21346af01d
[layout-cache] Adjust cost-function for recent change
2022-06-07 14:41:39 -06:00
bfee6839b0
[layout-cache] Cache lookahead, not input, classdef in ChainContextFormat2
...
From the commit:
+ /* For ChainContextFormat2 we cache the LookaheadClassDef instead of InputClassDef.
+ * The reason is that most heavy fonts want to identify a glyph in context and apply
+ * a lookup to it. In this scenario, the length of the input sequence is one, whereas
+ * the lookahead / backtrack are typically longer. The one glyph in input sequence is
+ * looked-up below and no input glyph is looked up in individual rules, whereas the
+ * lookahead and backtrack glyphs are tried. Since we match lookahead before backtrack,
+ * we should cache lookahead. This decisions showed a 20% improvement in shaping of
+ * the Gulzar font.
https://github.com/harfbuzz/harfbuzz/pull/3636
2022-06-07 11:41:05 -06:00
39820af72f
[layout] Add HB_NO_OT_LAYOUT_LOOKUP_CACHE to disable caching lookups
...
Enabled when optimize-size profile is enabled.
2022-06-07 10:27:55 -06:00
c8fb048f79
[gsubgpos] Document caching
2022-06-07 09:20:27 -06:00
5963cf4469
[gsubgpos] Merge cache_enter and cache_leave entry points
...
Saves a pointer per subtable
2022-06-07 09:17:14 -06:00
d4c09e9a87
[gsubgpos] Remove apply_cached() entry point
...
Just use a bool to apply()
2022-06-07 09:17:14 -06:00
b96622d15c
[layout] Use a cache for main input ClassDef of (Chain)ContextLookupFormat2
...
This commit adds a per-lookup caching infrastructure to GSUB/GPOS, and
uses it to cache input ClassDef.get_class value for (Chain)ContextLookupFormat2.
For fonts heavy on use of heave class-based2 context matching, this shows
a good speedup. For NotoNastaliqUrdu for example, I observe 17% speedup.
Unfortunately not many other lookups can use a cache like this :(.
https://github.com/harfbuzz/harfbuzz/pull/3636
2022-06-07 09:14:38 -06:00
e6409d3905
Revert "[layout] Use a cache for main input ClassDef of (Chain)ContextLookups"
...
This reverts commit 57d1c08739
2022-06-05 06:57:37 -06:00
57d1c08739
[layout] Use a cache for main input ClassDef of (Chain)ContextLookups
2022-06-05 06:54:19 -06:00
f9b643f6b2
[layout] s/hb_get_subtables_context_t/hb_accelerate_subtables_context_t/g
2022-06-04 09:27:31 -06:00
a42c624fca
Convert one final use of hashmap to unique_ptr
2022-06-03 01:22:34 -06:00
f13a79548f
[subset] Convert another use of hashmap to unique_ptr
2022-06-03 01:17:20 -06:00
997d9cc466
[map] Make unique_ptr hashable
2022-06-02 18:47:42 -06:00
bca710e8ad
[gsubgpos] Use map has() instead of get() when appropriate
2022-06-02 12:06:25 -06:00
e9407a2bd2
Use shared_ptr<hb_set_t> in one place
...
See if valgrind is happy...
2022-06-02 12:00:43 -06:00
9a2a857043
[ot-shape] Don't verify buffer if shaping failed
...
Fixes all of fuzzing verify failures, which were result of buffer failure
on super-long results, which fails unsafe-to-break because shorter strings
don't fail shaping.
2022-05-31 06:35:42 -06:00
f4a8b70016
More member initialization
2022-05-30 05:30:37 -06:00
97aa1ce6ba
[gsubgpos] Move some member initialization
2022-05-29 10:56:41 -06:00
3eb7eff487
Remove varstore cache use if HB_NO_VAR
2022-05-21 15:25:53 -06:00
880f50f7e4
Refactor varstore cache
2022-05-21 15:12:14 -06:00
f2a2fb91a3
[GPOS] Cache VarStore region scalars
2022-05-21 15:12:14 -06:00
202e6c4699
[subset] Remove unnecessary test
2022-05-18 17:12:43 -06:00
4cfc2d668e
[subset] Use a std::move on set_t when feasible
2022-05-18 15:32:19 -06:00
6695bf0560
[gsubgpos] Remove wrong condition in Context application
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/3545
2022-04-22 13:48:41 -06:00
a85461b9b6
Add link to Context LookupFlag discussion
...
https://github.com/harfbuzz/harfbuzz/discussions/3538
2022-04-20 12:13:40 -06:00
044d7a06db
[indic-like] Add per-lookup per-syllable flag
...
This allows mix-and-matching per-syllable and other lookups.
In fact, removes the clear-syllables call completely.
Fixes https://github.com/harfbuzz/harfbuzz/issues/3513
2022-03-28 17:39:54 -06:00
e8f3397f4e
[matcher] Simplify syllable initialization
2022-03-28 13:23:37 -06:00
ccd9161bfd
[apply-lookup] Try to fix the logic for contextual lookups
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1611
Notably, this fixes Myles's AdditionFont:
https://litherum.blogspot.com/2019/03/addition-font.html
Test with AdditionFont, eg.:
$ util/hb-view AdditionFont.otf =1112112+1113134=
2022-03-25 15:14:40 -06:00
bf2a845a17
[ot-layout] Comment
2022-03-24 13:09:53 -06:00
4e2f409bce
[subset] Don't hold references to members of the active_glyph_stack.
...
These references may get invalidated after the vector for the stack is resized. Fixes: https://oss-fuzz.com/testcase-detail/5422577634377728
2022-01-31 22:58:53 +02:00
bc899650c7
[subset] Fix for issue #3397 .
...
cur_intersected_glyphs gets modified during recursion leading to incorrect filtering of sub tables in some cases. So don't use cur_intersected_glyphs. Instead just add an additional entry onto the parent_active_glyphs () stack.
Additionaly expands NotoNastaliqUrdu tests to include coverage of the issue from #3397 .
2022-01-28 14:20:33 -08:00
f3c1f4f0da
[subset] convert active_glyphs_stack to be a vector of hb_set_t instead of hb_set_t*.
2022-01-28 14:20:33 -08:00
60006d3687
[gsubgpos] Adjust chaining unsafe-to-concat application
...
Fixes three tests.
2022-01-22 11:53:18 -07:00
78481b32c0
[gsubgpos] Combine input/backtrack/lookahead unsafe-to-concat
...
I feel like this is correct logic. Still have to prove.
Errors unchanged at 10.
2022-01-22 11:53:18 -07:00
3122c2cdc4
[buffer] Add HB_GLYPH_FLAG_UNSAFE_TO_CONCAT
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/1463
2022-01-22 11:53:18 -07:00
ac1bb3e39e
[machinery] Move accelerators to constructor/destructor
2022-01-20 12:10:05 -07:00
e062376ef1
[machinery] Make accelerator lazy-loader call Xinit/Xfini
...
Instead of init/fini. To isolate those functions. To be turned into
constructor/destructors, ideally one per commit (after some SFINAE
foo.)
2022-01-19 17:09:34 -07:00
c8195dbc01
Merge pull request #3365 from harfbuzz/gdef-fix
...
GDEF fix
2022-01-19 06:56:05 -08:00
5a91db1135
[map] Remove constexpr invalid items
...
These were non-workable in the general case, eg std::string.
2022-01-13 12:44:48 -07:00
fa12f1ae25
Rename method
2022-01-12 15:08:34 -07:00
9408e0d6d1
Rename variable
2022-01-12 15:07:34 -07:00
96c969c820
Preserve glyph props
2022-01-12 15:06:40 -07:00
77507a1d8d
[subset] Don't double count visit lookups in closure_lookups
...
Each recursed lookup was getting counted once in closure_lookups() and then again in hb_closure_lookups_context_t::recurse.
2021-12-10 15:13:49 -07:00
d67025705f
[subset] reset the visited lookup count before each closure iteration.
2021-12-10 15:13:49 -07:00
706014f69d
[subset] (Chain)ContextSubst glyph_closure fix
...
- When pos_glyphs is empty, use current full glyphs set as input for
subsequent recursive closure process
- Also increase max_lookup_visit_count to 35000 cause a real font file hit
previous limit 20000 and some lookups are dropped unexpectedly
2021-12-10 11:08:36 -07:00
98fefd5ded
Merge pull request #3301 from harfbuzz/hashmap-classes
...
Hashmap classes
2021-11-19 13:34:53 -07:00
ca418cac74
[subset] keep features that have FeatureParams and the tag is "size"
2021-11-19 13:31:25 -07:00
e88fc41ef3
[subset] inputSequence could be empty, change the sanity check
2021-11-19 12:57:43 -07:00
394f772937
[map] Allow storing classes in the hashmap
...
Fixes https://github.com/harfbuzz/harfbuzz/issues/3293
The trick was to change the type of the invalid key/value to be non-class.
2021-11-19 11:49:23 -07:00
f51b48c8e7
[subset] Fix fuzzer found memory leak.
...
Happens because an insert into a map with an invalid key reports successful, but this causes the set being inserted to be lost.
2021-11-02 17:00:07 -07:00
540f19b6fe
[subset] fix bug in (Chain)ContextFormat2
...
Only keep rulesets for glyphs class numbers that survived in coverage
2021-11-02 16:05:55 -07:00
364b6b3989
[subset] Don't assume FeatureList is sorted
...
Though the spec said FeatureRecords are sorted alphabetically by feature
tag, there're font files with unsorted FeatureList. And harfbuzz is not
able to subset these files correctly because we use binary search in
finding featureRecords when collecting lookups. Also
find_duplicate_features needs to be updated to handle this.
2021-11-02 15:15:02 -07:00
6d555ce82e
[meta] Use std::forward instead of hb_forward()
2021-11-02 00:18:22 -06:00
36ed56b6e8
[subset] code cleanup
...
- rename copy() method if it takes extra parameter
- get rid of same code blocks
2021-10-09 13:37:20 -06:00
c2cc566c9d
[subset] Fix subset_offset_array adding unused space to serializer.
...
ArrayOf.serialize_append allocates space for the new item, but ArrayOf.pop() does not recover the allocated space. So in the case where the revert path was entered the extra space added by serialize_append gets left in the serialization buffer. This moves the snapshot to before ArrayOf.serialize_append is called so that revert cleans up the buffer extend.
2021-09-22 16:07:18 -06:00
8c5c81746d
[subset] fix find_duplicate_features
...
We should check each feature against all other features with the same
tag for duplicates.
2021-09-22 15:35:38 -06:00
37379f8f7d
[subset] fox for (Chain)ContextFormat3: subset lookupRecord
...
skip copying a lookupRecord if lookup referenced in the lookupRecord is
not retained after subset
2021-09-08 13:55:21 -06:00
fad452bffb
[set] Move main functionality into hb_bit_set_t
...
To add inversion on top in hb_invertible_set_t and use that as hb_set_t.
2021-08-17 10:47:18 -06:00
30579f5a37
[set] Document & use open-ended del_range()
2021-07-29 11:19:13 -06:00
2337f0d047
Internally use hb_malloc/.../hb_free instead of malloc/.../free
...
Redefining those stock names as macros was conflicting with gcc 10
headers.
Fixes https://github.com/harfbuzz/harfbuzz/issues/3044
2021-07-08 10:54:09 -07:00
085aa65e09
[subset] Convert offset serialization in gsubgpos and gdef to serialize_serialize().
2021-06-14 16:47:45 -07:00
bc33b87f5e
[set] Use references, not pointers, in internal C++ API
2021-06-09 11:54:12 -06:00
8b686afeb5
[subset] add unlikely().
2021-04-06 12:34:44 -06:00
64122b5a44
[subset] don't visit lookup if covered glyph set has failed.
...
If covered glyph set is in error then the same lookup can be recursed into repeatedly potentially causing a fuzzer timeout. Fixes: https://oss-fuzz.com/testcase-detail/5416421032067072 .
2021-04-06 12:34:44 -06:00
c35d786397
[subset] never drop the 'pref' feature.
...
Never ever drop feature 'pref', even if it's empty. Harfbuzz uses it to choose the shaper for Khmer.
2021-04-01 16:27:35 -06:00
b10741ca7f
[subset] when closing glyphs in context lookups don't use a current glyph set if a lookup is already applied.
2021-03-31 21:03:33 -06:00
5639e253f9
Add Array16Of<>
2021-03-31 16:04:43 -06:00
5efe360986
Rename (Unsized)OffsetListOf
2021-03-31 15:33:22 -06:00
6c4e0491d7
s/OffsetArrayOf/Array16OfOffset16To/g
2021-03-31 15:31:32 -06:00
9b4b58493b
Fixup for recent OffsetTo<> changes
2021-03-31 13:27:21 -06:00
ad28f973f3
Rename offset types to be explicit about their size
...
Add Offset16To<>, Offset24To<>, and Offset32To<> for most use-cases.
2021-03-31 13:00:07 -06:00
adca4ce071
[subset] fixes https://oss-fuzz.com/testcase-detail/6173520787800064 .
...
Caused by incorrect bounds check in glyph closure for context lookups.
2021-03-30 15:44:41 -06:00
9f77a0c1ff
[subset] use hb_set_clear to avoid calling clear() on null pool set.
2021-03-30 15:12:52 -06:00
752e393ad2
[subset] avoid calling clear on null pool set.
2021-03-30 15:12:52 -06:00
9ed5f04a70
[subset] Simplify recent out-of-memory fixes
...
By checking return status of map->set().
2021-03-29 18:14:30 -06:00
52df6b9fd8
[subset] check for set insertion success.
2021-03-29 17:41:07 -06:00
c6adb90a27
[subset] fix nullptr deref.
2021-03-29 17:41:07 -06:00
56ca435787
[subset] fix for collect_features and remove_redundant_lamngsys
...
previously remove_redundant_sys () is missing in harfbuzz, after
redundant langsys removal, some features are removed as well in
prune_features() in fonttools. This change is trying to get the same
result between harfbuzz and fonttools.
2021-03-17 14:24:04 -07:00
0e1c0fa404
[subset] optimize glyph closure method: step 5
...
add testcase and some fixes
2021-03-17 15:22:07 -06:00
b8a58a0c0b
[subset] optimize glyph closure method: step 4
...
optimize recurse_lookups in Context/ChainContext
glyph closure, only the glyphs that the parent lookup
can apply the recursion to can participate in recursing
the lookup.
2021-03-17 15:22:07 -06:00
b4fc593c3c
[subset] optimize glyph closure method: step1
...
Previous GSUB glyph closure is done by recursively visiting
all reachable lookup tables that apply to any glyphs in the
current/input glyph set, but actually only the glyphs that
the parent lookup can apply the recursion to can participate
in recursing the lookup. This is step 1 for glyph closure
optimization:
1. Add stack of currently active glyph set pointers into hb_closure_context_t
2. Update closure() method in simple GSUB tables to use
this stack in order not to change current glyph set at different stages
of recursion
3. Add function may_have_non_1to1() to GSUB tables
2021-03-17 15:22:07 -06:00
8450f43ae1
[buffer] HB_NODISCARD next_glyph()
2021-03-15 16:01:35 -06:00
34a1204f10
[buffer] HB_NODISCARD output_glyph()
...
Also, generalize and use replace_glyphs() in morx where output_glyph() was used
in a loop.
2021-03-15 16:01:35 -06:00
cac6c86d2f
[buffer] HB_NODISCARD move_to()
2021-03-15 13:56:46 -06:00
3f1998a065
[buffer] HB_NODISCARD replace_glyph()
2021-03-15 13:33:44 -06:00
2902529b92
[subset] Fix HB_TINY build
...
Fixes https://github.com/harfbuzz/harfbuzzjs/issues/34#issuecomment-789247723
2021-03-02 15:05:39 -07:00
83b66bfb66
Another try to fix narrowing error
...
../src/hb-ot-layout-gsubgpos.hh: In instantiation of ‘void OT::ChainRule::serialize_array(hb_serialize_context_t*, OT::HBUINT16, Iterator) const [with Iterator = hb_map_iter_t<hb_array_t<const OT::IntType<short unsigned int> >, const hb_map_t*&, (hb_function_sortedness_t)0, 0>; typename hb_enable_if<hb_is_iterator_of<Lhs, typename Lhs::item_t>::value>::type* <anonymous> = 0; OT::HBUINT16 = OT::IntType<short unsigned int>]’:
../src/hb-ot-layout-gsubgpos.hh:2341:30: required from here
../src/hb-ot-layout-gsubgpos.hh:2326:15: error: narrowing conversion of ‘(unsigned int)g’ from ‘unsigned int’ to ‘short unsigned int’ inside { } [-Werror=narrowing]
c->copy (HBUINT16 {g});
~~~~~~~~^~~~~~~~~~~~~~
https://github.com/harfbuzz/harfbuzz/pull/2875
2021-02-23 13:04:38 -07:00
f4f35a4d5f
[constexpr] Use initializer instead of assignment
2021-02-22 22:32:42 -07:00
6d94194497
Use auto in range-for-loop more
2021-02-19 17:10:06 -07:00
08a4997fde
[subset] Add subset support for Extension lookups (GPOS 9, GSUB 7).
2021-02-11 13:41:31 -07:00
dfa9d7acdc
[minor] Use serializer->propagate_error() to simplify code
2021-02-11 11:19:51 -07:00
d7e2a51de2
[minor] Add unlikely() when checking for error
2021-02-11 10:55:03 -07:00
1da75afbe8
[minor] Add unlikely()
2021-02-10 00:03:41 -07:00
6e1afac64b
[minor] Rename internal variable
...
To address review comment:
https://github.com/harfbuzz/harfbuzz/pull/2699#discussion_r573370781
2021-02-09 18:48:46 -07:00
6a3fd94f3b
Merge pull request #2699 from googlefonts/gpos_8
...
[subset] Add a more complex layout subsetting test case and fix the issues it exposed.
2021-02-09 20:49:04 -05:00
8f47dd5779
[subset] don't set lookup visited in closure_lookups_context_t::recurse.
...
- Lookup::closure_lookups also checks if the lookups visited and sets the lookup to visited. If we set visited in 'recurse' then Lookup::closure_lookups will fail to recurse into the children of the lookup.
- Also when copying ChainRule's skip LookupRecord's that point to lookups which aren't retained. This matches FontTool's behaviour.
2021-02-09 20:34:52 -05:00
53806e5b83
Tiny improvement on previous commit
...
Functionally the same.
2020-11-25 11:51:37 -07:00
a5f6f869e8
Drop layout table if layout lookup accelerator allocation failed
...
If the table is kept, when the table is present but only the accelerator
allocation failed, then we run into nullptr dereferences downstream in
hb_ot_map_t::apply when trying to access proxy.accels[i].
To fix this, instead of only setting lookup_count of accelerator_t
itself, drop the whole table to avoid hb-ot-map construction assuming
that the lookups will work correctly despite accelerator allocation
having failed.
Fixes:
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=24490
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=24490
2020-11-25 10:18:04 -08:00
718bf5aab3
[subset] only keep features reachable from script in the final subset.
...
Matches fontTools behaviour.
2020-09-29 13:16:01 -07:00
e583505334
[subset] Use plan->glyphset_gsub instead of plan->glyphset for GSUB/GPOS
...
This matches fontTools behaviour. glyphset_gsub does not contain gids added from closing over composite glyphs in glyf, since these cannot particpate in GSUB/GPOS processing.
2020-09-29 11:16:15 -07:00
Behdad Esfahbod
Garret Rieger
Qunxin Liu
Dominik Röttsches