
2018-11-22 22:12:36 -05:00
2c8188bf59
[kerx] Make sure subtables are non-zero-length
...
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=11400
2018-11-22 22:02:19 -05:00
a9e0bdc35d
[GSUB] Don't flush glyphset during recursion in closure()
...
See comment.
Supercedes https://github.com/harfbuzz/harfbuzz/pull/1401
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=11366
2018-11-22 21:30:04 -05:00
d0e81b2cc8
[set] Rename
2018-11-22 21:20:39 -05:00
e866910579
Enforce requiring null_size even if min_size is 0
...
This concludes null-size enforcement changes
2018-11-22 18:07:59 -05:00
4d4fd64ff4
Allow non-nullable OffsetTo<> to non-Null'able objects
2018-11-22 18:07:36 -05:00
fa9f585ec5
[glyf] Don't mark structs UNBOUNDED
...
See comments.
2018-11-22 17:56:51 -05:00
f47a60a754
Mark UnsizedArrayOf<> as UNBOUNDED
...
Since min_size is 0, Null() still accepts this type.
2018-11-22 17:53:29 -05:00
199a0f7b41
[ot-shape] Simplify logic
2018-11-22 17:31:07 -05:00
014e4980ed
Move
2018-11-22 16:21:49 -05:00
62890dee18
[aat] Zero mark advances if decided so even if there's cross-kerning
...
Cross-kerning can only take care of positioning vertically. It doesn't
adjust mark advance...
2018-11-22 16:20:29 -05:00
a201fa74cd
[aat] Tweak fallback positioning logic when applying morx
...
Such that for Indic-like scripts (eg. Khmer), we don't do any fallback mark
advance-zeroing / positioning, but we do for Latin, etc. Reuses preferences
of our script-specific OpenType shapers for those.
Fixes regression: https://github.com/harfbuzz/harfbuzz/issues/1393
Which means, fixes again: https://github.com/harfbuzz/harfbuzz/issues/1264
While not regressing: https://github.com/harfbuzz/harfbuzz/issues/1357
2018-11-22 15:52:29 -05:00
fa0bd8964d
[myanmar] Minor move
2018-11-22 14:46:39 -05:00
7dc561984b
[myanmar] If there's no GSUB table, pick myanmar shaper
...
Needed for morx+kern mark-zeroing interaction. All other scripts
work this way.
2018-11-22 14:45:46 -05:00
25f52f58c2
[myanmar] Remove myanmar_old shaper
...
Over time it has become the same as default shaper. So, remove.
2018-11-22 14:41:01 -05:00
eeed802b1d
Fix spurious gcc warnings
...
../../src/hb-null.hh:53:39: warning: enum constant in boolean context [-Wint-in-bool-context]
2018-11-22 01:53:36 -05:00
b96ecb9971
More
...
This makes more of the gcc spurious warning:
../../src/hb-null.hh:53:39: warning: enum constant in boolean context [-Wint-in-bool-context]
But not going to let that defeat correct code. Type to switch to clang
as my main compiler...
2018-11-22 01:49:12 -05:00
2737aa81e5
Fix up recent change
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1300
2018-11-22 01:44:27 -05:00
fffea5aff7
Minor
2018-11-22 01:25:34 -05:00
209b58ef73
Minor
2018-11-22 01:22:33 -05:00
3b9fd176e8
Disallow taking Null() of unbounded structs
...
Not sure I've marked all such structs. To be done as we discover.
Fixes https://github.com/harfbuzz/harfbuzz/issues/1300
2018-11-22 01:18:55 -05:00
f2b91d6510
Use Type::null_size for our structs in Null(), sizeof() for other types
2018-11-22 01:10:22 -05:00
7dd945a876
One more time..
2018-11-22 01:05:17 -05:00
d062ad10de
Fix bots happy again, hopefully
...
So, our fallback static_assert cannot be had more than once per line
of source.
2018-11-22 00:39:14 -05:00
fb10c021c8
Revert alignof() == 1 check
...
Bots not happy with using "this" inside assertion...
This reverts 2656644887
2018-11-22 00:21:49 -05:00
8d778877b8
..
2018-11-21 23:46:09 -05:00
8cfeed9948
Minor
2018-11-21 23:42:31 -05:00
e5d954a2fb
Minor
2018-11-21 23:30:50 -05:00
e987059c61
Minor
2018-11-21 23:25:06 -05:00
a2b6d308a4
Remove DEFINE_SIZE_ARRAY2
2018-11-21 23:23:49 -05:00
2656644887
Check alignof() structs are 1
2018-11-21 23:23:21 -05:00
6321fdf704
Whitespace
2018-11-21 23:19:00 -05:00
ecdceea861
Merge pull request #1399 from harfbuzz/sharada-sandhi-mark
...
Fix USE categories for U+111C9 SHARADA SANDHI MARK
2018-11-21 16:12:32 -05:00
b89c7fd3dc
Allow defining HB_USE_ATEXIT to 0
...
That's better use of that value than requiring extra macro HB_NO_ATEXIT
2018-11-21 12:32:48 -05:00
b3d5b0a5d9
Fix USE categories for U+111C9 SHARADA SANDHI MARK
2018-11-21 11:35:44 -05:00
f48bb9a393
[var] Deprecated axis enumeration API and add new version
...
New version has axis flags.
New API:
+hb_ot_var_axis_info_t
+hb_ot_var_find_axis_info()
+hb_ot_var_get_axis_infos()
Deprecated API:
-HB_OT_VAR_NO_AXIS_INDEX
-hb_ot_var_axis_t
-hb_ot_var_find_axis()
-hb_ot_var_get_axes()
2018-11-20 20:46:14 -05:00
b2d803cef6
Remove newly-added -hb_ot_var_axis_get_flags()
2018-11-20 20:46:14 -05:00
736897d7a1
[var] Make sure hb_ot_var_axis_flags_t is int-sized
2018-11-20 20:46:14 -05:00
eab5d15f61
[var] Move code
2018-11-20 20:46:14 -05:00
8295118279
Shrink the emoji table by merging adjacent ranges
2018-11-20 15:41:45 -05:00
be1828daaa
[var] Fix type of coords returned
...
Ouch. Wonder how none of the bots caught the float->int truncation.
2018-11-20 11:16:23 -05:00
831ba74382
Fix Codacy "issues"
2018-11-20 01:16:08 -05:00
3c7792ca32
[use] Fix reordering
...
Fixes #1235 .
2018-11-19 16:49:40 -05:00
587d49fc65
[fvar] Add named-instance API
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1241
2018-11-19 14:51:34 -05:00
46c0da820f
Fix build
2018-11-19 13:32:48 -05:00
56c9238d3d
[fvar] Rewrite sanitize
2018-11-19 13:09:53 -05:00
4a6a692e3e
[fvar] Use hb_array_t for axes
2018-11-19 13:06:44 -05:00
e009739601
[fvar] Minor
2018-11-19 12:53:53 -05:00
1a2eb108b8
[ot-var] Add hb_ot_var_axis_get_flags()
...
Part of https://github.com/harfbuzz/harfbuzz/issues/1241
New API:
+hb_ot_var_axis_flags_t
+hb_ot_var_axis_get_flags
2018-11-19 12:36:56 -05:00
bd6b2ba1d3
[ot-var] Add flags
...
Unfortunate that we don't have room in hb_ot_var_axis_t to expose flags :(.
2018-11-19 11:34:56 -05:00
c076c7b85c
[ot-var] Use hb_ot_name_id_t
2018-11-19 11:30:40 -05:00
e3a1a8350a
2.1.3
2018-11-16 16:53:25 -08:00
9714e114b8
Fix recent commits
2018-11-16 16:52:42 -08:00
0328a1ce41
Revert b4c6113032
...
Was causing more trouble than it solved. We use unsigned for indexing,
and it's not helpful to allow that wrapping to negative integers on
32bit machines. The only way we could work around it would have been
by accepting int64_t arg, but that's overkill.
Ignore the MSVC 2008 build issue. We don't support that compiler.
2018-11-16 16:48:28 -08:00
52f61cdb87
Detect over/under-flow in UnsizedArray::operator[]
...
Was causing bad substitutions in mort table because of WordOffsetToIndex()
producing negative numbers that were cast to unsigned int and returned as
large numbers (which was desirable, so they would be rejected), but then
they were cast to int when passed to this operator and acting as small
negative integers, which was bad...
Detect overflow. Ouch, however, now I see this still fails on 32-bit.
Guess I'm going to revert an earlier change.
2018-11-16 16:41:59 -08:00
6910ff03e6
[aat] Fix mort shaping
...
Ouch!
2018-11-16 16:11:02 -08:00
fdb29ab2b0
2.1.2
2018-11-16 15:38:11 -08:00
cff4c6087f
Fix vertical fallback space sign
...
Ouch!
Follow-up to cf203af8a3https://github.com/harfbuzz/harfbuzz/issues/1343
2018-11-16 15:04:41 -08:00
7a97f70740
Don't apply GPOS if applying morx
...
That's what Apple does, and it wouldn't degrade our OpenType performance.
Part of https://github.com/harfbuzz/harfbuzz/issues/1348
2018-11-16 14:46:40 -08:00
eafd515639
Prefer morx table if GSUB is empty (no scripts)
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1348
2018-11-16 14:45:56 -08:00
aa06574823
Minor
2018-11-16 14:31:05 -08:00
50d1a41c08
[coretext] Hopefully the last one
2018-11-16 08:52:57 -05:00
0aab861f9c
[coretext] Another round
2018-11-16 08:43:25 -05:00
0809b76a9a
[coretext] One more try..
2018-11-16 08:29:47 -05:00
78bd447594
[coretext] One more try
2018-11-16 04:10:53 -05:00
fd27a23c82
[coretext] Another build fix attemt
2018-11-16 03:57:12 -05:00
729aedf0da
[directwrite] Fix build
2018-11-16 03:26:46 -05:00
cfb9771a3b
[coretext] Try to fix
2018-11-16 03:25:07 -05:00
e3e9547365
[coretext] Unbreak build
2018-11-16 02:55:29 -05:00
ce5da0f36a
[shaper] Rewrite shaper data code to be more template-driven than macro-driven
2018-11-16 02:52:20 -05:00
cb4bf85b14
[hdmx] Fix bounds checking
...
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=11351
2018-11-16 02:02:24 -05:00
af727b4e62
[hdmx] Minor
2018-11-16 01:55:39 -05:00
9e9f16c92d
[subset] Remove invalid glyphs from glypset to retain
...
Fixes https://github.com/harfbuzz/harfbuzz/pull/1388
2018-11-16 01:48:26 -05:00
11aa0468ac
[subset] minor, adjust spaces
2018-11-16 00:02:47 +03:30
d7c50ff0d6
[math] minor, adjust spaces
2018-11-16 00:02:47 +03:30
cabe433fbb
[base] Add TODO items
2018-11-15 02:48:50 -05:00
3bf1ce748f
[shaper] Rename
2018-11-14 21:08:54 -05:00
c221dc0ba7
[ot-shape] Move code around
2018-11-14 19:57:19 -05:00
7867c2bad0
[STAT] Add table parsing ( #1384 )
2018-11-14 22:13:50 +03:30
48d16c2ab2
[hmtx] Fix signedness issue
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1248#issuecomment-438689499
2018-11-14 09:57:05 -05:00
29db2a44a6
[ot-color/svg] Note that it can be gzipped
2018-11-14 12:13:16 +03:30
3c3eb5ea9c
[aat] Disable fallback mark advance zeroing and positioning if morx applied
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1357
2018-11-13 21:10:10 -05:00
dc4225ccd1
Don't retry creating again and again in lazy_loader if create failed
...
Still does that if get_null() returns nullptr. Our shaper data objects
are like that. Shrug.
2018-11-13 20:48:46 -05:00
086235f593
Merge pull request #1382 from punchcutter/master
...
Change USE Category for Grantha Virama
2018-11-13 19:50:25 -05:00
2092f595c7
Merge pull request #1380 from kbrow1i/cygwin
...
Don't use Win32 API on Cygwin
2018-11-13 19:49:06 -05:00
c565fc3fb3
Change USE Category for Grantha Virama
...
https://github.com/harfbuzz/harfbuzz/issues/1379
2018-11-13 12:51:10 -08:00
475be9d5c6
Fix Windows build
2018-11-13 13:01:13 -05:00
fc44dea341
Use atomic ints for upem and num_glyphs on face
2018-11-13 11:54:33 -05:00
9579ed9755
Make atomic types' internal values non-mutable
...
This resulted from confusion previously...
2018-11-13 11:46:20 -05:00
c52d5bcd94
[ot-face] Add 'head' table
2018-11-13 11:41:29 -05:00
eee5b5ed04
Don't use Win32 API on Cygwin
...
Cygwin is a Posix platform to the extent possible. It should use the
Posix API except in special circumstances.
2018-11-12 21:07:34 -05:00
56f541d000
[shape-plan] Remove unused code
2018-11-12 19:46:37 -05:00
6c22f3fd95
[shape-plan] Implement fine-grained caching plans with user-features
...
Only tag, value, and global-ness of features are considered, not their
start/end offsets.
2018-11-12 19:26:01 -05:00
cc8428756a
[shape-plan] Cache shape plans with variations based on variation indices
2018-11-12 18:48:10 -05:00
8284cb9fb3
[shape-plan] Refactor more
2018-11-12 18:18:20 -05:00
1082338525
[shape-plan] Only use shape-plan key to initialize hb_ot_shape_plan_t
...
Such that we don't accidentally use info not in the cache key.
2018-11-12 18:05:02 -05:00
7ac03f88a2
[shape-plan] Minor
2018-11-12 17:50:30 -05:00
c7be933439
[shape-plan] Refactor some more
2018-11-12 17:49:15 -05:00
fc27777833
[shape-plan] Refactor more
2018-11-12 17:27:34 -05:00
566612295b
[shape-plan] Turn hb_shape_plan_proposal_t into hb_shape_plan_key_t
...
And include it in hb_shape_plan_t itself.
2018-11-12 17:19:45 -05:00
Behdad Esfahbod
David Corbett
Ebrahim Byagowi
punchcutter
Ken Brown