Garret Rieger
4cdd1b16c9
[subset] added todo in test-subset-glyf.
2018-02-12 13:36:28 -08:00
Garret Rieger
1714feef4f
[subset] re-enable test-subset-glyf. Refactor to use common functions in hb-subset-test.h
2018-02-12 13:35:17 -08:00
Garret Rieger
d8d8bd8405
[subset] Add some helper functions for writing tests around subsetting.
2018-02-12 13:34:08 -08:00
Behdad Esfahbod
d33ba16361
Fix more warnings
2018-02-11 18:26:41 -08:00
Behdad Esfahbod
41906cd5ef
[subset] Another fixup
...
I broke this in c31fcf4c58
2018-02-11 19:46:06 -06:00
Ebrahim Byagowi
ae6d0259b4
Install fonttools on two bots and enable cmake hb-subset test ( #777 )
2018-02-11 15:40:55 +03:30
Behdad Esfahbod
6c1848b1e3
Misc warning fixes
2018-02-10 15:52:35 -06:00
Behdad Esfahbod
4c5bdb4b39
[test] Disable test-subset-glyf for now
2018-02-10 15:40:54 -06:00
Behdad Esfahbod
43314ff4c0
More fixes
...
test-subset-glyf still failing. No idea why :(
2018-02-10 15:40:03 -06:00
Behdad Esfahbod
3346ba9cd5
[test/text-rendering-tests] Update from upstream
2018-02-09 11:05:17 -06:00
Garret Rieger
3e81832432
Disable subset tests on cmake for now.
2018-02-08 18:05:48 -06:00
Behdad Esfahbod
e081c9e8fc
Oops!
2018-02-08 17:44:24 -06:00
Behdad Esfahbod
1c49afe179
Whitespace
2018-02-08 17:38:58 -06:00
Behdad Esfahbod
9682ef135f
Minor fixups
2018-02-08 17:35:57 -06:00
Garret Rieger
0f3c756cbf
Add CMake config for building a separate harfbuzz-subset.
2018-02-08 17:31:11 -06:00
Garret Rieger
d4d120ad79
Skip subset to fonttools comparison test if TTX is not present.
2018-02-08 17:31:11 -06:00
Garret Rieger
29d915284e
Whitespace
2018-02-08 17:31:11 -06:00
Garret Rieger
f9420d9eff
In the hb-subset to fontTools comparison, use ttx to compare the fonts. This allows for some binary differences such as re-ordered tables.
2018-02-08 17:31:11 -06:00
Garret Rieger
f2ceb5ee4d
Comment out failing assert in test-subset-glyf for now. Should be re-enabled once hb_subset is writing out a new face.
2018-02-08 17:31:11 -06:00
Garret Rieger
89dbebd4ad
Add a basic test for glyf subsetting.
2018-02-08 17:31:11 -06:00
Behdad Esfahbod
1242ad8424
[test/text-rendering-tests] Update from upstream
2018-02-08 15:08:46 -06:00
Behdad Esfahbod
c479a59988
[subset] Assemble font
2018-02-07 21:26:04 -06:00
Behdad Esfahbod
f70100417c
[test] Minor
2018-02-07 13:58:23 -05:00
Behdad Esfahbod
af02812fc5
[subset] Change hb_subset API to remove hb_subset_face_t
2018-02-07 13:07:46 -05:00
Garret Rieger
7481f0ce21
Add codepoints to subset input.
2018-02-07 12:41:21 -05:00
Ebrahim Byagowi
8cb8bfd5b5
[test] Fix test-subset on old distros ( #758 )
...
https://developer.gnome.org/glib/stable/glib-Testing.html#g-assert-cmpmem
2018-02-06 20:40:42 +03:30
Behdad Esfahbod
b7b7f0ac59
[test] Minor
2018-02-05 19:40:31 -05:00
Garret Rieger
5f6eb1256f
Add libharfbuzz-subset.la to linking when building api tests.
2018-02-05 10:23:38 -08:00
Behdad Esfahbod
257022b178
Merge pull request #745 from googlefonts/master
...
Interface for hb_subset, skeleton for the hb-subset cli, and basic testing rigging.
2018-02-04 20:01:35 -05:00
Behdad Esfahbod
3615f344ec
[test/text-rendering-tests] Update from upstream
2018-02-04 17:21:27 -05:00
Garret Rieger
7b01761ade
Add CMakeLists.txt to dist files for subset test.
2018-02-04 12:57:34 -08:00
Garret Rieger
f83a43b561
Add CMake configuration for all of the new subsetting code.
2018-02-04 12:57:34 -08:00
Garret Rieger
06fe297f2a
Properly include subset test data files.
2018-02-04 12:57:34 -08:00
Garret Rieger
058b1260ad
Re-write hb-subset utility to use main-font-text driver.
2018-02-04 12:57:34 -08:00
Garret Rieger
2763a2c598
Include subset test files in distribution.
2018-02-04 12:57:34 -08:00
Garret Rieger
76b84c36b9
Whitespace
2018-02-04 12:57:34 -08:00
Garret Rieger
a2965f2ea5
Add a basic implementation of hb-subset to enable compilation of test-subset.
2018-02-04 12:57:33 -08:00
Garret Rieger
38af23b8df
Make the expected output for subsetting basics test be equal to the input file for now.
2018-02-04 12:57:33 -08:00
Garret Rieger
b59c08eb04
Add the beginning of a unit test for hb-subset
2018-02-04 12:57:33 -08:00
Rod Sheeter
e9d154ac8d
tweak test failure output. write to a temp file not stdout. test still fails because expected is not just an identical copy of input
2018-02-04 12:57:33 -08:00
Garret Rieger
76351518ca
Remove basic subset test from XFAIL
2018-02-04 12:57:33 -08:00
Garret Rieger
b029b7c19a
Whitespace
2018-02-04 12:57:33 -08:00
Garret Rieger
0853260e99
Configure automake to run the new subset tests.
2018-02-04 12:57:33 -08:00
Garret Rieger
5c63c37b2b
WIP test runner for subset tests.
2018-02-04 12:57:33 -08:00
Garret Rieger
cc46cd88a1
In generate-expected-outputs read the test definition with utf8 encoding.
2018-02-04 12:57:33 -08:00
Garret Rieger
4cdae914e2
Add basic directory structure for subsetter integration tests. Plus a utility for generating expected output files.
2018-02-04 12:57:33 -08:00
David Corbett
ccb0367dae
Let VS1 follow U+1031 MYANMAR VOWEL SIGN E
2018-02-02 12:41:07 -08:00
Behdad Esfahbod
585b576098
Fix distcheck
2018-01-30 11:52:11 -08:00
Behdad Esfahbod
e00743b47d
[test/text-rendering-tests] Update from upstream
2018-01-30 11:01:18 -08:00
Ebrahim Byagowi
e849b8a85b
[cmake] Always put test/api/CMakeLists.txt on distributions
2018-01-19 16:12:24 +03:30
David Corbett
9b693212a8
Update record-test.sh to the new directory layout ( #716 )
2018-01-19 01:04:13 +03:30
Behdad Esfahbod
60f84ef014
Disable MORX-13/14
...
Really, need better way to catch these :(.
2018-01-15 20:46:58 -05:00
Behdad Esfahbod
54a1e250a8
[test/text-rendering-tests] Update from upstream
2018-01-13 14:44:39 +00:00
Behdad Esfahbod
0d39ac224c
[test/text-rendering-tests] Update README
2018-01-11 22:47:08 +01:00
Behdad Esfahbod
17a66f133c
[test] Disable MORX-12
...
Ouch. I need a better workflow for this.
2018-01-11 22:46:19 +01:00
Behdad Esfahbod
1387fe8f9c
[test/text-rendering-tests] Update from upstream
2018-01-11 19:23:14 +01:00
Behdad Esfahbod
1eb98c9916
[test/text-rendering-tests] Actually disable MORX-11
2018-01-11 17:59:03 +01:00
Behdad Esfahbod
1d20d109a0
[test/text-rendering-tests] Update from upstream
2018-01-11 15:19:01 +01:00
Behdad Esfahbod
fc405b7ca7
[test/text-rendering-tests] Update from upstream
2018-01-11 15:15:36 +01:00
Ebrahim Byagowi
4177f646aa
[cmake] Update and unbreak ( #700 )
2018-01-11 15:43:23 +03:30
Behdad Esfahbod
d6076747ac
[test/text-rendering-tests] Add DISABLED list
2018-01-11 10:51:46 +01:00
Behdad Esfahbod
de02b5093c
[test/in-house] Add Makefile.sources
2018-01-11 10:48:42 +01:00
Behdad Esfahbod
7c32e01d04
Minor
2018-01-11 10:47:31 +01:00
Behdad Esfahbod
a5adc5aa84
[test/text-rendering-tests] Update from upstream
...
Also adds an update.sh script that does this automatically.
2018-01-11 10:43:30 +01:00
Behdad Esfahbod
c861daacc4
[test/text-rendering-tests] Add README and COPYING
2018-01-11 10:43:30 +01:00
Behdad Esfahbod
ea3e86c640
Minor
2018-01-11 10:43:30 +01:00
Ebrahim Byagowi
1461965c22
[test] Better tests output by converting stderr to stdout ( #696 )
...
https://gitlab.kitware.com/cmake/cmake/issues/17630
2018-01-11 08:48:22 +03:30
Ebrahim Byagowi
ebb9b7d633
[test] Print shaping output on CI fails ( #695 )
2018-01-11 08:15:06 +03:30
Ebrahim Byagowi
71e0ed9cba
[cmake] Add header existence tests ( #685 )
2018-01-10 15:45:12 +03:30
Ebrahim Byagowi
a073621b5d
[cmake] Fix tests against latest changes ( #690 )
2018-01-10 12:13:28 +03:30
Behdad Esfahbod
6db0b17c4c
[test] Move test handling to sundirectories
2018-01-10 05:40:43 +01:00
Behdad Esfahbod
4c982b4867
[test] Whitelist one more passing test
2018-01-10 05:26:55 +01:00
Behdad Esfahbod
0b28e1199d
[test] Whitelist one more passing test
2018-01-10 05:02:42 +01:00
Behdad Esfahbod
6dc6f0c9f8
[test] Whitelist one passing test
2018-01-10 04:15:57 +01:00
Behdad Esfahbod
ec939761e4
Minor
2018-01-10 03:56:43 +01:00
Behdad Esfahbod
c4eac531df
Minor
2018-01-10 03:51:09 +01:00
Behdad Esfahbod
285a0bd799
[test] Another try
...
Making cmake happy this time.
2018-01-10 03:49:32 +01:00
Behdad Esfahbod
95b32b0055
[test] Fix text disabling
...
Ouch, make!
2018-01-10 03:48:09 +01:00
Behdad Esfahbod
a938d105b8
[test/text-rendering-tests] Disable failing tests
2018-01-10 03:41:55 +01:00
Behdad Esfahbod
ed95d5e374
[test/text-rendering-tests] Disable failing tests
2018-01-10 03:41:02 +01:00
Behdad Esfahbod
3b1e97f364
Add HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES
...
New API:
HB_BUFFER_FLAG_REMOVE_DEFAULT_IGNORABLES
hb-shape / hb-view --remove-default-ignorables
One more text-rendering-tests test passing. Eleven failing.
2018-01-10 03:36:03 +01:00
Behdad Esfahbod
46e4ed552f
[test/text-rendering-tests] Hook up font variations
...
Six more passing. Twelve failing now.
2018-01-10 03:22:08 +01:00
Behdad Esfahbod
2b3f62fc4d
[test] Add tests from text-rendering-tests repo
...
https://github.com/unicode-org/text-rendering-tests
2018-01-10 03:16:56 +01:00
Behdad Esfahbod
3f29ea9144
[util] Add hb-shape --ned
...
For not displaying extra data: no clusters, no advance. Just data pertaining
to where glyphs end up on the screen.
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
7020130665
[test] Minor
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
316a28f8f8
[test] Add script for extracting tests from text-rendering-tests
...
Work in progress...
https://github.com/unicode-org/text-rendering-tests
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
6b4d63f295
[test] Minor
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
504f913654
[test] Simplify test runner
...
Always pass hb-shape as first argument to run-tests.py.
Also require automake 1.13.
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
141db8b7a3
[test] Minor
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
44c65eee28
[test] Reorganize test suite
...
In anticipation of importing more test suites.
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
9a3ad307ea
[test] Remove unused hb-diff-ngrams
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
b6b460024d
[test] Remove unused manifest stuff
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
ebbee7c247
[test] Move test files around
2018-01-10 02:50:49 +01:00
Behdad Esfahbod
ed2f458b14
Add test font for previous commit
2018-01-05 12:55:39 +00:00
Behdad Esfahbod
127bcc6ece
Merge remote-tracking branch 'fdo/master'
2018-01-05 12:50:22 +00:00
Behdad Esfahbod
8c0d1916a4
Improve CGJ skipping logic
...
Previously we made CGJ unskippable. Now, if CGJ did NOT prevent
any reordering, allow skipping over it. To make this work we
had to make changes to the Arabic mark reordering algorithm
implementation to renumber moved MCM marks. See comments.
Fixes https://github.com/harfbuzz/harfbuzz/issues/554
2018-01-05 12:48:19 +00:00
Ebrahim Byagowi
72bec1cd09
Improve tests ( #679 )
2018-01-05 12:42:20 +03:30
Ebrahim Byagowi
70807ad274
Fix make check issue on out-of-tree builds and test it ( #672 )
2018-01-02 23:52:12 +03:30
Ebrahim Byagowi
4e3cf91af2
[cmake] Run rest of the tests on Windows ( #668 )
2018-01-01 11:17:51 +03:30
Ebrahim Byagowi
2dbdeb1421
[cmake] Run fuzzing and shaping tests ( #666 )
2017-12-29 23:13:29 +03:30
Ebrahim Byagowi
50163976d2
[cmake] Enable cmake tests on distribution and test it ( #664 )
2017-12-25 01:55:48 +03:30
David Corbett
7bfdf1a941
Forbid Batak killers after vowel signs
2017-12-24 14:21:22 -08:00
ebraminio
20e69c950d
[test] print commented line, like the original script ( #649 )
2017-12-07 12:24:12 +03:30
ebraminio
5f061d27d5
Rewrite run-tests in python ( #647 )
2017-12-07 11:22:55 +03:30
Khaled Hosny
c860955148
Add API tests to cmake build
2017-12-06 15:25:10 +02:00
Khaled Hosny
f19126dd35
We are already inside “if HAVE_GLIB” check
2017-12-06 15:25:10 +02:00
Behdad Esfahbod
20b4672754
[set] Fix recent changes
...
Ouch!
Add tests.
2017-12-02 15:15:50 -08:00
Behdad Esfahbod
d9066afd6b
Add another test for previous commit
...
From https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=4273
2017-11-29 16:18:18 -08:00
Behdad Esfahbod
16fba5b215
[ot] Fix Extension type recurse to disallow recursing to another Extension
...
Particularly hazardous if the second layer mixes forward and backward lookups.
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=4336
2017-11-29 16:08:11 -08:00
David Corbett
9f259769a6
Categorize U+A8B4 as Consonant_Medial
2017-11-21 13:48:32 -08:00
Behdad Esfahbod
8d55340593
Fix fallback positioning of double diacritic marks
...
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=787284
2017-11-20 19:43:22 -08:00
David Corbett
87d74605f5
Categorize U+0C80 as Consonant_Placeholder
2017-11-20 13:10:28 -08:00
ebraminio
7c6937e7c7
Move all references of old url to the new address ( #622 )
2017-11-20 14:49:22 -05:00
Behdad Esfahbod
baf7779d2d
Limit how much recursion GSUB/GPOS does
...
This only counts recursions right now. Good start.
Hopefully...
Fixes https://github.com/behdad/harfbuzz/issues/429
2017-11-14 21:55:37 -08:00
Behdad Esfahbod
b9c96b5d7c
Add test for previous commit
...
At some point we should run tests under valgrind...
2017-11-14 15:56:21 -08:00
David Corbett
87f0ad183c
Allow U+20F0 in USE clusters
2017-11-14 15:04:10 -08:00
Behdad Esfahbod
8eed9cb11e
[kern] Fix invalid memory access if offset is zero
...
If offset is zero, we return Null() object. Wasn't prepared for that.
Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=4088
2017-11-09 18:09:40 -08:00
Khaled Hosny
b68fba4dc8
[test] Use glyph ids instead of forcing ft funcs
...
We were using ft funcs because these are CFF fonts, but we can avoid
glyph names instead and keep testing both ot and ft.
2017-11-09 18:01:15 -05:00
Khaled Hosny
2fcfbcf805
[test] Fix another instance of component rounding
...
See https://github.com/behdad/harfbuzz/pull/590#issuecomment-341194673
2017-11-09 18:01:15 -05:00
Behdad Esfahbod
822035ce73
[tests/arabic-stch.tests] Don't show flags
...
We don't need to test the actual flag values. For unsafe-to-break,
the --verify in run-tests.sh checks what we are interested in.
2017-11-08 17:03:55 -08:00
Behdad Esfahbod
53f0c27ddc
Minor
2017-11-08 17:02:54 -08:00
David Corbett
5a9cba9744
Mark non-initial left matras as unsafe to break
2017-11-08 19:12:10 -05:00
David Corbett
74e82adaea
Mark glyphs involved in 'stch' as unsafe to break
2017-11-06 15:06:23 -05:00
Khaled Hosny
0e63cbed9c
[test] Fix difference between ft and ot font funcs
...
See https://github.com/behdad/harfbuzz/pull/590#issuecomment-341194673 .
I simply removed the composite glyph and use the referenced simple
glyph directly.
2017-11-01 14:00:52 -06:00
Khaled Hosny
0b8f8dd869
Revert "[test] We don’t care about glyph positions here"
...
This reverts commit 70137e2b4b
.
2017-11-01 14:00:52 -06:00
Khaled Hosny
f1245013bf
[test] Run shaping tests with both ot and ft funcs
2017-11-01 10:33:10 -06:00
Khaled Hosny
70137e2b4b
[test] We don’t care about glyph positions here
...
Freetype and OT font functions give different positions for some glyphs
in this font (OT seems to be correct), but that is not what we are
interested in in this test.
See https://github.com/behdad/harfbuzz/pull/590#issuecomment-341045223 .
2017-11-01 10:33:10 -06:00
Khaled Hosny
9234b364b1
[test] Use ft functions for this test
...
https://github.com/behdad/harfbuzz/pull/590#issuecomment-340967271
2017-11-01 10:33:10 -06:00
Khaled Hosny
afbfb9279c
[test] Always use ft font funcs for these tests
...
These are CFF fonts and ot functions don’t support CFF glyph names yet.
The next commit will run all tests with ot functions.
2017-11-01 10:33:10 -06:00
David Corbett
d016c5bdb9
Categorize Grantha candrabindu for use in Tamil
2017-10-31 16:54:23 -06:00
David Corbett
729c9e1335
Include all relevant glyphs in the subsetted font
2017-10-26 13:49:56 -05:00
Behdad Esfahbod
cb6c6b0c42
Fix warning
...
49a41dc759 (commitcomment-25203194)
2017-10-26 11:43:33 -06:00
Jonathan Kew
7339926525
[test] Extend hb_set test for issue 579 (currently failing)
2017-10-26 12:33:05 -05:00
Jonathan Kew
3d6f7df42a
[test] Additional coverage in test_set_iter.
...
This adds a couple extra entries to the set used, such that iteration
would fail in 1.6.2 due to the bug in hb_set_t::page_t::next().
2017-10-26 12:30:52 -05:00
Khaled Hosny
9d4d2fb9af
[ot-font] Implement hb_ot_get_glyph_name
...
Turns out we already have support for “post” table, it just needed to be
activated and put in use.
2017-10-25 19:29:51 -04:00
Behdad Esfahbod
bf133497e5
Merge commit '49a41dc75931cc2d2f7e74b7801f8cc327039e8e'
2017-10-25 17:23:20 -06:00
Behdad Esfahbod
49a41dc759
[test] Add more hb-set tests
...
https://github.com/behdad/harfbuzz/issues/579
2017-10-25 16:17:45 -06:00
David Corbett
fd786c768f
Use `mktemp -d` instead of `mktemp --directory`
2017-10-25 16:53:37 -04:00
David Corbett
c4944920c0
Fix and run indic-script-extensions.tests
2017-10-25 16:53:16 -04:00
David Corbett
d8df714251
Fix code point iteration in narrow Python
2017-10-25 16:53:16 -04:00
Behdad Esfahbod
deed4a48d1
Faster hb_set_t
...
Fixes https://github.com/behdad/harfbuzz/pull/23
2017-10-15 16:45:33 -03:00
Behdad Esfahbod
81e2b9b8a9
Never skip over CGJ
...
We might want to tweak this some more. For now, never skipping over
it is better behavior than always skipping.
Part of https://github.com/behdad/harfbuzz/issues/554
2017-10-15 10:49:06 +02:00
Behdad Esfahbod
fa48ccbe12
[indic] Special-case Kannada Ra,H,ZWJ sequence
...
Fixes https://github.com/behdad/harfbuzz/issues/435
2017-10-12 14:07:37 +02:00
Behdad Esfahbod
ab8d70ec70
[arabic] Implement Unicode Arabic Mark Ordering Algorithm UTR#53
...
Fixes https://github.com/behdad/harfbuzz/issues/509
2017-10-04 14:47:10 +02:00
David Corbett
3ca9c92aa6
Test Consonant_With_Stacker in Kannada ( #530 )
2017-10-03 14:07:02 +02:00
Behdad Esfahbod
5d98de1f38
Tweak mark-base attachment
...
Apparently a base glyph can also become an attached component of a
ligature if the ligature-forming lookup used IgnoreBase. This was
being confused with a non-first component of a MultipleSubst and
hence not matched for mark-attachment. Tweak test to fix.
Fixes https://github.com/behdad/harfbuzz/issues/543
2017-10-03 13:49:40 +02:00
Behdad Esfahbod
09cbeb2246
Make bots happy
...
Fixes https://github.com/behdad/harfbuzz/issues/551
2017-10-03 13:22:07 +02:00
Behdad Esfahbod
1633513996
Add test for U+0A51
...
New Indic numbers are:
BENGALI: 353725 out of 354188 tests passed. 463 failed (0.130722%)
DEVANAGARI: 707307 out of 707394 tests passed. 87 failed (0.0122987%)
GUJARATI: 366355 out of 366457 tests passed. 102 failed (0.0278341%)
GURMUKHI: 60729 out of 60747 tests passed. 18 failed (0.0296311%)
KANNADA: 951201 out of 951913 tests passed. 712 failed (0.0747968%)
KHMER: 299071 out of 299124 tests passed. 53 failed (0.0177184%)
MALAYALAM: 1048136 out of 1048334 tests passed. 198 failed (0.0188871%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271662 out of 271847 tests passed. 185 failed (0.068053%)
TAMIL: 1091754 out of 1091754 tests passed. 0 failed (0%)
TELUGU: 970555 out of 970573 tests passed. 18 failed (0.00185457%)
Before 71c0a1429d
GURMUKHI used to be at 15,
because Uniscribe seems to allow this character standalone, but that looks
wrong.
2017-10-02 20:28:56 +02:00
Behdad Esfahbod
8b2c94c43f
Tweak ligature component matching for ligature formation
...
If two marks want to ligate and they belong to different components of the
same ligature glyph, and said ligature glyph is to be ignored according to
mark-filtering rules, then allow.
Example Burmese senquence:
U+1004,U+103A,U+1039,U+101B,U+103D,U+102D
Test font provided by Norbert Lindenberg.
Fixes https://github.com/behdad/harfbuzz/issues/545
2017-10-02 20:03:35 +02:00
Behdad Esfahbod
71c0a1429d
[indic] Fix shaping of U+0A51
...
Mark it as matra below to allow the sequence U+0A15, U+0A51, U+0A47.
Oh well...
Fixes https://github.com/behdad/harfbuzz/issues/524
2017-10-02 18:57:03 +02:00