organicmaps/harfbuzz
13
Fork 0
mirror of https://github.com/harfbuzz/harfbuzz.git synced 2025-04-05 05:25:05 +00:00
Code Issues 1 Projects Releases Wiki Activity Actions 9
16291 commits 24 branches 188 tags 126 MiB
Commit graph

16291 commits

This branch
This branch
All branches
Author SHA1 Message Date
Behdad Esfahbod
7297680fe6 [map] Keep one byte of the hash in a separate array 
Still no speedup. Just for the record.
 2023-06-01 13:28:54 -06:00
Behdad Esfahbod
41c3057c23 [map] Keep is_used in a separate bitmap as well 
Slows things down. Keeping for the record.
 2023-06-01 13:16:08 -06:00
Garret Rieger
f41c5ec961 [map] update map benchmark to also test lookups that hit. 2023-06-01 12:39:58 -06:00
Qunxin Liu
4acf6a82bf [instancer-solver] port solver fix from fonttools 
See https://github.com/fonttools/fonttools/issues/3139
 2023-05-30 18:58:22 +02:00
Garret Rieger
f3b4d35f36 [subset] Fix fuzzer crash. 
https://oss-fuzz.com/testcase-detail/6608005089853440
 2023-05-30 18:56:48 +02:00
Garret Rieger
ff326fbe8f [repacker] check the result of add_buffer() in other places where it's called. 2023-05-30 01:40:57 +02:00
Behdad Esfahbod
66a84355ab [hash] Use unsigned types 
Hopefully ubsan wouldn't complain about overflows now.
 2023-05-29 17:17:15 -06:00
Behdad Esfahbod
4a93576bca Revert "[map] Speedup for int types" 
This reverts commit c1b75f5ff0.

I was seeing very long chains (140) and a 17% slowdown in
BM_subset/subset_codepoints/RobotoFlex-Variable.ttf/512

Bummer.
 2023-05-28 13:24:02 -06:00
Behdad Esfahbod
c1b75f5ff0 [map] Speedup for int types 
Since our int hash is now good, we don't need the modulo prime
here, so forgo it.  Major speedup:

Comparing before to after
Benchmark                              Time             CPU      Time Old      Time New       CPU Old       CPU New
-------------------------------------------------------------------------------------------------------------------
BM_MapInsert/16                     -0.4136         -0.4137             6             4             6             4
BM_MapInsert/64                     -0.4442         -0.4441             6             3             6             3
BM_MapInsert/512                    -0.5382         -0.5383             8             4             8             4
BM_MapInsert/4096                   -0.4160         -0.4162             8             5             8             5
BM_MapInsert/32768                  -0.3256         -0.3258            12             8            12             8
BM_MapInsert/262144                 -0.1723         -0.1727            11            10            11             9
BM_MapInsert/1048576                -0.2310         -0.2309            28            22            28            22
BM_MapLookup/16                     -0.0247         -0.0247             3             3             3             3
BM_MapLookup/64                     -0.1039         -0.1038             3             3             3             3
BM_MapLookup/512                    -0.1076         -0.1079             4             3             4             3
BM_MapLookup/4096                   -0.3729         -0.3732             9             6             9             6
BM_MapLookup/32768                  +0.2467         +0.2468             9            12             9            12
BM_MapLookup/262144                 -0.1862         -0.1868            14            11            14            11
BM_MapLookup/1048576                +0.1159         +0.1160            15            17            15            17
OVERALL_GEOMEAN                     -0.2414         -0.2416             0             0             0             0
 2023-05-28 13:03:04 -06:00
Behdad Esfahbod
7f111787e4 [benchmark-map] Remove overhead 2023-05-27 12:41:33 -06:00
Behdad Esfahbod
5666807328 [benchmark-map] Improve Insert benchmark 
Previously it was enlarging the map depending on whatever
number of runs the benchmark-runner decided to run the loop.
That wasn't very useful...
 2023-05-27 12:37:36 -06:00
Behdad Esfahbod
00900f761f [hash] Enable better hash for integers 
https://github.com/harfbuzz/harfbuzz/pull/4228#issuecomment-1565079537
 2023-05-27 10:14:09 -06:00
Garret Rieger
20c564bc76 [repacker] Fix fuzzer memory leak. 
https://oss-fuzz.com/testcase-detail/6419865171525632
 2023-05-27 01:16:46 +02:00
Behdad Esfahbod
5abe713203 [map] Resize map on long chain probes 2023-05-27 00:49:20 +02:00
Garret Rieger
a652281ed6 [subset] Fix fuzzer timeout. 
Fixes https://oss-fuzz.com/testcase-detail/5979721620652032. Timeout was caused by degenerate map insert behaviour due to poor integer hash function. Presize the map to avoid it. Also fixes collect_mapping() for cmap format 13.
 2023-05-27 00:25:47 +02:00
Behdad Esfahbod
208c9490cb [VarComposite] Limit number of axes that are copied 
Bandaid for https://oss-fuzz.com/testcase-detail/4591122882887680
 2023-05-26 12:03:34 -06:00
Behdad Esfahbod
5872bdf64d [vector] Add copy-constructor for array_t's 2023-05-26 11:44:18 -06:00
Chun-wei Fan
42aba5ff30 harfbuzz-config.cmake.in: Support Windows usage 
Apply the appropriate library prefix/suffix by whether we are building
as a statib build in Meson, or by using
${CMAKE_[SHARED|IMPORT]_LIBRARY_PREFIX} and
${CMAKE_[SHARED|IMPORT]_LIRBARY_SUFFIX} as appropriate according to
the target platform for shared builds.
 2023-05-25 18:24:20 +03:00
Behdad Esfahbod
84354d3f23 [cff] When max-ops is reached return false 
Slight speedup too.
 2023-05-23 11:57:53 -06:00
Behdad Esfahbod
2655b7607c [map] Micro-optimize 
This shows a speedup.
 2023-05-23 10:57:58 -06:00
Behdad Esfahbod
12b0680e60 [map] Fix set() 
This was broken in 2133aa2407.

We have to keep probing over tombstones.
 2023-05-23 10:49:41 -06:00
Behdad Esfahbod
4056315c86 [font] Remove redundant avar mapping line 
The avar mapping happens in hb_ot_var_normalize_coords() call.
 2023-05-22 14:17:40 -06:00
Behdad Esfahbod
592a3d0e96
Merge pull request #4237 from harfbuzz/layout-font-extents 
[layout] Add hb_ot_layout_get_font_extents()
 2023-05-21 22:00:53 +02:00
Khaled Hosny
0d354e0442 [layout] Add tests for baseline 2 APIs 2023-05-21 19:01:04 +03:00
Behdad Esfahbod
e047b1abbc [layout] Improve docs 2023-05-20 10:27:31 -06:00
Behdad Esfahbod
132391ad62 [layout] Add test for font_extents 2023-05-19 16:10:21 -06:00
Behdad Esfahbod
6996e64f72 [BASE] Sanitize and get_min_max fixups 
Seems to work now.
 2023-05-19 14:22:13 -06:00
Behdad Esfahbod
02e0e9394b [layout] Fetch default font extents if localized extents not found 2023-05-19 13:18:06 -06:00
Behdad Esfahbod
8ccc1ff21e [layout] Mark language as nullable 
GI needs it apparently.
 2023-05-19 13:15:58 -06:00
Behdad Esfahbod
792ca22ca3 [layout] Document new API 2023-05-19 11:42:20 -06:00
Behdad Esfahbod
0894813a38 [layout] Add "2" version of BASE table API 
...that do the script/language resolution.

Part of https://github.com/harfbuzz/harfbuzz/pull/4237

Ref https://github.com/harfbuzz/harfbuzz/issues/3439
 2023-05-19 11:35:33 -06:00
Behdad Esfahbod
3e110c69c4 [layout] Add hb_ot_layout_get_font_extents() 
Untested.

Fixes https://github.com/harfbuzz/harfbuzz/issues/3290
 2023-05-19 11:06:26 -06:00
Simon Cozens
2f21dc23b5 [wasm] [docs] Improve build flags information 2023-05-18 07:14:28 +01:00
Simon Cozens
134cc8edfa [wasm] [docs] How to build 2023-05-17 16:58:49 +01:00
Behdad Esfahbod
5543d05885
[subset] Use vector instead of map for glyph bounds (#4232) 
* [subset] Use vector instead of map for glyph bounds

Is faster.

Part of https://github.com/harfbuzz/harfbuzz/issues/4231

* [subset] initialize bounds_vec value to 0xFFFFFFFF

Some non-EMPTY glyph might have 0 bounds width/height

---------

Co-authored-by: Qunxin Liu <qxliu@google.com>
 2023-05-17 09:29:44 -06:00
Behdad Esfahbod
ccfd7ef08c
Merge pull request #4131 from harfbuzz/wasm 
[wasm] WebAssembly shaper
 2023-05-17 16:00:33 +02:00
Behdad Esfahbod
6ed792ef4b [meson] Report WebAssembly build 2023-05-17 07:58:27 -06:00
Simon Cozens
56af88d048 [wasm] More thorough API docs 2023-05-17 14:32:17 +01:00
Behdad Esfahbod
c248fd9133 [hash] Comment 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
42f4dab8db [hash] Switch to Knuth multiplicative hash 
Wow, the old adhoc one was why the cmap test was timing out!
 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
0270e3e974 [map] Reinstate quadratic probing 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
19b628bdf0 Revert "[map] Implement Robinhood hashing" 
This reverts commit 7056178568.
 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
18c3ba5018 [map] Implement Robinhood hashing 
Numbers are not improved though. :(
 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
2133aa2407 [map] Inline code for set() 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
5bf5188ea2 [map] Simplify del() 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
2ffec3a6f4 [map] Shuffle fetch_item 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
1dc99128b9 [map] Separate has() code from set() code 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
2dd0803c85 [map] Downgrade from quadratic to linear probing 2023-05-16 10:12:02 +02:00
Behdad Esfahbod
498197671a [map] Write get() in terms of has() 2023-05-16 10:12:02 +02:00
Qunxin Liu
f37941bb52 [instancer] add tests for instancer-solver from python test suite 
Also fix bug in the solver code
 2023-05-15 14:41:57 -06:00