|
ed35dea8
|
2019-12-06T04:37:11
|
|
Fourth try... sighs
|
|
|
9168310b
|
2019-12-06T04:28:06
|
|
Fix build, third times...
|
|
|
34f5cc2c
|
2019-12-06T04:09:33
|
|
Second try at fixing build
|
|
|
14ce5ab0
|
2019-12-06T03:54:46
|
|
First try at fixing build errors
|
|
|
2274270c
|
2019-12-06T03:42:21
|
|
[algs] Streamline bsearch some more
|
|
|
bd55d4b4
|
2019-12-06T03:35:24
|
|
[algs] Streamline bsearch() API more towards hb_array_t::bsearch_impl()
Preparing to merge the two finally!
|
|
|
fd6df520
|
2019-12-06T03:00:23
|
|
[array] Isolate bsearch implementation more
|
|
|
06d3c201
|
2019-12-06T02:52:54
|
|
[array] Simplify bfind() positioning
I had copied the old scheme from fontconfig's fccharset.c. I just
convinced myself that this change is correct and produces exact
same results. But I also am skeptical. Anyone else feel like
convincing themselves as well please?
|
|
d67ba649
|
2019-12-05T13:15:21
|
|
Rename hb_array_t::in_range to hb_array_t::check_range
|
|
|
72d83a02
|
2019-12-05T13:09:48
|
|
Make hb_array_t::in_range similar to hb_sanitize_context_t::check_range
|
|
|
b1167d19
|
2019-12-05T15:46:58
|
|
Fix!
|
|
|
5ff410b4
|
2019-12-05T15:37:56
|
|
Add files, oops
|
|
|
2dc20e63
|
2019-12-05T15:28:42
|
|
Implement fallback vertical shaping from Firefox
Fixes https://github.com/harfbuzz/harfbuzz/issues/355
|
|
|
b6d0f152
|
2019-12-05T12:19:52
|
|
[indic] Fix old-spec base-finding logic w vatu feature
Fixes https://github.com/harfbuzz/harfbuzz/issues/1587
|
|
|
0414341a
|
2019-12-05T11:26:39
|
|
Prefer AAT shaping over OT if both available
Fixes https://github.com/harfbuzz/harfbuzz/issues/1529
|
|
82545c5e
|
2019-11-26T10:24:32
|
|
more coding style fixes
|
|
|
2e7c1239
|
2019-11-26T09:57:53
|
|
minor: coding style
|
|
|
ed94e410
|
2019-11-20T10:42:50
|
|
fixed api test test-ot-glyphname
|
|
|
7a4745ba
|
2019-11-20T10:15:41
|
|
added api test test-ot-glyphname
|
|
|
ab525ac6
|
2019-11-19T20:36:56
|
|
fix HB_TINY buid
|
|
|
ff64652c
|
2019-11-19T18:50:57
|
|
fix shape-fuzzer failure
|
|
|
a333b0f7
|
2019-11-19T18:01:53
|
|
add hb-ot-cff1-std-str.hh to Makefile.sources
|
|
|
61707266
|
2019-11-19T17:09:12
|
|
fix get_sid crash
|
|
|
4c4e08a1
|
2019-11-19T16:50:28
|
|
add cff1::accelerator_t::fini
|
|
|
ff87465b
|
2019-11-19T16:18:33
|
|
add get_glyph_from_name & test. use hb_string_array
|
|
|
51e77285
|
2019-11-17T02:25:39
|
|
minor
|
|
|
6d9eb9c0
|
2019-11-17T02:20:22
|
|
fixed failure with language-tags.tests
|
|
|
59fe3797
|
2019-11-16T19:47:31
|
|
fixed hb-shape-fuzzer failures in get_sid
|
|
|
e26df436
|
2019-11-16T19:12:34
|
|
fix test-ot-face crash
|
|
|
7b49042e
|
2019-11-16T18:20:16
|
|
add cff1::get_glyph_name
|
|
|
d5338ba1
|
2019-11-22T16:22:27
|
|
[ft] Add locking FT_Face API
Now that we mutex our internal FT_Face, share it with clients.
New API:
- hb_ft_font_lock_face()
- hb_ft_font_unlock_face()
|
|
|
e5c7ee9f
|
2019-11-20T14:22:01
|
|
[set] Fix undefined-behavior shift in _previous()
harfbuzz/src/hb-set.hh:138:43: runtime error: shift exponent 64 is too large for 64-bit type 'hb_set_t::page_t::elt_t' (aka 'unsigned long long')
|
|
|
5fddc5f1
|
2019-11-20T13:07:26
|
|
Use foreach_cluster
|
|
|
e4a101ed
|
2019-11-19T20:04:00
|
|
[ft] Fix for HB_NO_VAR build
|
|
76c27462
|
2019-11-19T16:51:33
|
|
Don’t modify a ccc to 6
ccc=6 will be used in Unicode 13.0.
|
|
8fdafee2
|
2019-11-18T16:44:25
|
|
undef HB_STRING_ARRAY_LENG_NAME
|
|
|
5e9863f3
|
2019-11-18T15:40:43
|
|
ARRAY_LENGTH fails with -Wglobal-constructors
|
|
|
a7598369
|
2019-11-18T15:09:24
|
|
calculate hb_string_array length automatically
|
|
57da16bc
|
2019-11-17T21:32:12
|
|
[cff] Check rcurveline and rlinecurve arguments
Make sure the number of arguments is not too many or too few, otherwise
skip the command.
|
|
|
9f1524ce
|
2019-11-17T16:17:20
|
|
[cff] Fix a typo in processing rlinecurve
We want to process all pairs of arguments except the last 6 as lines, so
should have been subtracting 6 here, otherwise if the number of
arguments happens to be multiples of 6 they will be all treated as
curves.
See https://github.com/harfbuzz/harfbuzz/pull/2016#issuecomment-554640098
|
|
|
78d5eca8
|
2019-11-18T13:09:29
|
|
[set] Fix range-based-loop condition
Alternative to https://github.com/harfbuzz/harfbuzz/pull/2046
|
|
4dae7cee
|
2019-11-13T10:08:35
|
|
[minor] Fix common typos: substition, the the.
s/substition/substitution/
s/the the/the/
|
|
|
e6351d9b
|
2019-11-11T17:39:55
|
|
Add ms-use/IndicShapingInvalidCluster.txt
|
|
|
b372c3e9
|
2019-11-08T20:59:48
|
|
Insert a dotted circle within <U+0B85, U+0BC2>
|
|
|
64a45be5
|
2019-11-09T12:25:33
|
|
[ubsan] Don't decrease pointer if match_glyph_data is null (#2048)
Similar to fix on https://github.com/harfbuzz/harfbuzz/pull/2022
Fixes https://crbug.com/1023070
|
|
878e3588
|
2019-11-05T15:26:01
|
|
[cmake] Add harfbuzz-icu library
|
|
|
84b86a12
|
2019-11-06T09:22:34
|
|
[fuzz] Remove just added case
It didn't fail locally but on bots causing timeout, let remove for now.
|
|
|
a8f049c9
|
2019-11-05T22:26:36
|
|
[fuzz] Upload testcase of https://crbug.com/oss-fuzz/18529
Apparently false alarm per last comment and was ok locally also but lets have it here also
|
|
|
7cde68f1
|
2019-11-02T19:04:14
|
|
[ci] install pip in fedora bot
|
|
|
46178c03
|
2019-11-02T14:36:30
|
|
[ci] fix fedora bot
|
|
|
486754a8
|
2019-09-23T23:48:08
|
|
[serialize] Extract iterable copy, copy_all
|
|
|
35218c48
|
2019-10-31T13:19:44
|
|
Minor
Allow empty HB_VAR_ARRAY definition. Though, doesn't compile with any
compiler I know of.
|
|
|
1a1d64bc
|
2019-10-31T12:50:47
|
|
If applying morx, apply kern table even if GPOS is present and has kern
Since we won't be applying GPOS if morx...
To be adjusted as I receive more information from Ned. But for now
fixes this:
$ ./hb-shape GillSans.ttc Ty
[T=0+1109|y=1@-128,0+769]
Fixes https://github.com/harfbuzz/harfbuzz/issues/1982 for now.
|
|
|
3a74ee52
|
2019-10-29T12:27:03
|
|
2.6.4
|
|
|
3958f6fb
|
2019-10-29T22:36:50
|
|
Add in_range in hb_bytes_t to merge range_checker_t with it
|
|
|
7915c5d6
|
2019-10-29T22:55:34
|
|
[ci] Check the recipt introduce in 7152ac3
|
|
|
aa3831e2
|
2019-10-29T21:37:07
|
|
[glyf] Use contour_bounds_t::empty to simplify get_extents logic
|
|
|
d59ae583
|
2019-10-29T21:30:04
|
|
[glyf] Refactor contour_bounds_t use to make its fields protected
|
|
|
773ee804
|
2019-10-29T10:58:56
|
|
Merge pull request #2038 from ebraminio/glyf-contour
Remove contour_point_t so we won't need to zero its unused var
|
|
|
0efbda7a
|
2019-10-29T21:03:26
|
|
Remove contour_point_t so we won't need to zero its unused var
|
|
|
47b4074d
|
2019-10-29T03:16:34
|
|
Merge pull request #2034 from khaledhosny/cmap-notdef
[cmap] Check GID before adding ranges in format 4 & 12
|
|
dd8a8460
|
2019-10-29T09:20:41
|
|
Use proper y-scale factor for height value
|
|
|
7152ac3f
|
2019-10-28T17:12:55
|
|
Fix build
$ make CPPFLAGS="-DHB_TINY -DHB_NO_OT_FONT" libharfbuzz-subset.la
|
|
|
dd288840
|
2019-10-29T01:45:49
|
|
[cmap] Check GID before adding ranges in format 4 & 12
Fixes https://github.com/harfbuzz/harfbuzz/issues/2031
|
|
|
fd71c045
|
2019-10-28T17:06:53
|
|
Hopefully fix up previous commit and bots
|
|
|
e832dc4c
|
2019-10-28T17:02:40
|
|
[config] Make HB_NO_ERRNO work with systems defining errno as a macro
|
|
|
ce11df1b
|
2019-10-28T14:45:31
|
|
2.6.3
|
|
|
5a59de2b
|
2019-10-28T14:02:47
|
|
Actually add +hb_font_get_nominal_glyphs()
New API:
+hb_font_get_nominal_glyphs()
|
|
|
03028a5f
|
2019-10-28T13:46:56
|
|
Revert "Don't include codepoint 0 in the results of collect_unicodes."
This reverts commit 14ad96ffbf77c33d8d33d2686d17c2375381989e.
This was wrong. My bad!
https://github.com/harfbuzz/harfbuzz/issues/2031
|
|
|
2714dc31
|
2019-10-28T13:01:34
|
|
Merge pull request #2032 from googlefonts/collect_unicodes
Don't include codepoint 0 in the results of collect_unicodes.
|
|
14ad96ff
|
2019-10-28T12:56:04
|
|
Don't include codepoint 0 in the results of collect_unicodes.
It is always assumed to be the notdef glyph.
|
|
|
aca63902
|
2019-10-22T00:06:46
|
|
[gpos] Don't move pointer when match_glyph_data is not set
|
|
defe9b6d
|
2019-10-25T10:07:26
|
|
crash fix : Heap-buffer-overflow READ 2
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=18513
|
|
|
b2fcca6e
|
2019-10-24T15:15:26
|
|
fuzzer crash fix
https://oss-fuzz.com/testcase-detail/5643107869917184
|
|
|
9815ca03
|
2019-10-24T21:05:38
|
|
[ci] Use custom subset fuzzer timeout for failing sanitizer bots
|
|
|
b03b62cb
|
2019-10-23T12:18:21
|
|
[ci] Add HB_NO_MT on non HB_TINY no-build-system builds
|
|
|
5e489670
|
2019-10-22T15:03:47
|
|
Remove hb-warning.cc
Fixes https://github.com/harfbuzz/harfbuzz/issues/2005
|
|
|
05bcdb39
|
2019-10-21T16:10:06
|
|
Add a specialization of array_t:hash for hb_bytes_t and hb_ubytes_t.
|
|
|
95ab110c
|
2019-10-21T13:15:46
|
|
Optimize intersects_array to fix fuzzer timeout.
|
|
|
be5cdcdf
|
2019-10-22T11:55:04
|
|
Store the hash value of a key in item_t in hb_map to reduce the number of hash computations.
|
|
|
b33a0d62
|
2019-10-21T13:24:52
|
|
fuzzer crash fix: Null-dereference WRITE
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=18363
|
|
|
cf414e36
|
2019-09-27T09:55:17
|
|
Added profiling instructions to TESTING.md
|
|
|
1f0a9d9b
|
2019-09-17T11:10:08
|
|
[subset] GPOS Lookup Type 2: PairPos
|
|
|
e7667831
|
2019-10-21T22:17:06
|
|
[fuzz] Add https://crbug.com/oss-fuzz/17898 testcase
|
|
|
b66094ad
|
2019-09-30T16:19:18
|
|
[subset] GSUB Lookup Type 6: ChainContextSubst
|
|
|
eff91bbb
|
2019-10-16T16:58:57
|
|
Add missing pop() for the failure case of script subsetting.
|
|
|
eab191fd
|
2019-10-02T14:28:53
|
|
Rewrite hash and == function for hb_array_t as a plain loop.
Profling showed that the current implementation were accounting for nearly all processing time in some cases. These implementations look to be about 10x faster.
|
|
|
831daf4c
|
2019-10-01T17:46:27
|
|
Enforce HB_MAX_LANGSYS limit during layout subsetting.
|
|
|
9871461b
|
2019-10-01T16:29:24
|
|
Enforce HB_MAX_SCRIPTS during subsetting of the script table.
|
|
|
029775bc
|
2019-10-09T11:00:09
|
|
[buffer] Minor; improve HB_NO_BUFFER_MESSAGE
|
|
|
9a68e6b9
|
2019-10-16T13:28:15
|
|
[doc] minor
Addresses https://github.com/harfbuzz/harfbuzz/pull/2013/files#r335196908
|
|
|
b0b8551a
|
2019-10-12T21:12:19
|
|
[glyf] Use range_checker_t in composite_iter_t
|
|
|
06e35ce0
|
2019-10-12T15:16:31
|
|
[glyf] minor
|
|
|
d13bb4bd
|
2019-10-12T12:13:57
|
|
[glyf] other cleanups
|
|
|
e7aa8c5d
|
2019-10-12T11:25:32
|
|
[glyf] Introduce glyf::Glyph, a hb_bytes_t/GlyphHeader pair
|
|
|
546ffc9f
|
2019-10-12T09:51:57
|
|
[glyf] minor
|
|
|
4299ea26
|
2019-10-12T09:51:27
|
|
[glyf] Move get_contour_points to GlyphHeader
|
|
b918cd8c
|
2019-10-10T15:30:48
|
|
Document hb_feature_t.
This documents hb_feature_t. This is motivated mostly by the ambiguity
of the units for 'start' and 'end' (clusters) and whether they are
inclusive or exclusive. This also documents that for lookup type 3 the
value is the one based index into the alternates and that in a list of
features later feature values override previous feature values with the
same tag.
|
|
|
e637a4b3
|
2019-10-11T10:28:58
|
|
Merge pull request #2012 from harfbuzz/lepcha-oo
Reorder U+1C29 LEPCHA VOWEL SIGN OO
|
|
|
670fec23
|
2019-10-10T09:21:08
|
|
Minor, add a zero length tolerant memset, hb_memset
|