Commit Graph

2648 Commits

Author SHA1 Message Date
Behdad Esfahbod fabd3113a9 [OT] Port Arabic fallback shaping to synthetic GSUB
All of init/medi/fina/isol and rlig implemented.

Let there be dragons... ⻯
2012-09-06 00:51:44 -04:00
Behdad Esfahbod f0b8ed1b6d [Indic] Allow "H,ZWJ,M"
Uniscribe accepts a Halant,ZWJ before matras.  Allow that.

BENGALI down from 295 to 291
DEVANAGARI down from 69 to 57
GUJARATI down from 19 to 17
KANNADA down from 871 to 867
MALAYALAM down from 340 to 337
TELUGU down from 20 to 16

Currently at:

BENGALI: 353897 out of 354188 tests passed. 291 failed (0.0821598%)
DEVANAGARI: 707337 out of 707394 tests passed. 57 failed (0.00805774%)
GUJARATI: 366440 out of 366457 tests passed. 17 failed (0.00463902%)
GURMUKHI: 60704 out of 60747 tests passed. 43 failed (0.0707854%)
KANNADA: 951046 out of 951913 tests passed. 867 failed (0.0910798%)
KHMER: 299077 out of 299124 tests passed. 47 failed (0.0157125%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1047997 out of 1048334 tests passed. 337 failed (0.0321462%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271666 out of 271847 tests passed. 181 failed (0.0665816%)
TAMIL: 1091754 out of 1091754 tests passed. 0 failed (0%)
TELUGU: 970557 out of 970573 tests passed. 16 failed (0.00164851%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2012-09-05 17:41:08 -04:00
Behdad Esfahbod 4ed717ef61 [Indic] Relax grammar
Now that we insert dotted-circle, tests break more easily when our indic
machine breaks.

In particular, a few Devanagari tests were having sequences like
"C,H,ZWJ,N", and because of the ZWJ the Nukta does NOT get reordered to
before the Halant as the grammar used to expect...  Fixup.

Another case is as simple as "C,ZWJ,SM".

Fixes 10 out of 79 failures:

DEVANAGARI: 707325 out of 707394 tests passed. 69 failed (0.00975411%)
2012-09-05 17:21:17 -04:00
Behdad Esfahbod aa7141efe4 [Indic] Fix Khmer syllable-final coeng-consonant
Brings down Khmer failures from 162 to 47.

KHMER: 299077 out of 299124 tests passed. 47 failed (0.0157125%)

Also rebaselined some of the test files that had only-inherited lines.
Removing those, the stats are:

BENGALI: 353893 out of 354188 tests passed. 295 failed (0.0832891%)
DEVANAGARI: 707315 out of 707394 tests passed. 79 failed (0.0111678%)
GUJARATI: 366438 out of 366457 tests passed. 19 failed (0.00518478%)
GURMUKHI: 60704 out of 60747 tests passed. 43 failed (0.0707854%)
KANNADA: 951042 out of 951913 tests passed. 871 failed (0.0915%)
KHMER: 299077 out of 299124 tests passed. 47 failed (0.0157125%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1047994 out of 1048334 tests passed. 340 failed (0.0324324%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271666 out of 271847 tests passed. 181 failed (0.0665816%)
TAMIL: 1091754 out of 1091754 tests passed. 0 failed (0%)
TELUGU: 970553 out of 970573 tests passed. 20 failed (0.00206064%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)

Still some regressions, but some of the more egregious cases are
addressed.
2012-09-05 17:14:52 -04:00
Behdad Esfahbod 27bd55bd2c [Indic] Tamil does not have half-forms either
The Win7 Tamil font does not realy on this behavior, but the WinXP
version does.  Handle Tamil like Malayalam: Matras always move to
before base.

WinXP Tamil failures went down from 168964 (15.4752%) to 167
(0.0152953%) (two orders of magnitude reduction!).

Included in this is a minor fixup that actually fixed a few tests
with non-Tamil too.  Numbers at:

BENGALI: 353997 out of 354285 tests passed. 288 failed (0.0812905%)
DEVANAGARI: 707339 out of 707394 tests passed. 55 failed (0.00777502%)
GUJARATI: 366489 out of 366506 tests passed. 17 failed (0.0046384%)
GURMUKHI: 60769 out of 60809 tests passed. 40 failed (0.0657797%)
KANNADA: 951086 out of 951913 tests passed. 827 failed (0.0868777%)
KHMER: 299106 out of 299124 tests passed. 18 failed (0.00601757%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1048104 out of 1048416 tests passed. 312 failed (0.0297592%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271747 out of 271847 tests passed. 100 failed (0.0367854%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970558 out of 970573 tests passed. 15 failed (0.00154548%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2012-09-05 15:22:02 -04:00
Behdad Esfahbod 87b75d0a4a [OT] Allow adding features with fallback implementation 2012-09-04 23:06:38 -04:00
Behdad Esfahbod 1d3947a6bd Minor 2012-09-04 22:42:17 -04:00
Behdad Esfahbod b3b89b6658 [OT] Add SubstLookup serialize API 2012-09-04 21:28:33 -04:00
Behdad Esfahbod 715e03bc21 Minor 2012-09-04 20:10:17 -04:00
Behdad Esfahbod 652d1e0d64 [OT] Start adding Lookup-level serialize API 2012-09-04 20:00:44 -04:00
Behdad Esfahbod a930c68e9c [OT] More serialize. Implements all basic GSUB subtables 2012-09-04 19:16:09 -04:00
Behdad Esfahbod 1b38b4e817 Minor 2012-09-04 18:17:21 -04:00
Behdad Esfahbod 2bd9fe3598 Refactor 2012-09-04 15:15:19 -04:00
Behdad Esfahbod a5ddd9e31c [OT] Really fix possible NULL dereference this time 2012-09-04 14:55:00 -04:00
Behdad Esfahbod 2941683358 [OT] Implement serialize() for AlternateSubst 2012-09-03 23:31:14 -04:00
Behdad Esfahbod 1f07e3382a [OT] Implement serialize() for MultiSubst 2012-09-03 23:28:34 -04:00
Behdad Esfahbod 4912030dfb Minor 2012-09-03 21:00:48 -04:00
Behdad Esfahbod f8fa2b5cf6 Fix possible NULL dereference
As reported by Kenichi Ishibashi.
2012-09-03 20:19:46 -04:00
Behdad Esfahbod 4b312fb288 [OT] Remove serialize alignment
Will reintroduce in a different way when we actually need it.
2012-09-01 21:56:06 -04:00
Behdad Esfahbod c61be03d6d [OT] A bit more serialize 2012-09-01 21:49:44 -04:00
Behdad Esfahbod abcc5ac1fd [OT] Improve serialize syntax
For some definition of improvement...
2012-09-01 21:30:17 -04:00
Behdad Esfahbod bc5be24014 [OT] Restart work on serialize() 2012-09-01 21:25:20 -04:00
Behdad Esfahbod 6912e476dd [OT] Insert dotted-circle for run-initial marks
Unfortunately if the font has GPOS and 'mark' feature does
not position mark on dotted-circle, our inserted dotted-circle
will not get the mark repositioned to itself.  Uniscribe cheats
here.

If there is no GPOS however, the fallback positioning kicks in
and sorts this out.

I'm not willing to address the first case.
2012-09-01 20:38:45 -04:00
Behdad Esfahbod 1d581ec384 [OT] Fallback-position ccc=0 Thai / Lao marks
Not perfect, but so is fallback positioning in 2012...
2012-09-01 20:06:26 -04:00
Behdad Esfahbod 3992b5ec4c Move code around 2012-09-01 19:20:41 -04:00
Behdad Esfahbod b85800f9de [Indic] Implement dotted-circle insertion for broken clusters
No panic, we reeally insert dotted circle when it's absolutely broken.

Fixes most of the dotted-circle cases against Uniscribe. (for Devanagari
fixes 80% of them, for Khmer 70%; the rest look like Uniscribe being
really bogus...)

I had to make a decision.  Apparently Uniscribe adds one dotted circle
to each broken character.  I tried that, but that goes wrong easily with
split matras.  So I made it add only one dotted circle to an entire
broken syllable tail.  As in: "if there was a dotted circle here, this
would have formed a correct cluster."  That works better for split
stuff, and I like it more.
2012-08-31 19:18:20 -04:00
Behdad Esfahbod 327d14ef18 [Indic] Start adding dotted-circle instrastructure 2012-08-31 16:49:34 -04:00
Behdad Esfahbod 1be368e96f Minor 2012-08-31 16:29:17 -04:00
Behdad Esfahbod 784f29d061 Minor 2012-08-31 14:06:26 -04:00
Behdad Esfahbod 5a7f18767a [OT] Better fallback-position Thai / Lao ccc!=0 marks 2012-08-30 22:53:29 -04:00
Behdad Esfahbod 9f2348de58 [OT] Add serialize() for Coverage 2012-08-29 21:08:59 -04:00
Behdad Esfahbod e901b954c6 [OT] Start adding serialize() API 2012-08-29 20:26:08 -04:00
Behdad Esfahbod 965c280de0 Add HB_BUFFER_ASSERT_VAR
To be used in places we access buffer vars...
2012-08-29 14:02:37 -04:00
Behdad Esfahbod 0ccf9b6473 Move code around 2012-08-29 14:02:37 -04:00
Behdad Esfahbod 2fcbbdb41a Port Arabic fallback ligating to share code with GSUB
This will eventually allow us to skip marks, as well as (fallback)
attach marks to ligature components of fallback-shaped Arabic.
That would be pretty cool.  I kludged GDEF props in, so mark-skipping
works, but the produced ligature id/components will be cleared later
by substitute_start() et al.

Perhaps using a synthetic table for Arabic fallback shaping was a better
idea.  The current approach has way too many layering violations...
2012-08-29 14:01:22 -04:00
Behdad Esfahbod 5e399a8a45 Minor 2012-08-29 10:40:49 -04:00
Behdad Esfahbod a177d027d1 [GSUB] Move ligation logic over 2012-08-28 23:18:22 -04:00
Behdad Esfahbod 191fa885d9 [GSUB] Merge Ligature and context input matching
Looks better now...
2012-08-28 22:58:55 -04:00
Behdad Esfahbod 93814ca7dc Start converging Ligature and match_input 2012-08-28 22:39:10 -04:00
Behdad Esfahbod 2eef71737e [hb-icu-le] Add visbility 2012-08-28 19:16:38 -04:00
Behdad Esfahbod d59e28e492 Minor 2012-08-28 19:08:36 -04:00
Behdad Esfahbod af169d2813 Minor 2012-08-28 19:08:22 -04:00
Behdad Esfahbod 52ff2681d8 Use VisualStudio-style atomic intrinsics on mingw32 2012-08-28 18:03:35 -04:00
Behdad Esfahbod 7c8e844d92 Use namespace for OpenType tables
Avoids USHORT, SHORT, ULONG, LONG clashes with Windows API.
2012-08-28 17:57:49 -04:00
Behdad Esfahbod dc5df5af6b Revert "Minor"
This reverts commit 3e0a03978b.

I know remember why that line is there :).
2012-08-28 16:31:23 -04:00
Behdad Esfahbod 3e0a03978b Minor 2012-08-27 17:10:02 -04:00
Behdad Esfahbod 667218a5b1 Minor 2012-08-27 17:00:44 -04:00
Behdad Esfahbod 30dd62251f Only fallback-position glyphs if we have the ccc
Previously, ccc=0 Thai / Lao marks were being
mispositioned.  Don't touch them.
2012-08-27 16:54:34 -04:00
Behdad Esfahbod e1ba62811a Center unknown marks horizontally 2012-08-27 16:28:05 -04:00
Behdad Esfahbod 23b0e9d7dc [Indic] Fix switch
D'oh.  Was working by pure chance :)).
2012-08-26 14:30:38 -04:00
Behdad Esfahbod 56e878ab87 [graphite2] Cleanup scratch buffer allocation 2012-08-24 00:41:51 -04:00
Behdad Esfahbod 2f7586c622 [icu-le] Implement icu layout engine shaper 2012-08-24 00:00:33 -04:00
Behdad Esfahbod ba7f6c3797 [icu-le] Hook up to hb_face_t 2012-08-24 00:00:33 -04:00
Behdad Esfahbod e96bb36995 [icu-le] Actually use the FontTableCache 2012-08-24 00:00:33 -04:00
Behdad Esfahbod 7d242364ea [icu-le] Start adding a icu-layout-engine backend
Import PortableFontInstance and add shaper stub.
2012-08-24 00:00:29 -04:00
Behdad Esfahbod b5584ee4be [Indic] For old-spec, match non-zero context
Fixes consonant-position with old-spec Malayalam.  Uniscribe seem to be
doing this.  Fixes below-base La (eg. Pa,H,La) with AnjaliNewLipi.ttf.
Doesn't regress new-spec or other scripts.
2012-08-23 16:26:07 -04:00
Behdad Esfahbod d9b204d3d2 [GSUB] Allow non-zero-context matching in would_apply()
To be used in the next patch.
2012-08-23 16:22:28 -04:00
Behdad Esfahbod 1f2bb172fe Revert "[Indic/GSUB] Ignore context when matching would_apply()"
This reverts commit 24dd4e5674.

Oops.  My bad.  The change _regressed_ Malayalam test suite, not
improved it.  I'll redo it, differentiating between old-spec and
new-spec cases.
2012-08-23 16:10:37 -04:00
Behdad Esfahbod 24dd4e5674 [Indic/GSUB] Ignore context when matching would_apply()
The MS Indic specs say "...all classifications are determined ... using
context-free substitutions."  However, testing shows that MS's Malayalam
shapers (both old and new), "match" even if there is no zero-context rule.
We follow.

Fixes below-base La (eg. Pa,H,La) with AnjaliNewLipi.ttf (old spec).
Moreover, test suite Malayalam failures are down to 312 from 875!  No
change in other scripts.

Current numbers:

BENGALI: 353996 out of 354285 tests passed. 289 failed (0.0815727%)
DEVANAGARI: 707339 out of 707394 tests passed. 55 failed (0.00777502%)
GUJARATI: 366489 out of 366506 tests passed. 17 failed (0.0046384%)
GURMUKHI: 60769 out of 60809 tests passed. 40 failed (0.0657797%)
KANNADA: 951086 out of 951913 tests passed. 827 failed (0.0868777%)
KHMER: 299106 out of 299124 tests passed. 18 failed (0.00601757%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1047541 out of 1048416 tests passed. 875 failed (0.0834592%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271726 out of 271847 tests passed. 121 failed (0.0445103%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970558 out of 970573 tests passed. 15 failed (0.00154548%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2012-08-23 15:47:10 -04:00
Behdad Esfahbod 6732d62e78 [Indic] Implement pre-base reordering Ra for old-spec Malayalam
Fixes Pa,H,Ra sequence with AnjaliNewLipi.ttf.
2012-08-23 15:32:12 -04:00
Behdad Esfahbod 80cd92326f [Indic] Only apply basic features per-syllable
Free up syllables and let features work across syllables for the
presentation forms features and GPOS.

Fixed:
- 1 GURMUKHI test (remains 40)
- 12 KHMER tests (remains 18)
- 11 SINHALA tests (remains 121)

Regresses:
- 5 MALAYALAM tests (up to 312)

Current numbers:

BENGALI: 353996 out of 354285 tests passed. 289 failed (0.0815727%)
DEVANAGARI: 707339 out of 707394 tests passed. 55 failed (0.00777502%)
GUJARATI: 366489 out of 366506 tests passed. 17 failed (0.0046384%)
GURMUKHI: 60769 out of 60809 tests passed. 40 failed (0.0657797%)
KANNADA: 951086 out of 951913 tests passed. 827 failed (0.0868777%)
KHMER: 299106 out of 299124 tests passed. 18 failed (0.00601757%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1048104 out of 1048416 tests passed. 312 failed (0.0297592%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271726 out of 271847 tests passed. 121 failed (0.0445103%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970558 out of 970573 tests passed. 15 failed (0.00154548%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2012-08-23 12:06:14 -04:00
Behdad Esfahbod df5d5c68f3 Whitespace 2012-08-23 09:33:30 -04:00
Behdad Esfahbod 2f1747ed7d Add comment 2012-08-16 11:46:46 -04:00
Behdad Esfahbod bd08d5d126 [OT] Fix Arabic shaper OOB access
https://bugzilla.mozilla.org/show_bug.cgi?id=782908
2012-08-16 11:35:50 -04:00
Behdad Esfahbod daf0731865 [ICU] Fix includes
As reported by Steven Loomis, including uversion.h works everywhere.
2012-08-16 07:32:59 -04:00
Behdad Esfahbod a67ba9c0fe Whitespace 2012-08-15 18:52:17 -04:00
Behdad Esfahbod 45c1383cc7 Minor 2012-08-14 09:33:18 -04:00
Behdad Esfahbod 4ac4c6f2e1 Fix ICU build with older ICUs 2012-08-13 10:52:52 -04:00
Behdad Esfahbod d5045a5f40 [ICU] Use new normalizer2 compose/decompose API
It's considerably faster than the fallback implementation we had
previously!
2012-08-11 21:27:15 -04:00
Behdad Esfahbod 9f9f04c222 [OT] Unbreak Thai shaping and fallback Arabic shaping
The merger of normalizer and glyph-mapping broke shapers that
modified text stream.  Unbreak them by adding a new preprocess_text
shaping stage that happens before normalizing/cmap and disallow
setup_mask modification of actual text.
2012-08-11 18:34:13 -04:00
Behdad Esfahbod e9f28a38f5 [OT] Add shape_plan to Arabic shaper 2012-08-11 18:20:54 -04:00
Behdad Esfahbod daf13afb08 [OT] Implement fallback mark positioning for "double" combining marks 2012-08-10 16:38:44 -04:00
Behdad Esfahbod d345313104 [OT] Fix fallback mark positioning with left-to-right text
Ouch!
2012-08-10 16:34:04 -04:00
Behdad Esfahbod f4cb476298 [OT] Slightly adjust normalizer
The change is very subtle.  If we have a single-char cluster that
decomposes to three or more characters, then try recomposition, in
case the farther mark may compose with the base.
2012-08-10 03:51:44 -04:00
Behdad Esfahbod 07d6828063 Minor 2012-08-10 03:28:50 -04:00
Behdad Esfahbod b00321ea78 [OT] Avoid calling get_glyph() twice
Essentially move the glyph mapping to normalization process.
The effect on Devanagari is small (but observable).  Should be more
observable in simple text, like ASCII.
2012-08-09 22:33:32 -04:00
Behdad Esfahbod 12c0875eaf [OT] Remove redundant check 2012-08-09 22:02:54 -04:00
Behdad Esfahbod 5c60b70c89 [OT] More code shuffling around
Preparing for merging map_glyphs() and normalize().
2012-08-09 21:58:07 -04:00
Behdad Esfahbod cd0c6e148f Shuffle buffer variable allocations around
To room for more allocations, coming.
2012-08-09 21:48:55 -04:00
Behdad Esfahbod 8d1eef3f32 Minor 2012-08-09 21:35:47 -04:00
Behdad Esfahbod 56c9e7c004 Fill out combining class resetting for fallback shaping Thai/Lao/Tibetan 2012-08-09 21:14:23 -04:00
Behdad Esfahbod a321e1d51e Revert "Reject lookups with no subTable"
This reverts commit 30ec9002d8.

See previous commit.
2012-08-09 18:30:34 -04:00
Behdad Esfahbod 2eaf482b37 Revert "[GSUB/GPOS] Reject Context/ChainContext lookups with zero input"
This reverts commit 0981068b75.

I was confused.  Even if we access coverage[0] unconditionally, we don't
need bound checks since the array machinary already handles that.
2012-08-09 18:30:05 -04:00
Behdad Esfahbod a02d86484b Add check-exported-symbols.sh
And misc linking fixes.
2012-08-08 18:04:29 -04:00
Behdad Esfahbod 4c8ac4f47e Misc minor fixes 2012-08-08 17:44:19 -04:00
Behdad Esfahbod 560d68af81 Use a export-file for Windows builds
Apparently even that doesn't make check-internal-symbols.sh happy with
mingw32.  Going to disable that for DLLs again, but hopefully the
export-file is doing *something*.
2012-08-08 17:16:01 -04:00
Behdad Esfahbod f8751cf8e0 [hb-old] speed-up build 2012-08-08 17:15:44 -04:00
Behdad Esfahbod 5f4c52867c Minor 2012-08-08 16:53:37 -04:00
Behdad Esfahbod 7e7d245b33 Make default_language threadsafe 2012-08-08 15:23:48 -04:00
Behdad Esfahbod 06b192c458 Minor 2012-08-08 15:23:45 -04:00
Behdad Esfahbod 37191ede75 Minor 2012-08-08 14:59:09 -04:00
Behdad Esfahbod 6d9a329a8a Adjust a couple source checks 2012-08-08 14:48:41 -04:00
Behdad Esfahbod 9c929abdcf Minor renaming 2012-08-08 14:33:37 -04:00
Behdad Esfahbod 801298b590 Fix cast
https://bugs.freedesktop.org/show_bug.cgi?id=53233
2012-08-08 14:26:36 -04:00
Behdad Esfahbod 21756934a1 [OT] Implement fallback positioning
Implemented for Arabic, Hebrew, and generic marks.
Activated if no GPOS table present.
2012-08-08 01:20:45 -04:00
Behdad Esfahbod fb56e76283 [hb-old] Fix warnings 2012-08-07 23:44:47 -04:00
Behdad Esfahbod affaf8a0e5 [OT] Start adding fallback positioning
Used when there is no GPOS.
2012-08-07 22:43:07 -04:00
Behdad Esfahbod 7e4920fd15 Minor 2012-08-07 22:32:23 -04:00
Behdad Esfahbod 472f229a63 [GSUB] Generalize would_apply()
Fixes logic also, where before we were always matching if glyphs_len==1
and a ligature started with the glyph.
2012-08-07 22:25:24 -04:00
Behdad Esfahbod 6f3a300138 Add hb_font_glyph_from/to_string 2012-08-07 22:13:25 -04:00
Behdad Esfahbod eb56f6ae96 Minor 2012-08-07 21:44:25 -04:00
Behdad Esfahbod f4e48adcdd [OT] Apply 'rclt' feature in horizontal mode
'rclt' is "Required Contextual Forms" being proposed by Microsoft.
It's like 'calt', but supposedly always on.  We apply 'calt' anyway,
and now apply this too.
2012-08-07 21:12:49 -04:00
Behdad Esfahbod b1914b8bd0 Fix warnings 2012-08-07 16:57:48 -04:00
Behdad Esfahbod 0f8881d6bb More refactoring 2012-08-07 16:57:02 -04:00
Behdad Esfahbod 428dfcab66 Minor refactoring 2012-08-07 16:51:48 -04:00
Behdad Esfahbod 61f41849af Add Hebrew presentation forms shaping
Lifted from https://bugzilla.mozilla.org/show_bug.cgi?id=728866
2012-08-07 16:45:27 -04:00
Behdad Esfahbod 32d71dc133 [Graphite] Minor 2012-08-07 14:21:12 -04:00
Behdad Esfahbod 030ac5022e Remove enum trailing comma
...again.
2012-08-07 13:01:12 -04:00
Behdad Esfahbod 368b4e7649 Minor 2012-08-06 23:06:04 -04:00
Behdad Esfahbod ade7459ea7 [util] Fix leaks 2012-08-06 19:49:42 -07:00
Behdad Esfahbod 2fef993460 [Graphite] Fix graphite2 backend with RTL text
Patch from Martin Hosken.
2012-08-06 19:35:04 -07:00
Behdad Esfahbod e4992e13e1 [Graphite] Port graphite2 backend to new shaper infrastructure 2012-08-06 19:29:53 -07:00
Behdad Esfahbod 66591ececf Remove unnecessary lifecycle bits
We already set recount to INVALID when destroying.
This block was not necessary.
2012-08-06 17:07:19 -07:00
Behdad Esfahbod 167b625d98 [Indic] Minor, move 'blwf' after 'half'
We don't apply them together anyway.  Should not make any difference
right now.
2012-08-05 21:16:26 -07:00
Behdad Esfahbod 048e3b596f Speed up hb_set_digest_lowest_bits_t calcs 2012-08-04 20:46:45 -07:00
Behdad Esfahbod 3d1b66a35e Speed up hb_set_digest_common_bits_t calcs 2012-08-04 17:42:28 -07:00
Behdad Esfahbod 25326c2359 Rewrite ARRAY_LENGTH as a template function
Such it wouldn't apply to pointers accidentally.
2012-08-04 16:43:18 -07:00
Behdad Esfahbod 8ba8042821 [Indic] Fix consonant position font lookup logic
Oops.  I broken this badly and the test suite did not notice.  That
worries me.  Have to investigate.
2012-08-03 18:54:54 -07:00
Behdad Esfahbod abd0c05f1f Minor 2012-08-03 18:45:05 -07:00
Behdad Esfahbod 46ee108ef8 Fix leak 2012-08-03 18:21:13 -07:00
Behdad Esfahbod 71baea0062 [OT] Use general-category, not GDEF class, to decide to zero mark advances
At this point, the GDEF glyph synthesis looks pointless.  Not that I
have many fonts without GDEF lying around.

As for mark advance zeroing when GPOS not available, that also is being
replaced by proper fallback mark positioning soon.
2012-08-03 17:40:07 -07:00
Behdad Esfahbod 3a7e137a68 Dn't use gint 2012-08-03 17:23:40 -07:00
Behdad Esfahbod 11b0e20ba4 [Indic] Add per-script configuration tables
This concludes the Indic shape_plan work.  May do for Arabic also...
2012-08-02 14:21:40 -04:00
Behdad Esfahbod 85fc6c483f [Indic] Move more stuff to the shape_plan
Almost done.  Need to add per-script static tables.
2012-08-02 12:21:44 -04:00
Behdad Esfahbod 914ffaa40f [Indic] Move more repeated work into shape_plan 2012-08-02 11:05:32 -04:00
Behdad Esfahbod a8c6da90f4 [OT] Add per-complex-shaper shape_plan data
Hookup some Indic data to it.  More to come.
2012-08-02 10:46:34 -04:00
Behdad Esfahbod 8bb5deba96 [OT] Pipe shape_plan down to pause_callbacks 2012-08-02 10:07:58 -04:00
Behdad Esfahbod 3e38c0f288 More massaging 2012-08-02 09:44:18 -04:00
Behdad Esfahbod 16c6a27b4b [OT] Port complex_shaper to planner/plan 2012-08-02 09:38:28 -04:00
Behdad Esfahbod 5393e3a62b [OT] Minor refactoring 2012-08-02 09:24:35 -04:00
Behdad Esfahbod 24eacf17c8 [Indic] Move consonant-position-setting into initial_reordering() 2012-08-02 08:42:51 -04:00
Behdad Esfahbod afbcc24be0 [GSUB] Wire the font, not just the face, down to substitute()
We need the font for glyph lookup during GSUB pauses in Indic shaper.
Could perhaps be avoided, but at this point, we don't mean to support
separate substitute()/position() entry points (anymore), so there is
no point in not providing the font to GSUB.
2012-08-02 08:36:40 -04:00
Behdad Esfahbod b0e6a26a10 [OT] Hide some API
It was impossible to meaningfully use them from the outside these days.
2012-08-02 08:11:14 -04:00
Behdad Esfahbod 305246744e Minor 2012-08-02 08:08:04 -04:00
Behdad Esfahbod 8ef3d53255 [Indic] More refactoring of consonant position peeking in the font
To be moved to initial_reordering next...
2012-08-02 07:59:19 -04:00
Behdad Esfahbod 3eb6f81fd3 [Indic] Refactor
Move all the logic that needs to eventually move into the indic table
into hb-ot-shape-complex-indic-private.hh.
2012-08-02 07:38:39 -04:00
Behdad Esfahbod 3614ba242f [Indic] Rename 2012-08-02 07:23:42 -04:00
Behdad Esfahbod 610e5e8f71 [Indic] Streamline feature would_apply()
Comes with some 10% speedup for Devanagari even!
2012-08-02 05:41:18 -04:00
Behdad Esfahbod 1d002048d5 [Indic] Minor 2012-08-02 05:02:53 -04:00
Behdad Esfahbod 6f76113755 [GSUB/GPOS] Check array size before accessing digests 2012-08-02 04:00:31 -04:00
Behdad Esfahbod 22148b8c4a Use Coverage digests in would_apply 2012-08-02 03:51:51 -04:00
Behdad Esfahbod 6c459c8fef Minor 2012-08-02 03:45:53 -04:00
Behdad Esfahbod e2b8d75fa6 Use wider set digests on 64-bit archs 2012-08-01 22:17:48 -04:00
Behdad Esfahbod 0120ce9679 [GSUB/GPOS] Remove unused get_coverage() methods 2012-08-01 21:56:35 -04:00
Behdad Esfahbod 1336ecdf8e [GSUB/GPOS] Use Coverage digests as gatekeeper
Gives me a good 10% speedup for the Devanagari test case.  Less so
for less lookup-intensive tests.

For the Devanagari test case, the false positive rate of the GSUB digest
is 4%.
2012-08-01 21:46:36 -04:00
Behdad Esfahbod a878c58a8f [GSUB/GPOS] Add add_coverage() 2012-08-01 21:46:19 -04:00
Behdad Esfahbod 60a3035ac5 Add hb_set_digest_t
Implement two set digests, and one that combines the two.
2012-08-01 21:46:19 -04:00
Behdad Esfahbod c8accf1dd2 [OT] Templatize Coverage::add_coverage() 2012-08-01 21:05:57 -04:00
Behdad Esfahbod 8fbfda920e Inline font getters 2012-08-01 19:03:46 -04:00
Behdad Esfahbod 6adf417bc1 Use a lookup table for modified_combining_class 2012-08-01 18:07:42 -04:00
Behdad Esfahbod 208f70f055 Inline Unicode callbacks internally 2012-08-01 17:13:10 -04:00
Behdad Esfahbod 7470315a3e Move unicode accessors around 2012-08-01 17:01:59 -04:00
Behdad Esfahbod 21fdcee001 Add hb_unicode_combining_class_t 2012-08-01 16:28:50 -04:00
Behdad Esfahbod 84186a6400 Add commentary on the compatibility decomposition in the normalizer 2012-08-01 13:32:39 -04:00
Behdad Esfahbod 0834d95201 [hb-old] Adjust mark positioning parameters
Fallback mark positioning works now...  With hb-ft and hb-view /
hb-shape at least.
2012-08-01 00:21:09 -04:00
Behdad Esfahbod 4ca743dfb8 [old] Implement fontMetrics 2012-08-01 00:03:41 -04:00
Behdad Esfahbod 1e7d860613 [GPOS] Adjust mark advance-width zeroing logic
If there is no GPOS, zero mark advances.

If there *is* GPOS and the shaper requests so, zero mark advances for
attached marks.

Fixes regression with Tibetan, where the font has GPOS, and marks a
glyph as mark where it shouldn't get zero advance.
2012-07-31 23:41:06 -04:00
Behdad Esfahbod a8842e4a44 Remove some TODO items 2012-07-31 23:17:23 -04:00
Behdad Esfahbod 2bc3b9a616 [OT] Zero mark advances if the shaper desires so
Enabled for all shapers except for Indic.
2012-07-31 23:17:22 -04:00
Behdad Esfahbod 5fecd8b035 [OT] Synthesize glyph classes 2012-07-31 23:17:22 -04:00
Behdad Esfahbod 03b09214c0 [GSUB] Minor 2012-07-31 22:43:58 -04:00
Behdad Esfahbod f0fc1df8fc [hb-old] Implement getGlyphMetrics()
Still working on it.
2012-07-31 22:43:32 -04:00
Behdad Esfahbod 378d279bbf Implement Unicode compatibility decompositions
Based on patch from Philip Withnall.
https://bugs.freedesktop.org/show_bug.cgi?id=41095
2012-07-31 21:36:16 -04:00
Behdad Esfahbod 321ec29cc2 Remove unused function 2012-07-31 21:10:16 -04:00
Behdad Esfahbod 69cc492dc1 [buffer] Minor 2012-07-31 14:51:36 -04:00
Behdad Esfahbod 693918ef85 [OT] Streamline complex shaper enumeration
Add a shaper class struct.
2012-07-30 21:08:51 -04:00
Behdad Esfahbod c2e42c3db6 Minor 2012-07-30 19:54:50 -04:00
Behdad Esfahbod 03f67bc012 More refactoring glyph class access 2012-07-30 19:47:53 -04:00
Behdad Esfahbod 300c7307eb [OT] Don't crash if no GDEF available 2012-07-30 19:37:44 -04:00
Behdad Esfahbod 3dcbdc2125 Minor 2012-07-30 19:32:42 -04:00
Behdad Esfahbod 05bd1b6342 [GSUB/GPOS] Move glyph props matching around 2012-07-30 19:30:01 -04:00
Behdad Esfahbod 2fca1426ca [GSUB] Don't erase glyph classes if GDEF does not have glyph classes 2012-07-30 18:46:41 -04:00
Behdad Esfahbod fd42257f8c Minor 2012-07-30 18:44:10 -04:00
Behdad Esfahbod 7fbbf86efe [GSUB] Minor 2012-07-30 18:36:42 -04:00
Behdad Esfahbod 713914d320 [Uniscribe] Clean up a bit 2012-07-30 17:54:38 -04:00
Behdad Esfahbod 301168dae7 [CoreText] Port to shape_plan infrastructure 2012-07-30 17:48:04 -04:00
Behdad Esfahbod 6cdfd14bb1 Fix build on Mac 2012-07-30 17:22:17 -04:00
Behdad Esfahbod 7e34601ded Unbreak Hangul jamo composition
When we removed the separate Hangul shaper, the specific normalization
preference of Hangul was lost.  Fix that.  Also, the Thai shaper was
copied from Hangul, so had the fully-composed normalization behavior,
which was unnecessary.  So, fix that too.
2012-07-30 14:53:41 -04:00
Behdad Esfahbod 7afb14407e [Indic] Recategorize Telugu length marks
Fixes 8 more Telugu tests.  Failures at 15 (0.00154548%).
2012-07-30 13:54:46 -04:00
Behdad Esfahbod f2377155e3 [hb-old] Fix misc leaks
Backport (forward-port?!) from upstream:

commit 3ab7b37bdebf0f8773493a1fee910b151c4de30f
Author: Behdad Esfahbod <behdad@behdad.org>
Date:   Mon Jul 30 10:50:22 2012 -0400

    Fix misc leaks

    https://bugs.freedesktop.org/show_bug.cgi?id=31992
    https://bugs.freedesktop.org/show_bug.cgi?id=31993
    https://bugs.freedesktop.org/show_bug.cgi?id=31994
    https://bugs.freedesktop.org/show_bug.cgi?id=31995
2012-07-30 10:50:57 -04:00
Behdad Esfahbod 3f4764bb56 Don't lock user_data set during destruction if empty 2012-07-30 10:06:42 -04:00
Behdad Esfahbod 4ba647eecf Fix leak 2012-07-30 09:53:06 -04:00
Behdad Esfahbod f860366456 [OT] Gain back some lost speed 2012-07-30 03:16:38 -04:00
Behdad Esfahbod 11f4c87d01 [OT] Remove hb_ot_layout_ensure()
I didn't like it from the beginning.
2012-07-30 02:36:46 -04:00
Behdad Esfahbod 578e42182b Minor 2012-07-30 02:35:07 -04:00
Behdad Esfahbod a973b5ce86 [GSUB] Further adjustments to mark-attachment vs ligation interaction
The d1d69ec52e change broke Kannada badly,
since it was ligating consonants, pushing matra out, and then ligating
with the matra.  Adjust for that.  See comments.
2012-07-30 01:47:46 -04:00
Behdad Esfahbod 0aef425e25 [GSUB] Minor 2012-07-30 00:55:15 -04:00
Behdad Esfahbod d1d69ec52e [GSUB] Don't ligate glyphs attached to different components of ligatures
This concludes the mark-attachment vs ligating interaction fixes (for now).
2012-07-30 00:51:47 -04:00
Behdad Esfahbod 4751dec8be Minor 2012-07-30 00:42:07 -04:00
Behdad Esfahbod f24bcfbed1 Minor 2012-07-30 00:39:00 -04:00
Behdad Esfahbod fe20c0f84f [GSUB] Fix mark component stuff when ligatures form ligatures!
See comments.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=437633
2012-07-30 00:00:59 -04:00
Behdad Esfahbod 2ec3ba46a3 [GSUB/GPOS] Minor
Start squeezing more out of lig_id/lig_comp.
2012-07-29 22:16:15 -04:00
Behdad Esfahbod ef6e9cec33 Fixup bb0e4ba3e9 2012-07-29 21:35:22 -04:00
Behdad Esfahbod cb3d340631 [GSUB] Don't set new lig_id on mark ligatures
If two marks form a ligature, retain their previous lig_id, such that
the mark ligature can attach to ligature components...

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=676343

In fact, I noticed that we should not let ligatures form between glyphs
coming from different components of a previous ligature.  For example,
if the sequence is: LAM,SHADDA,LAM,FATHA,HEH, the LAM,LAM,HEH form a
ligature, putting SHADDA and FATHA next to eachother.  However, it would
be wrong to ligate them.  Uniscribe has this bug also.
2012-07-29 20:37:38 -04:00
Behdad Esfahbod a15b70a81a [hb-old] Fix cluster formation in RTL
Unlike Uniscribe, hb-old returns glyphs in logical order, so the logic
does not need to duplicated for RTL.
2012-07-29 20:09:22 -04:00
Behdad Esfahbod 8a7e70ef65 [Minor] 2012-07-29 19:56:54 -04:00
Behdad Esfahbod bb0e4ba3e9 Minor 2012-07-29 17:34:14 -04:00
Behdad Esfahbod a00ad60bc0 [Uniscribe] Remove hb_uniscribe_font_ensure()
Wasn't a huge fan of putting the burden on the user.  Just remove it and
do what we've got to do transparently.
2012-07-28 21:16:08 -04:00
Behdad Esfahbod 5d874d566f [GPOS] Fix mark-to-mark positioning when one of the marks is a ligature
This commit: a3313e5400 broke MarkMarkPos
when one of the marks itself is a ligature.  That regressed 26 Tibetan
tests (up from zero!).  Fix that.  Tibetan back to zero.
2012-07-28 21:05:25 -04:00
Behdad Esfahbod 338fe662b5 [GSUB] Minor 2012-07-28 18:53:01 -04:00
Behdad Esfahbod e6f7479fe3 [GSUB] Simplify would-apply 2012-07-28 18:34:58 -04:00
Behdad Esfahbod dadede012e Minor 2012-07-28 18:13:09 -04:00
Behdad Esfahbod 0b99429ead [GSUB/GPOS] Add get_coverage() and use it to speed up main loop
And use it to speed up the hotspot by checking coverage directly in
the main loop, not 10 functions deep in.

Gives me a solid 20% boost with Indic test suite.  Less so for less
lookup-intensive scenarios.

Remove the "fast_path" hack from before.
2012-07-28 17:46:35 -04:00
Behdad Esfahbod 30ec9002d8 Reject lookups with no subTable 2012-07-28 17:25:20 -04:00
Behdad Esfahbod 0981068b75 [GSUB/GPOS] Reject Context/ChainContext lookups with zero input 2012-07-28 17:01:59 -04:00
Behdad Esfahbod 2f87cebe10 Implement shape_plan caching
Should give us some performance boost.
2012-07-27 04:20:39 -04:00
Behdad Esfahbod e9eb9503e9 Add default_shaper_list to shape_plan 2012-07-27 03:16:22 -04:00
Behdad Esfahbod 3b7c4e2706 Don't fail choosing shaper on planning failure
Shapers have a chance to reject a font in face shaper_data creation.
No need to allow failing during planning.
2012-07-27 03:12:23 -04:00
Behdad Esfahbod cfe9882610 Add hb_ot_layout_ensure() and hb_uniscribe_font_ensure() 2012-07-27 03:06:30 -04:00
Behdad Esfahbod c5b668fb92 Choose one shaper per plan 2012-07-27 02:49:39 -04:00
Behdad Esfahbod e82061e8db Move ot shaper completely to shape_plan 2012-07-27 02:29:32 -04:00
Behdad Esfahbod ea278d3895 Partially switch ot shaper to shape_plan 2012-07-27 02:12:28 -04:00
Behdad Esfahbod b6b7ba1313 Switch old and uniscribe backends to shape_plan 2012-07-27 01:37:18 -04:00
Behdad Esfahbod c32c096a42 Switch to shape_plan
Not optimized yet.  Eats babies.  And no shaper uses the shape_plan.
2012-07-27 01:13:53 -04:00
Behdad Esfahbod 5b95c148cc Start implementing shape_plan 2012-07-27 01:02:24 -04:00
Behdad Esfahbod bd26b4d21f Minor 2012-07-26 22:18:24 -04:00
Behdad Esfahbod 027857d041 Start adding a unified shaper access infrastructure
Add global shape_plan.  Unused so far.
2012-07-26 21:14:02 -04:00
Behdad Esfahbod fa2dfcd560 Fix visibility warnings with MinGW32 2012-07-26 16:06:16 -04:00
Jonathan Kew ac2085d4b3 [CoreText] Ensure cluster indices in output buffer are non-decreasing.
Does not provide Uniscribe-compatible results, but should at least avoid
breaking hb-view due to out-of-order cluster values.

For RTL runs, ensure cluster values are non-increasing (instead of
non-decreasing).
2012-07-26 15:58:45 -04:00
Behdad Esfahbod 441d3bb7de Minor 2012-07-26 12:01:12 -04:00
Behdad Esfahbod 2e7f223054 [hb-old] Fix Arabic cursive positioning
Backporting from upstream:

commit b847f24ce855d24f6822bcd9c0006905e81b94d8
Author: Behdad Esfahbod <behdad@behdad.org>
Date:   Wed Jul 25 19:29:16 2012 -0400

    [arabic] Fix Arabic cursive positioning

    This was clearly broken in testing.  Who knows...  Fixes for me.
    Test with a Nastaleeq font, or with Arabic Typesetting.

    Backporting from Chromium.
2012-07-25 19:30:15 -04:00
Behdad Esfahbod 9550a8c4e8 [hb-old] Fixup not-enough-space handling 2012-07-25 19:22:57 -04:00
Behdad Esfahbod 91e721ea86 [hb-old] Fix clusters
Unlike its "documentation", hb-old's log_clusters are, well, indeed
logical, not visual.  Fixup.  Adapted / copied from hb-uniscribe.
2012-07-25 19:20:34 -04:00
Behdad Esfahbod a3313e5400 [GPOS] Fix MarkMarkPos applied to results of MultipleSubst
This was broken as a result of 7b84c536c1.
As Khaled reported, MarkMark positioning was broken with glyphs
resulting from a MultipleSubst.  Fixed.  Test with the ALLAH character
in Amiri.
2012-07-25 18:37:51 -04:00
Behdad Esfahbod 35bdab3cf1 Minor 2012-07-25 11:59:52 -04:00
Behdad Esfahbod 8fe4c7405b [hb-old] Add HarfBuzz.old shaper
Choose using shaper name "old".
2012-07-25 11:11:22 -04:00
Behdad Esfahbod 5e1987005e [hb-old] Define Unicode funcs in terms of new HarfBuzz 2012-07-25 11:11:22 -04:00
Behdad Esfahbod 4a31166b28 [hb-old] Shovel out the line-breaking / word-segmentation stuff 2012-07-25 11:11:22 -04:00
Behdad Esfahbod 0bcbe88cf3 [hb-old] Add visibility attributes 2012-07-25 11:11:22 -04:00
Behdad Esfahbod 6a9d43c317 [hb-old] Remove unused header file 2012-07-25 11:11:22 -04:00
Behdad Esfahbod fb47209c5b [hb-old] Rename hb_buffer_* to HB_Buffer_* 2012-07-25 11:11:22 -04:00
Behdad Esfahbod 1512a73575 [hb-old] Start adding HarfBuzz-old as a new backend 2012-07-25 11:11:16 -04:00
Behdad Esfahbod 478fd0529b Minor 2012-07-24 17:09:01 -04:00
Behdad Esfahbod 8979a7f6f2 [Mongolian] Remove Mongolian Vowel Separator at the end of shaping
Results match Uniscribe now.
2012-07-24 17:03:55 -04:00
Jonathan Kew aa6d849838 [CoreText] Add basic Core Text backend for comparison with our native shaping
Does not attempt to handle clusters in a Uniscribe- or HarfBuzz-compatible way;
just returns the original string indexes that CT maintains. These may even be
out-of-order in the case of reordrant glyphs.
2012-07-24 15:52:32 -04:00
Behdad Esfahbod ec8d249469 Make data members of various OpenType structs protected instead of private
Should fix warnings generated when building with -Wunused-private-field.
Based on patch from Jonathan Kew.
2012-07-24 15:40:37 -04:00
Behdad Esfahbod 97aa0b738a Minor const correctness shuffling 2012-07-24 15:02:34 -04:00
Behdad Esfahbod 6411e74caf [Indic] Reposition Gurmukhi top matras to after post
The font is forming a post-base consonant in some samples, and Uniscribe
positions top matra on the post-base.  Do the same.

Gurmukhi failures down from 59 to 41 (0.0674242%).
2012-07-24 13:48:49 -04:00
Behdad Esfahbod 65c43accdc [Indic] Better position left-matra in Malayalam
Just put it before base, which is what's expected.

Malayalam failures down from 1559 to 1197 (0.114172%).

BENGALI: 353988 out of 354285 tests passed. 297 failed (0.0838308%)
DEVANAGARI: 693571 out of 693628 tests passed. 57 failed (0.00821766%)
GUJARATI: 366489 out of 366506 tests passed. 17 failed (0.0046384%)
GURMUKHI: 60750 out of 60809 tests passed. 59 failed (0.0970251%)
KANNADA: 950956 out of 951913 tests passed. 957 failed (0.100534%)
KHMER: 299094 out of 299124 tests passed. 30 failed (0.0100293%)
MALAYALAM: 1047219 out of 1048416 tests passed. 1197 failed (0.114172%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271699 out of 271847 tests passed. 148 failed (0.0544424%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970524 out of 970573 tests passed. 49 failed (0.00504856%)
2012-07-24 03:36:47 -04:00
Behdad Esfahbod 88f413b56f [Indic] Implement Reph+Ya-Phalaa interaction
The sequence Ra,H,Ya in Bengali is ambigious and Unicode encoded that to
get Ya-Phalaa, one would place ZWJ before Halant.  Ie. a ZWJ,H sequence
requests subjoining, while a H,ZWJ requests Half form.  Implement that.

Bengali failures go down from 377 to 297 (0.0838308%).
Gujarati is down by 4 to 17 (0.0046384%).
Kannada is down by 226 to 957 (0.100534%).

Current status:

BENGALI: 353988 out of 354285 tests passed. 297 failed (0.0838308%)
DEVANAGARI: 693571 out of 693628 tests passed. 57 failed (0.00821766%)
GUJARATI: 366489 out of 366506 tests passed. 17 failed (0.0046384%)
GURMUKHI: 60750 out of 60809 tests passed. 59 failed (0.0970251%)
KANNADA: 950956 out of 951913 tests passed. 957 failed (0.100534%)
KHMER: 299094 out of 299124 tests passed. 30 failed (0.0100293%)
MALAYALAM: 1046857 out of 1048416 tests passed. 1559 failed (0.148701%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271699 out of 271847 tests passed. 148 failed (0.0544424%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970524 out of 970573 tests passed. 49 failed (0.00504856%)
2012-07-24 03:04:36 -04:00
Behdad Esfahbod dff0ece11d [Indic] Limit matras to 4 per syllable
Also limit joiners.

This limits our syllable length to a constant, and is
closer to what Uniscribe does anyway.

Two Devanagari tests regressed, but who cares about tests with 20
joiners in a row?!  Devanagari at 57 (0.00821766%) now.
2012-07-24 02:37:42 -04:00
Behdad Esfahbod 330b329c89 [Indic] Unmark U+17D1 KHMER SIGN VIRIAM to NOT be a Virama
Fixes another 1 Khmer failure.  Down to 30 (0.0100293%) now.
2012-07-24 02:25:26 -04:00
Behdad Esfahbod 6824a7194e [Indic] Recategorize Khmer various signs as top matras
Khmer failures down from 39 to 31 (0.0103636%).
2012-07-24 02:22:18 -04:00
Behdad Esfahbod d90b8e841e [Indic] Reposition Khmer prebase-reordering Ra around split matras
In Khmer coeng model, a V,Ra can go *after* matras.  If it goes after a
split matra, it should be reordered to *before* the left part of such matra.

Khmer failures down from 136 to 39 (0.0130381%).
2012-07-24 02:11:18 -04:00
Behdad Esfahbod 0afb84c125 [Indic] Fix minor bug in pre-base Ra positioning 2012-07-24 01:44:47 -04:00
Behdad Esfahbod 7573799126 [Indic] Position Khmer U+17CE
Fixes another 6 Khmer failures.  Now at 136 (0.0454661%).
2012-07-24 01:32:07 -04:00
Behdad Esfahbod 8d00e8d0e7 [Indic] Don't reposition Khmer Bindu
Khmer Bindu doesn't like to move to syllable end.  Leave it where it
was.

Brings down Khmer failures from 510 to 142 (0.047572%).
2012-07-24 01:15:34 -04:00
Behdad Esfahbod 2278eefcdb [Indic] In Sinhala, form forced Reph even if no other consonant found
Fixes another 10 Sinhala failures.  Down to 148 (0.0544424%).
2012-07-24 00:31:10 -04:00
Behdad Esfahbod 71fd5e80ad [Indic] Further adjust base algorithm for Sinhala
Apparently if there is C,V,ZWJ,C, the first C will be base, but if
it's C,ZWJ,V,C, the second one will be.

Note that Uniscribe implements this differently, by breaking syllable in
the case of C,ZWJ,V,C and putting the first consonant in one syllable
and the rest in the next syllable.

Sinhala failures down from 208 to 158 (0.0581209%).  No changes to
Khmer.
2012-07-24 00:21:16 -04:00
Behdad Esfahbod 73d71cc527 [Indic] End Vowel-based syllable at ZWJ
One Devanagari test regressed, plus 10 Malayalam (at 1545 now).

Fixed 120 Sinhala failures.  Now at 208 (0.0765136%).
2012-07-24 00:09:12 -04:00
Behdad Esfahbod 34c215036f [Indic] Improve Sinhala base algorithm and reph positioning
Sinhala does not have half forms.  And most (all?) consonants can be
base, except when preceded by ZWJ, which would request a subjoined form.
Hence switch the base algorithm to categorize with Khmer, start search
at start, and stop at a ZWJ.

Also, mark all pos=base consonants after base to be subjoined.  Mark
base itself to have pos=base.

Finally, adjust Sinhala's reph position to after-main.

Brings down Sinhala failures from 455 to 328 (0.120656%).
2012-07-23 23:51:29 -04:00
Behdad Esfahbod 2ec934c6c2 [Indic] Change "unknown" position to end of syllable 2012-07-23 23:49:04 -04:00
Behdad Esfahbod b70021f7c8 When removing zero-width marks, don't remove ligatures
If a mark ligated, it probably should NOT be removed.
2012-07-23 20:18:17 -04:00
Behdad Esfahbod 49c5ec5144 Minor refactoring 2012-07-23 20:14:13 -04:00
Behdad Esfahbod c3e6fdc379 [Indic] Improve check on ligatures
Only skip actual ligatures, not marks in-between ligature components.
2012-07-23 20:11:42 -04:00
Behdad Esfahbod 771a8f5028 [Indic] exclude ligatures when matching on Indic category
If, say, a H,ZWJ,C ligature was formed, we don't want the code to detec
that as a Halant.  So, ignore ligatures when matching category in
final_reordering.

Sinhala failures down from 514 to 455 (0.167374%).
2012-07-23 20:09:30 -04:00
Behdad Esfahbod d1af9e82e5 [GSUB/GPOS] Const correctness 2012-07-23 19:55:35 -04:00
Behdad Esfahbod baacd090df [Indic] Minor refactoring 2012-07-23 19:51:48 -04:00
Behdad Esfahbod c7c4de2fb9 [Indic] Remove syllable length check before sorting
We now limit syllable lengths in the machine.  No need to match here.
2012-07-23 18:25:02 -04:00
Behdad Esfahbod 9fa052733e [Indic] Limit syllables to at most five consonants
Seems to be about what Uniscribe does.  Not exactly.  But close enough.
More consonants will start a new cluster.

A few scripts went way down in failures.  In particular:

  - Devanagari failures went down from 490 to 56.
  - Telugu went down from 113 to 49.

Other scripts went down slightly or didn't change.  New numbers:

BENGALI: 353908 out of 354285 tests passed. 377 failed (0.106412%)
DEVANAGARI: 693572 out of 693628 tests passed. 56 failed (0.00807349%)
GUJARATI: 366485 out of 366506 tests passed. 21 failed (0.00572978%)
GURMUKHI: 60750 out of 60809 tests passed. 59 failed (0.0970251%)
KANNADA: 950730 out of 951913 tests passed. 1183 failed (0.124276%)
KHMER: 298613 out of 299124 tests passed. 511 failed (0.170832%)
MALAYALAM: 1046881 out of 1048416 tests passed. 1535 failed (0.146411%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271333 out of 271847 tests passed. 514 failed (0.189077%)
TAMIL: 1091837 out of 1091837 tests passed. 0 failed (0%)
TELUGU: 970524 out of 970573 tests passed. 49 failed (0.00504856%)

Some of the remaining Telugu and Devanagari issues seem to be Uniscribe
eating Anusvara when placed before a non-joiner.  Ouch!
2012-07-23 18:19:17 -04:00
Behdad Esfahbod 093cd58326 [Thai] Fix SARA AM handling
Oops, thinko.
2012-07-23 14:04:42 -04:00
Behdad Esfahbod 42848453bf [Thai] Reorder U+0E3A THAI VOWEL SIGN PHINTHU
Uniscribe reorders U+0E3A to be after U+0E38 and U+0E39.  We do that by
modifying the ccc for U+0E3A.

Fixes the two remaining Thai failures (see previous commit).
2012-07-23 13:52:07 -04:00
Behdad Esfahbod 4a7f4f3e56 [Thai] Adjust SARA AM reordering to match Uniscribe
Adjust the list of marks before SARA AM that get the reordering
treatment.  Also adjust cluster formation to match Uniscribe.

With Wikipedia test data, now I see:

  - For Thai, with the Angsana New font from Win7, I see 54 failures out
    of over 4M tests  (0.00129107%).  Of the 54, two are legitimate
    reordering issues (fix coming soon), and the other 52 are simply
    Uniscribe using a zero-width space char instead of an unknown
    character for missing glyphs.  No idea why.  The missing-glyph
    sequences include one that is a Thai character followed by an Arabic
    Sokun.  Someone confused it with Nikhahit I assume!

  - For Lao, with the Dokchampa font from Win7, 33 tests fail out of
    54k (0.0615167%).  All seem to be insignificant mark positioning
    with two marks on a base.  Have to investigate.
2012-07-23 13:15:33 -04:00
Behdad Esfahbod 2cc933aff9 [Indic] Fix cluster formation with left-matras and conjunct forms
Test case was: <U+0D15,U+0D4D,U+0D15,U+0D4A>.
2012-07-23 08:23:44 -04:00
Behdad Esfahbod e6b01a878c [Indic] Further streamline cluster formation
This should address all possible cluster misformations that I had in
mind.
2012-07-23 00:11:26 -04:00
Behdad Esfahbod 7b2a7dadd6 [Indic] Merge clusters before sorting
This should fix any instabilities in cluster formation that we were
speculating may happen with surrounding syllables.  Or most of it
perhaps.
2012-07-22 23:58:55 -04:00
Behdad Esfahbod abb3239ef9 [Indic] Update clusters for left-matra even if matra didn't move
Fixes crashes reported with left matra under
non-uniscribe-bug-compatibilty mode.
2012-07-22 23:55:19 -04:00
Behdad Esfahbod 92a1ad7bef [Indic] Stop searching for base if a post form is found before below form
Improves Bengali and Gurmukhi.  Malayalam regressed a bit.  We will deal
with that later.
2012-07-20 18:55:15 -04:00
Behdad Esfahbod 4c450c703f [Indic] Recompose Bengali Ya,Nukta
This is a bunch of hacks for now.

Improves Bengali a bit.
2012-07-20 18:13:04 -04:00
Behdad Esfahbod e9c0f152a3 [Uniscribe] Fix script fallback
Gurmukhi failures half now.  Others changed slightly.
2012-07-20 17:37:48 -04:00
Behdad Esfahbod 5791f32915 [Indic] Allow a ZWNJ after SM's
Malayalam failures go way down.  Other scripts benefitted slightly too.
Sinhala had one or two test regressions, but...
2012-07-20 16:26:55 -04:00
Behdad Esfahbod 34ae336f3f [Indic] Improve Reph AfterMain positioning
Fixes 20 out of 48 failing Oriya tests.  Failure rate down to 0.066% now.
2012-07-20 16:17:28 -04:00
Behdad Esfahbod bdd080431a [Indic] Reposition Oriya Candrabindu
Oriya failures down from 0.65% to 0.20%.
2012-07-20 16:03:09 -04:00
Behdad Esfahbod 5f0eaaad12 [Indic] Fix base search in final_reordering
Fixes most Malayalam failures.  Down from 1.6% to 0.38% now.  Fixes a
few more in other scripts too.
2012-07-20 15:47:24 -04:00
Behdad Esfahbod 81202bd860 [Indic] Don't attach SM/VD to other characters 2012-07-20 15:14:51 -04:00
Behdad Esfahbod efb4ad7356 Fix compiler warnings
If x is not constant, we cannot ASSERT_STATIC on it.
2012-07-20 14:27:38 -04:00
Behdad Esfahbod f31d97e44e [Indic] Form Telugu Reph out of Ra,Virama,ZWJ
Apparently this was approved in Feb 2012.  No font yet.
2012-07-20 14:13:35 -04:00
Behdad Esfahbod 2e193b240e [Indic] Don't split U+0AC9
Althought IndicMatraCategory.txt classifies it as Top_And_Right matra,
it does not have Unicode decomposition, and Uniscribe does not do
anything special about it either.

Gujarati failures down from 0.672% to 0.0130966%.
2012-07-20 14:02:35 -04:00
Behdad Esfahbod 30c3d5e9fc [Indic] Simplify Uniscribe cluster emulation
Now that we break syllables on Halant,ZWNJ, this code can be simplified.
2012-07-20 13:56:32 -04:00
Behdad Esfahbod decf6ffca4 [Indic] Minor! 2012-07-20 13:51:31 -04:00
Behdad Esfahbod 9e4f94a72c [Indic] Break syllables at Halant,ZWNJ
That's really what Uniscribe does, and explains a lot of pecularities of
Halant,ZWNJ before the base.

Sent Telugu from 1% failures to 0.03%.  Improved Kannada and Malayalam
slightly.  Fixed half of Bengali, and did NOT break anything!
2012-07-20 13:48:03 -04:00
Behdad Esfahbod 2c372b80f6 [Indic] Better check for applying 'init'
Specifically, don't apply 'init' if previous char is a joiner.

Fixes some more of Bengali.
2012-07-20 13:37:48 -04:00
Behdad Esfahbod 34a7440b7c [GPOS] Don't zero mark advances
Fixes more of Telugu, Kannada, and Oriya.

May break things (outside Indic...), but we cannot think of any font relying
on this immediately.
2012-07-20 12:40:39 -04:00
Behdad Esfahbod 8ed248de77 [Indic] Minor 2012-07-20 11:42:24 -04:00
Behdad Esfahbod d0e68dbd0b [Indic] Implement reph positioning step 5
Not tuned, just copied from step 2.  Fixes another 0.5% of Kannada
failures.  1% to go.
2012-07-20 11:25:41 -04:00
Behdad Esfahbod a9e45c32e4 [Indic] Don't let ZWNJ at the end of syllable affect base search
Fixes a few Devanagari, half of remaining Kannada failures, quarter for
Telugu, and others slightly improved or unchanged.
2012-07-20 11:04:15 -04:00
Behdad Esfahbod 20b68e699f [Indic] Apply 'cjct' globally
Fixes 5 Devanagari failures, and no regressions.
2012-07-20 10:47:46 -04:00
Behdad Esfahbod 51e764de44 [Indic] Unbreak old scriptures
Brings down failures with Lohit-Telugu from 57% to 1.40%.
2012-07-20 10:30:24 -04:00
Behdad Esfahbod 900cf3d449 Minor 2012-07-20 10:18:23 -04:00
Behdad Esfahbod 87cd63266e [Indic] Recategorize some Kannada right matras
Kannada failures down from 3.5% to 2.93%.
2012-07-19 21:25:46 -04:00
Behdad Esfahbod 3604d64ced [Indic] Recategorize GURMUKHI ADDAK
It's not in IndicSyllabicCategory.txt.  Fixes most of Gurmukhi failures.
Failures down from 7.7% to 0.222%!
2012-07-19 21:13:04 -04:00
Behdad Esfahbod 8932858123 Minor 2012-07-19 21:02:38 -04:00
Behdad Esfahbod 47ef931f13 [buffer] Make sure out_info = info during GPOS 2012-07-19 20:52:44 -04:00
Behdad Esfahbod ae63cf2062 Print line number during return when tracing 2012-07-19 20:45:41 -04:00
Behdad Esfahbod 5249f3aee1 [Indic] Unbreak Khmer
For Khmer, all consonants are subjoining.  No need to look in the font.
We were looking in the wrong order anyway.
2012-07-19 20:30:22 -04:00
Behdad Esfahbod e0475345d5 [Indic] Apply 'akhn' globally
Fixes 1.5% more failures for Telugu, 2% for Kannada.
Breaks one test in Devanagari.
2012-07-19 20:24:14 -04:00
Behdad Esfahbod fa247ebe52 [Indic] Better position U+0CD5
Fixes another 5% of Kannada failures.
2012-07-19 19:52:19 -04:00
Behdad Esfahbod f055442716 [Indic] Lookup consonant position in the font
Fixes most failures of Oriya, and improves others a bit.
2012-07-19 16:20:21 -04:00
Behdad Esfahbod 74d1d88781 [GSUB] Fix would_apply() for LigatureSubst 2012-07-19 16:14:23 -04:00
Behdad Esfahbod be73a5f936 Add src/test-would-substitute tool 2012-07-19 15:12:18 -04:00
Behdad Esfahbod e72b360ac6 Refactor / finish would_apply() operation
Untested.
2012-07-19 14:44:46 -04:00
Behdad Esfahbod 8c973ebf0f [Indic] Implement per-script matra positioning
Following what the spec says.

Brings down Telugu failures from 40% to 3.75%, and Kannada failures from
44% to 10%.  Does NOT affect other scripts' test results.
2012-07-19 13:25:08 -04:00
Behdad Esfahbod 8bb32458f9 [Indic] More refactoring 2012-07-19 13:04:44 -04:00
Behdad Esfahbod 9ccc6382ba [Indic] Minor refactoring 2012-07-19 12:45:31 -04:00
Behdad Esfahbod f83aaa3133 [Indic] Minor 2012-07-19 12:23:23 -04:00
Behdad Esfahbod be8b9f5f71 [Indic] Start refactoring different matra positions per script 2012-07-19 12:11:12 -04:00
Behdad Esfahbod b01d9b3d90 [Indic] Disallow decomposition of a couple characters
This is a hack for now.  Will be fixed when we do complex-shaper-driven
normalization properly.

The results with or without decomposition are the same, but Uniscribe
does not normalize, so this matches better.
2012-07-19 11:25:49 -04:00
Behdad Esfahbod 422ecd2d3c [Indic] Accept a forced Rakar sequence at the end of syllable
In Sinhala, Rakar is formed by Al-Lakuna,ZWJ,Ra.  If you put that at the
end of a Consonant,Matra syllable, you get a dotted-circle from
Uniscribe.  Apparently adding a ZWJ before the Al-Lakuna "fixes" that.
And people have been encoding that sequence...  So, allow a forced
"ZWJ,Virama,ZWJ,Ra" sequence at the of syllables.

Fixes some 100 or more of Sinhala failures.  Now at 622 only (0.23%).
2012-07-18 23:25:58 -04:00
Behdad Esfahbod 6fc1732003 [Indic] Allow joiners on both sides of Halant at the same time
The sequence <ZWJ,Al-Lakuna,ZWJ> is used in Sinhala to explicitly ask
for Rakar.  Fixes two-thousand Sinhala tests.  Not many left.
2012-07-18 17:49:19 -04:00
Behdad Esfahbod 10cdc94eee [Indic] In final reordering, find base, even if it disappeared
POS_BASE can disappear if base ligated backward.  Define base as last
with position not after base.

Fixes a few hundred of Sinhala failures with Iskoola Pota.
2012-07-18 17:43:23 -04:00
Behdad Esfahbod 9c4d24a3a6 [Indic] Minor 2012-07-18 17:29:10 -04:00
Behdad Esfahbod 3285e107c9 [Indic] Implement Sinhala "Al Lakuna" Reph behavior
In Sinhala, Reph is formed only explicitly, by the presence of a ZWJ.
2012-07-18 17:22:14 -04:00
Behdad Esfahbod 91cade7555 [Indic/Unicode] Decompose Sinhala split matras the way Uniscribe likes
Makes no visual difference.

Fixes most of the failures.  Down from 15% to 1.3%!
2012-07-18 16:50:41 -04:00
Behdad Esfahbod d8942dcbb4 Apply Tibetan (global) features.
Fixes all Tibetan failures.  All 180k of them!

Merges back Hangul into the default shaper.
2012-07-18 16:34:10 -04:00
Behdad Esfahbod 552d19b7a1 [Indic] Treat Register Shifters like Nukta
Really this time.

Fixes another 18 Khmer tests.
2012-07-18 16:02:33 -04:00
Behdad Esfahbod e8cd81f76d [Indic] Minor 2012-07-18 16:00:20 -04:00
Behdad Esfahbod 69f26bf39c [Indic] Fix Matra reordering when base is at end of syllable
For example: U+915,U+200c,U+93f

Fixes last Tamil failure!
2012-07-18 15:47:51 -04:00
Behdad Esfahbod d16ccc4ae7 Leave one extra item at the end of buffer allocation
Just in case, for the times we do out-of-bounds access.

jk
2012-07-18 15:43:55 -04:00
Behdad Esfahbod 075d671f10 [Indic] Fix out-of-bounds array access 2012-07-18 15:41:53 -04:00
Behdad Esfahbod dcb527242b [Indic] Allow joiners before matras
Fixes 1 more Devanagari test!
2012-07-18 15:32:26 -04:00
Behdad Esfahbod 391cc03317 [Indic] Allow halant group in Vowel and placeholder syllables
Fixes 2 out of 560 Devanagari failures.  AND:
Fixes 1 out of 2 Tamil failures.
2012-07-18 15:12:49 -04:00
Behdad Esfahbod ca4e3d3eab [Indic] Streamline halant/joiner in grammar 2012-07-18 15:05:40 -04:00
Behdad Esfahbod 418d00dffd [Indic] Minor 2012-07-18 14:57:28 -04:00
Behdad Esfahbod 4c3691d2a3 [Indic] Hopefully minor!
Refactoring Indic machin.  No semantic change.
2012-07-18 14:23:55 -04:00
Behdad Esfahbod e092c556fb [Indic] Minor 2012-07-18 14:09:25 -04:00
Behdad Esfahbod 14dbdd9e39 [Indic] Unbreak Tamil
Tamil has only about 150 failures now!
2012-07-18 13:13:03 -04:00
Behdad Esfahbod db8981f1e0 [Indic] Position Khmer Robat
It's a visual Repha.

Still not positioning logical Repha as occurs in Malayalam.

Another 200 Khmer failures fixed.  547 to go.  That's better than
Devanagari!
2012-07-17 23:42:04 -04:00
Behdad Esfahbod 25bc489498 [Indic] Better categorize Register Shifters and Khmer Various signs
Down another 500 or so Khmer failures!
2012-07-17 17:53:03 -04:00
Behdad Esfahbod 39b17837b4 Add hb_buffer_normalize_glyphs() and hb-shape --normalize-glyphs
This reorders glyphs within the cluster to a nominal order.  This should
have no visible effect on the output, but helps with testing, for
getting the same hb-shape output for visually-equal glyphs for each
cluster.
2012-07-17 17:09:29 -04:00
Behdad Esfahbod 25e302da9a [Indic] Minor 2012-07-17 14:25:14 -04:00
Behdad Esfahbod 5d32690a34 [Indic] For scripts without Half forms, always choose first consonant as base
In such scripts (ie. Khmer), a ZWJ/ZWNJ shouldn't stop the search for
base.  So, instead just choose the first consonant as base directly.

Test sequence:
U+1798,200c,U+17C9,U+17D2,U+179B,U+17C1,U+17C7
2012-07-17 14:23:28 -04:00
Behdad Esfahbod 34b5714906 [Indic] Treat Khmer Register Shifters more like Nuktas
Except that there may be a ZWNJ before a Register Shifter.
2012-07-17 14:09:32 -04:00
Behdad Esfahbod 11e2a601b1 [Indic] Minor 2012-07-17 14:02:28 -04:00
Behdad Esfahbod 0201e0a464 [Indic] Apply 'cfar' for Khmer
Mark stuff after a pre-base reordering Ro 'cfar'.  Used in Khmer.
This allows distinguishing the following cases with MS Khmer fonts:

  U+1784,U+17D2,U+179A,U+17D2,U+1782
  U+1784,U+17D2,U+1782,U+17D2,U+179A
2012-07-17 13:56:24 -04:00
Behdad Esfahbod 55f70ebfb9 [Indic] Position final subjoined consonants (and vowels) after matras
In Khmer, a final subjoined consonant or independent vowel can occur
after matras.  This final subjoined thing should NOT be reordered to
before the matra even though it's subjoined.

Fixes another 1k of the Khmer failures.  Not much left really.
2012-07-17 12:50:13 -04:00
Behdad Esfahbod c50ed71e9a [Indic] Recategorize Khmer coeng sign as a separate category OT_Coeng
Amend the syllable structure to allow a final subscripted consonant
(Coeng+C) and a final subscripted independent vowel (Coeng+V).
Fixes another 2k of Khmer failures.
2012-07-17 11:54:28 -04:00
Behdad Esfahbod deb521dee4 [Indic] Add a separate Coeng class
No characters recategorized yet.  No semantic change.
2012-07-17 11:37:32 -04:00
Behdad Esfahbod 74ccc6a132 [Indic] Move Halant with after-base consonants
Normally, we attach the Halant to the previous character and move it
with it.  For after-base consonants however, the Halant "belongs" to the
consonant after, so attach it so.

This fixes Bengali sequences involving post-base consonant Ya, which
should ligate with the Halant to form Ya Phala, but previously a
reordered matras was blocking the ligation.
2012-07-17 11:16:19 -04:00
Behdad Esfahbod d5c4edcdd6 [Indic] Apply presentation-forms features all at once
Seems like this is what Uniscribe is doing, and does not break any fonts
we tested (with Devanagari, Malayalam, Khmer, and Bengali), while fixing
some Ra Phala sequences for Bengali with Vrinda.  Fixes another 2% of
Bengali failures (a couple more to go).
2012-07-17 10:40:59 -04:00
Behdad Esfahbod 559f706678 Fix MarkAttachmentType matching
Fixes issue reported by Khaled Hosny with his Hussaini Nastaleeq font
and sequences like those added in the previous commit.
2012-07-16 22:46:52 -04:00
Behdad Esfahbod ad4494759f Minor 2012-07-16 22:40:21 -04:00
Behdad Esfahbod af92b4cc90 [Indic] Disable 'kern' in Uniscribe bug compatibility mode
Uniscribe does not apply 'kern' in the Indic module.  Some of the Khmer
fonts they ship have small adjustments in the 'kern' table.  Disable
'kern' in the Indic module under Uniscribe bug compatibility mode.

Fixes some 10% of the Khmer failures.  Remains under 3% (excluding
dotted-circle ones).
2012-07-16 20:31:24 -04:00
Behdad Esfahbod d96838ef95 Allow complex shapers overriding common features
In a new callback...  Currently unused by all complex shapers.
2012-07-16 20:26:57 -04:00
Behdad Esfahbod df50b84740 [Indic] Categorize other Khmer marks
Mark them the same as the Register Shifters for now.  Need to rename
that category to something more sensible after all is settled.

Fixes another percent of Khmer failures.  Down to under 3%!
2012-07-16 20:14:50 -04:00
Behdad Esfahbod 8e7b5882fb [Indic] Recognize pre-base reordering Ra anywhere in the syllable
We were doing that only immediately after base.

Fixes another percent in the Khmer failures.  About three more to go...
2012-07-16 17:04:46 -04:00
Behdad Esfahbod 7d09c98a1f [Indic] Recognizer Register Shifter marks
Fixes another 6% of the Khmer failures.
2012-07-16 16:45:22 -04:00
Behdad Esfahbod 60da763dfa [GSUB/GDEF] Guess glyph classes after substitution only if no GDEF
Brings down Khmer failures with Daun Penh font from 36% to 20%.
2012-07-16 16:14:40 -04:00
Behdad Esfahbod fcdc5f1c88 [Indic] Categorize Khmer Ro
Khmer failures down from 58% to 36%.
2012-07-16 15:52:54 -04:00
Behdad Esfahbod 78818124b1 [Indic] Reoder pre-base reordering Ra
Brings down Malayalam failures from 14% down to 3%.
2012-07-16 15:49:08 -04:00
Behdad Esfahbod 1a1dbe9a27 [Indic] Rename 2012-07-16 15:41:33 -04:00
Behdad Esfahbod 46e645ec4b [Indic] Start implementing pre-base reordering 2012-07-16 15:30:05 -04:00
Behdad Esfahbod 921ce5b17d [Indic] Rename
No semantic change.
2012-07-16 15:26:56 -04:00
Behdad Esfahbod b504e060f0 [Indic] Implement After-Main Reph positioning
Almost...
2012-07-16 15:21:12 -04:00
Behdad Esfahbod 17d7de91d7 [Indic] Apply 'pref' to pre-base reodering Ra
No reordering yet.
2012-07-16 15:20:15 -04:00
Behdad Esfahbod 362d3db8d3 [Indic] Minor
Should not be any semantic change.  In preparation for implementing
pre-base reordering Ra.
2012-07-16 15:15:28 -04:00
Behdad Esfahbod 70fe77bb9a Minor 2012-07-16 14:52:18 -04:00
Behdad Esfahbod 2f903215c5 Minor 2012-07-16 13:54:43 -04:00
Behdad Esfahbod a3e04bee2c [Indic] Reorder virama only for old Indic spec 2012-07-16 13:47:19 -04:00
Behdad Esfahbod 0de771b72d [Indic] Categorize Khmer consonants 2012-07-16 13:39:36 -04:00
Behdad Esfahbod d487fff266 Split matras without a Unicode decomposition
This is a hack for now, to get us going with Khmer.  This will be
refactored properly later to move the complex logic into complex
shapers.
2012-07-16 13:25:57 -04:00
Behdad Esfahbod 8aa801a6fd [Indic] Adjust position for split matras
We are going to split matras without a Unicode decompositions in a way
that the second half takes the codepoint of the whole matra.  So,
position them where the second half is supposed to end up.
2012-07-16 13:24:26 -04:00
Behdad Esfahbod 1feb8345a5 [GSUB] Allow 1-to-1 ligature substitutions!
Apparently Uniscribe allows these, and they are used in some Khmer fonts
shipped with Windows, namely, Daun Penh.
2012-07-16 13:23:40 -04:00
Behdad Esfahbod 29f106d7fb [Indic] Apply Above Forms 2012-07-16 12:05:35 -04:00
Behdad Esfahbod fa2bd9fb63 Further simplify atomic ops on Visual Studio 2012-07-14 12:15:54 -04:00
Behdad Esfahbod 0a49235701 Minor 2012-07-13 13:20:49 -04:00
Behdad Esfahbod 11c4ad439e Add -Wcast-align 2012-07-13 11:29:31 -04:00
Behdad Esfahbod a98d0ab186 Make sure HB_BEGIN_DECLS / HB_END_DECLS is only used in public headers
So we can use them to switch default visibility to internal if desired,
and use these to make only declared symbols public.
2012-07-13 10:19:10 -04:00
Behdad Esfahbod 5c5bc96216 Allow overriding HB_BEGIN_DECLS / HB_END_DECLS 2012-07-13 10:15:37 -04:00
Behdad Esfahbod 50a4e78b53 Check for exported weak symbols
Ouch, all our C++ inline functions are being exported (weakly) already.
Fix coming.
2012-07-13 09:48:39 -04:00
Behdad Esfahbod b5aeb95afe Make hb_in_range() static 2012-07-13 09:45:54 -04:00
Behdad Esfahbod 271c8f8907 Minor 2012-07-13 09:32:30 -04:00
Behdad Esfahbod 391f1ff5d8 Fix _InterlockedCompareExchangePointer on x86 2012-07-13 09:04:07 -04:00
Behdad Esfahbod 2023e2b54d [ft] Disable ppem setting
The calculations were wrong.

FreeType makes it really hard to set size and ppem independently.
For now, disable it.  Need to come up with a fix later.
2012-07-11 19:01:26 -04:00
Behdad Esfahbod cdf7444505 [ft] Use unfitted kerning if x_ppem is zero 2012-07-11 18:52:39 -04:00
Behdad Esfahbod 6d08c7f1b3 Revert "Towards templatizing common Lookup types"
This reverts commit 727135f3a9.

This is work-in-progress.  Didn't mean to push it out just yet.
2012-07-11 18:01:27 -04:00
Behdad Esfahbod 552bf3a9f9 Bump WINNT version requested from 500 to 600
Since we use the OpenType versions of Uniscribe functions, we are
relying on that version of the WINNT API.  Otherwise, usp10.h will hide
those symbols.
2012-07-11 18:00:28 -04:00
Behdad Esfahbod 9a5b421a64 Fix build with no Unicode funcs implementations provided 2012-07-11 18:00:28 -04:00
Behdad Esfahbod 727135f3a9 Towards templatizing common Lookup types 2012-07-11 18:00:28 -04:00
Behdad Esfahbod 12f5c0a222 Fix check for Intel atomic ops 2012-06-26 11:16:13 -04:00
Behdad Esfahbod 6932a41fb6 Use octal-escaped UTF-8 characters instead of plain text
https://bugs.freedesktop.org/show_bug.cgi?id=50970
2012-06-26 10:46:31 -04:00
Behdad Esfahbod 8c0ea7bcb4 Disable introspection again
Until I figure out the build issues.  Sigh...
2012-06-24 13:20:56 -04:00
Behdad Esfahbod 49f8e0cd9a GStaticMutex is deprecated 2012-06-16 15:40:03 -04:00
Behdad Esfahbod 1bc1cb3603 Make source more digestable for gobject-introspection 2012-06-16 15:21:55 -04:00
Behdad Esfahbod 84d781e54c Flesh out gobject-introspection stuff a bit 2012-06-16 15:21:41 -04:00
Behdad Esfahbod 2cf301968c Add hb_object_lock/unlock() 2012-06-09 14:58:01 -04:00
Behdad Esfahbod f211d5c291 More Oops! Fix fast-path with sub-type==0 2012-06-09 03:11:22 -04:00
Behdad Esfahbod b1de6aa1f3 Oops! 2012-06-09 03:07:59 -04:00
Behdad Esfahbod b12e2549cb Minor 2012-06-09 03:05:20 -04:00
Behdad Esfahbod faf0f20253 Add sanitize() logic for fast-paths 2012-06-09 03:02:36 -04:00
Behdad Esfahbod 4e766ff28d Add fast-path for GPOS too
Shaves another 3% for DejaVu Sans long Latin strings.
2012-06-09 02:53:57 -04:00
Behdad Esfahbod 993c51915f Add fast-path to GSUB to check coverage
Shaves a good 10% off DejaVu Sans with simple Latin text for me.
Now, DejaVu is very ChainContext-intensive, but it's also a very
popular font!
2012-06-09 02:48:16 -04:00
Behdad Esfahbod f19e0b0099 Match input before backtrack
Makes more sense, optimization-wise.
2012-06-09 02:26:57 -04:00
Behdad Esfahbod 67bb9e8cea Add set add_coverage() to Coverage() 2012-06-09 02:02:46 -04:00
Behdad Esfahbod 4952f0aa5b Minor 2012-06-09 01:43:20 -04:00
Behdad Esfahbod ad6a6f2240 Minor 2012-06-09 01:43:20 -04:00
Behdad Esfahbod 46617a4213 Fix cache implementation 2012-06-09 01:43:20 -04:00
Behdad Esfahbod ce47613889 Micro-optimize
I know...
2012-06-09 01:43:15 -04:00
Behdad Esfahbod 70416de298 Minor 2012-06-09 00:56:41 -04:00
Behdad Esfahbod 99159e52a3 Use linear search for small counts
I see about 8% speedup with long strings with DejaVu Sans.
2012-06-09 00:50:40 -04:00
Behdad Esfahbod caf0412690 Minor 2012-06-09 00:26:32 -04:00
Behdad Esfahbod 0f8fea71a6 Minor. Hide _hb_ot_layout_get_glyph_property() 2012-06-09 00:24:38 -04:00
Behdad Esfahbod 44b8ee0c90 Minor 2012-06-09 00:23:24 -04:00
Behdad Esfahbod 7b84c536c1 In MarkBase attachment, only attach to first of a MultipleSubst sequence
This is apparently what Uniscribe does.  Test case is:

  SEEN FATHA TEH ALEF

with Arabic Typesetting.  Originally reported by Khaled Hosny.
2012-06-08 22:04:23 -04:00
Behdad Esfahbod ec57e0c565 Set lig_comp for MultipleSubst components
To be used for correct mark attachment to first component of a
MultipleSubst output.  That's what Uniscribe does.
2012-06-08 21:47:23 -04:00
Behdad Esfahbod e085fcf7ca Remove unused buffer->replace_glyphs_be16 2012-06-08 21:45:00 -04:00
Behdad Esfahbod 3ec77d6ae0 Don't use replace_glyphs_be for MultipleSubst 2012-06-08 21:44:06 -04:00
Behdad Esfahbod 4b7192125f Minor 2012-06-08 21:41:46 -04:00
Behdad Esfahbod 4508789f4b Add test for static initializers and other C++ stuff 2012-06-08 21:32:43 -04:00
Behdad Esfahbod 56bd259b9a Minor 2012-06-08 21:29:18 -04:00
Behdad Esfahbod bc8357ea7b Merge clusters during normalization 2012-06-08 21:01:20 -04:00
Behdad Esfahbod fe3dabc08d Minor 2012-06-08 20:56:05 -04:00
Behdad Esfahbod e88e14421a Use merge_clusters instead of open-coding 2012-06-08 20:55:21 -04:00
Behdad Esfahbod 330a2af3ff Use merge_clusters when forming Unicode clusters 2012-06-08 20:40:02 -04:00
Behdad Esfahbod bd300df9ad Minor 2012-06-08 20:36:37 -04:00
Behdad Esfahbod e51d2b6ed1 Extend into main buffer if extension hit end of out-buffer merging clusters 2012-06-08 20:36:33 -04:00
Behdad Esfahbod 5ced012d9f Extend end when merging clusters in out-buffer 2012-06-08 20:31:32 -04:00
Behdad Esfahbod 72c0a18783 Extend clusters backward in out-buffer 2012-06-08 20:30:03 -04:00
Behdad Esfahbod cd5891493d Extend clusters backwards, into the out-buffer too 2012-06-08 20:28:59 -04:00
Behdad Esfahbod 77471e0371 Clear output buffer before calling GSUB pause functions 2012-06-08 20:21:02 -04:00
Behdad Esfahbod cafa6f3727 When merging clusters, extend the end 2012-06-08 20:17:10 -04:00
Behdad Esfahbod 28ce5fa454 Merge clusters when ligating 2012-06-08 20:17:06 -04:00
Behdad Esfahbod 2bb1761ccb Minor, use next_glyph() 2012-06-08 19:29:44 -04:00
Behdad Esfahbod 5f68f8675e Minor 2012-06-08 19:23:43 -04:00
Behdad Esfahbod 8729691267 Increase Uniscribe MAX_ITEMS 2012-06-08 14:39:31 -04:00
Behdad Esfahbod dbffa4c83d Fix Uniscribe charset matching
Previously was failing to match fonts that didn't support CHARSET_ANSI.

There still remains a problem with the Uniscribe backend, in that if a
font with the same family name is installed, and is newer, the native
one is preferred over the font we provide.  Fixing it requires rewriting
the name table with a unique family name...
2012-06-08 14:39:31 -04:00
Behdad Esfahbod 82e8bd8628 Remove unused code 2012-06-08 14:39:31 -04:00
Behdad Esfahbod 6da9dbff21 Remove zero-width chars in the fallback shaper too 2012-06-08 10:53:35 -04:00
Behdad Esfahbod 68b76121f8 Fix regressions introduced by sed. Ouch!
Introduced in 99c2695759.
Broken mark-mark and mark-ligature stuff.
2012-06-08 10:47:00 -04:00
Behdad Esfahbod 0dd86f9f68 Whitespace 2012-06-08 10:23:03 -04:00
Behdad Esfahbod 8e7beba7c3 Fix Uniscribe clusters with direction-overriden Arabic 2012-06-08 10:22:06 -04:00
Behdad Esfahbod b069c3c31b Really fix override-direction in Uniscribe 2012-06-08 10:10:29 -04:00
Behdad Esfahbod fcd6f53261 Unbreak Uniscribe
Oops.  hb_tag_t and OPENTYPE_TAG have different endianness.  Perhaps
something to add API for in hb-uniscribe.h
2012-06-08 09:59:43 -04:00
Behdad Esfahbod 29eac8f591 Override direction in Uniscribe backend
Matches OT backend now.
2012-06-08 09:26:17 -04:00
Behdad Esfahbod 1c1233e576 Make Uniscribe backend respect selected script 2012-06-08 09:20:53 -04:00
Behdad Esfahbod 0bb0f5d419 Add note re _NullPool 2012-06-07 17:42:48 -04:00
Behdad Esfahbod 2a3d911fe0 Fix alignment-requirement missmatch
Detected by clang and lots of cmdline options.
2012-06-07 17:31:46 -04:00
Behdad Esfahbod 6095de1635 Fix clang warning with NO_MT path 2012-06-07 15:48:18 -04:00
Behdad Esfahbod a18280a8ce Fix warnings produced by clang analyzer 2012-06-07 15:44:12 -04:00
Behdad Esfahbod 73cb02de2d Minor 2012-06-06 11:29:25 -04:00
Behdad Esfahbod 79e2b4791f Fix ASSERT_POD on clang
As reported by bashi.  Not tested.
2012-06-06 11:27:17 -04:00
Behdad Esfahbod 6220e5fc0d Add ASSERT_POD for most objects 2012-06-06 03:30:09 -04:00
Behdad Esfahbod a00a63b5ef Add macros to check that types are POD 2012-06-06 03:07:01 -04:00
Behdad Esfahbod 61eb60c129 Don't link to libstdc++
New try.
2012-06-05 21:22:36 -04:00
Behdad Esfahbod 81a4b9fd4e Remove unused hb_static_mutex_t 2012-06-05 20:53:00 -04:00
Behdad Esfahbod 4a3a9897b3 Disable Intel atomic ops on mingw32
Apparently the configure test is not enough...
2012-06-05 20:39:07 -04:00
Behdad Esfahbod 0594a24484 Cleanup TRUE/FALSE vs true/false 2012-06-05 20:35:40 -04:00
Behdad Esfahbod e1ac38f8dd Fix inert buffer set_length() with zero
Oops!
2012-06-05 20:31:49 -04:00
Behdad Esfahbod 04bc1eebe7 Add configure tests for Intel atomic intrinsics 2012-06-05 20:16:56 -04:00
Behdad Esfahbod f64b2ebf82 Remove last static initializer
We're free!  Lazy or immediate...
2012-06-05 20:15:27 -04:00
Behdad Esfahbod 04aed572f1 Make hb-ft static-initializer free 2012-06-05 18:45:36 -04:00
Behdad Esfahbod be4560a3b5 Undo default unicode-funcs to avoid static initializer again 2012-06-05 18:43:57 -04:00
Behdad Esfahbod 093171ccec Implement lock-free hb_language_t
Another static-initialization down.  One more to go.
2012-06-05 18:00:45 -04:00
Behdad Esfahbod 6843ce01be Add atomic-pointer functions
Gonig to use these for lock-free linked-lists, to be used for
hb_language_t among other things.
2012-06-05 17:27:20 -04:00
Behdad Esfahbod cdafe3a7d8 Add gcc intrinsics implementations for atomic and mutex 2012-06-05 16:40:23 -04:00
Behdad Esfahbod d970d2899b Add gcc implementation for atomic ops 2012-06-05 16:06:28 -04:00
Behdad Esfahbod 0e253e97af Add a mutex to object header
Removes one more static-initialization.  A few more to go.
2012-06-05 15:54:43 -04:00
Behdad Esfahbod a2b471df82 Remove static initializers from indic 2012-06-05 15:17:44 -04:00
Behdad Esfahbod f06ab8a426 Better hide nil objects and make them const 2012-06-05 14:49:14 -04:00
Behdad Esfahbod bf93b636c4 Remove constructor from hb_prealloced_array_t
This was causing all object types to be non-POD and have static
initializers.  We don't need that!

Now, most nil objects just moved from .bss to .data.  Fixing for that
coming soon.
2012-06-05 14:17:32 -04:00
Behdad Esfahbod f1971a2174 Fix warnings 2012-06-05 14:06:04 -04:00
Behdad Esfahbod 9fc7a11469 Remove comma at the end of enum
As reported by Jonathan Kew on the list.
2012-06-04 08:28:19 -04:00
Behdad Esfahbod 3b8fd9c48f Remove const from ref_count.ref_count
According to Tom Hacohen this was breaking build with some compilers.

In file included from hb-buffer-private.hh:35:0,
                 from hb-ot-map-private.hh:32,
                 from hb-ot-shape-private.hh:32,
                 from hb-ot-shape.cc:29:
hb-object-private.hh: In constructor '_hb_object_header_t::_hb_object_header_t()':
hb-object-private.hh:97:8: error: uninitialized const member in 'struct hb_reference_count_t'
hb-object-private.hh:51:25: note: 'hb_reference_count_t::ref_count' should be initialized
In file included from hb-ot-shape.cc:33:0:
hb-set-private.hh: In constructor '_hb_set_t::_hb_set_t()':
hb-set-private.hh:37:8: note: synthesized method '_hb_object_header_t::_hb_object_header_t()' first required here
hb-ot-shape.cc: In function 'void hb_ot_shape_glyphs_closure(hb_font_t*, hb_buffer_t*, const hb_feature_t*, unsigned int, hb_set_t*)':
hb-ot-shape.cc:521:12: note: synthesized method '_hb_set_t::_hb_set_t()' first required here
2012-06-03 15:54:19 -04:00
Behdad Esfahbod 70600dbf62 Minor 2012-06-03 15:52:51 -04:00
Behdad Esfahbod 96a9ef0c9f Remove tab character like other "zero-width" characters
Uniscribe does that, this make comparing results to Uniscribe
easier.
2012-06-01 13:46:26 -04:00
Behdad Esfahbod 0558d55bac Remove hb_atomic_int_set/get()
We never use them in fact...

I'm just adjusting these as I better understand the requirements of
the code and the guarantees of each operation.
2012-05-28 10:46:47 -04:00
Behdad Esfahbod bce095524b Add hb_font_get_glyph_name() and hb_font_get_glyph_from_name() 2012-05-28 10:45:50 -04:00
Behdad Esfahbod bc145658bd Warn if no Unicode functions implementation is found 2012-05-28 10:45:50 -04:00
Behdad Esfahbod a3547330fa Cleanup atomic ops on OS X 2012-05-27 10:20:47 -04:00
Behdad Esfahbod e4b6d503c5 Don't use atomic ops in hb_cache_t
We don't care about linearizability, so unprotected int read/write
are enough, no need for expensive memory barriers.  It's a cache,
that's all.
2012-05-27 10:11:13 -04:00
Behdad Esfahbod 819faa0530 Minor 2012-05-27 10:09:18 -04:00
Behdad Esfahbod 303d5850ec Fix Windows atomic get/set
According to:
http://msdn.microsoft.com/en-us/library/65tt87y8.aspx

MemoryBarrier() is the right macro to protect these, not _ReadBarrier()
and/or _WriteBarrier().
2012-05-27 10:01:13 -04:00
Behdad Esfahbod 29ce446d31 Add set iterator 2012-05-25 14:17:54 -04:00
Behdad Esfahbod 62c3e111fc Add set symmetric difference 2012-05-25 13:48:00 -04:00
Behdad Esfahbod 27aba594c9 Minor 2012-05-24 15:00:01 -04:00
Behdad Esfahbod cde1c0114b Fix hb_atomic_int_set() implementation for HB_NO_MT
As pointed out by Jonathan Kew.
2012-05-24 10:46:39 -04:00
Behdad Esfahbod ed2f1363a3 Fix substitution glyph class propagation
The old code was doing nothing.

Still got to find an example font+string that makes this matter, but
need this for fixing synthetic GDEF anyway.
2012-05-22 22:12:22 -04:00
Behdad Esfahbod 20fdb0f41d Add a lock-free cache type for int->int functions
To be used for cmap and advance caching if desired.
2012-05-17 22:04:45 -04:00
Behdad Esfahbod bd908b4f10 Implement hb_atomic_int_set() for OS X 2012-05-17 22:02:08 -04:00
Behdad Esfahbod 022a05ae90 Minor 2012-05-17 21:53:24 -04:00
Behdad Esfahbod 22afd66a30 Add hb_atomic_int_set() again 2012-05-17 21:23:49 -04:00
Behdad Esfahbod 4aa7258cb1 Fix type conflicts on Windows without glib 2012-05-17 21:01:04 -04:00
Behdad Esfahbod f039e79d54 Don't use min/max as function names
They can be macros on some systems.  Eg. mingw32.
2012-05-17 20:55:12 -04:00
Behdad Esfahbod 34961e3198 Prefer native atomic/mutex ops to glib's 2012-05-17 20:50:38 -04:00
Behdad Esfahbod ec3ba4b96f Move atomic ops into their own header 2012-05-17 20:30:46 -04:00
Behdad Esfahbod 1d6846db9e [Indic] Apply vatu feature after cjct
Testing with old Deva spec this reduces failures.
Test sequence: U+0915,U+094D,U+0930.
2012-05-13 18:09:29 +02:00
Behdad Esfahbod 617f4ac46f Refactor 2012-05-13 16:48:03 +02:00
Behdad Esfahbod 5e4e21fce4 Revert "[Indic] Refactoring"
This reverts commit 0831061efb.
2012-05-13 16:46:08 +02:00
Behdad Esfahbod 3f18236a03 Fix more warnings 2012-05-13 16:20:10 +02:00
Behdad Esfahbod 9f377ed321 Fix more unused-var warnings 2012-05-13 16:13:44 +02:00
Behdad Esfahbod d993e72331 Fix hb_face_set_index() 2012-05-13 16:04:36 +02:00
Behdad Esfahbod 93345edcbe Fix warnings 2012-05-13 16:01:08 +02:00
Behdad Esfahbod eace47b173 Minor 2012-05-13 15:54:43 +02:00
Behdad Esfahbod 99c2695759 Add accessort to buffer for current info, current pos, and prev info 2012-05-13 15:45:18 +02:00
Behdad Esfahbod 6736f3c5b0 Minor 2012-05-13 15:21:06 +02:00
Behdad Esfahbod 5df809b655 [GSUB/GPOS] Remove context_length
The spec doesn't say contextual matching should be done this way,
and AOTS doesn't do it either.  It was inherited from old HarfBuzz.
Remove it.
2012-05-13 15:17:51 +02:00
Behdad Esfahbod 28b9d502bb Minor 2012-05-13 15:04:00 +02:00
Behdad Esfahbod 737dded2e0 Fix compiler warnings 2012-05-12 15:40:11 +02:00
Behdad Esfahbod 7f852b644b Fix compiler warnings 2012-05-11 23:10:31 +02:00
Behdad Esfahbod f7e8dcfd4f [Indic] Unbreak Devanagari
And this, concludes the HarfBuzz Massala Hackfest.

I like to specially thank Jonathan Kew for doing all the decription and
letting me get commit points.
2012-05-11 22:01:33 +02:00
Behdad Esfahbod 6a091df9b4 [Indic] Disambiguate sub vs post vs above matras
Bengali is at *just* above 5% now.
2012-05-11 21:42:27 +02:00
Behdad Esfahbod 9d0d319a4a [Indic] Position Bengali Reph before matras 2012-05-11 21:36:32 +02:00
Behdad Esfahbod f893672511 [Indic] Start categorizing Reph per script 2012-05-11 21:10:03 +02:00
Behdad Esfahbod a913b024d8 [Indic] Apply 'init' feature for Bengali
Error down from 20% to 7%.
2012-05-11 20:59:26 +02:00
Behdad Esfahbod eed903b164 [Indic] Refactor for the arrival of 'init' feature
Yep, on Bengali now!
2012-05-11 20:50:53 +02:00
Behdad Esfahbod 18c06e189b [Indic] Add Uniscribe bug feature for dotted circle
For dotted-circle independent clusters, Uniscribe does no Reph shaping
for the exact sequence Ra+Halant+25CC.  Which also is the only possible
sequence with 25CC at the end.
2012-05-11 20:02:14 +02:00
Behdad Esfahbod 0831061efb [Indic] Refactoring 2012-05-11 19:07:58 +02:00
Behdad Esfahbod 7ea58db311 Minor 2012-05-11 18:58:57 +02:00
Behdad Esfahbod 9c09928989 [Indic] Allow multiple Consonants in Vowel/NBSP syllables
Uniscribe allows multiple Halant+Consonant after a Vowel.
Tests:
↦       * U+0905,U+094D,U+092B,U+094D,930,94d,930
2012-05-11 18:46:35 +02:00
Behdad Esfahbod 8c0aa486f3 [Indic] Allow two Nuktas per consonant
Uniscribe allows up to two nuktas per consonant and one per matra. It does so
indepent of whether the consonant already has a nukta in it.  Tests:

        * U+0916,U+093C,U+0941
        * U+0959,U+093C,U+0941
        * U+0916,U+093C,U+093C,U+0941
        * U+0959,U+093C,U+093C,U+0941
        * U+0916,U+093C,U+093C,U+093C,U+0941
        * U+0959,U+093C,U+093C,U+093C,U+0941
        * 915,93c,93c,,94d,U+0916,U+093C,U+093C,U+093e,93c,93c
2012-05-11 18:13:42 +02:00
Behdad Esfahbod 3399a06e70 [Indic] Fix U+0952 and similar classification to match Uniscribe
See comments.
2012-05-11 17:54:26 +02:00
Behdad Esfahbod 11aa3ef18d [Indic] Treat U+0951..U+0954 all similar to U+0952 2012-05-11 17:30:48 +02:00
Behdad Esfahbod 5f131d3226 [GSUB/GPOS/Indic] Apply GSUB/GPOS within syllables only
This does not apply to the context matchings.

This regresses tests right now.  And we are not sure whether this is
the right thing to do for GPOS.  But we'll figure out.
2012-05-11 17:29:40 +02:00
Behdad Esfahbod 8fd83aaf6e [GSUB/GPOS] Fix wrong buffer access in backward skippy mask matching 2012-05-11 17:18:37 +02:00
Behdad Esfahbod ff24d1081a [Indic] Don't use syllable serial value 0 2012-05-11 17:07:08 +02:00
Behdad Esfahbod 892eb78782 [Indic] Implement Uniscribe Reph+Matra+Halant bug feature 2012-05-11 16:54:40 +02:00
Behdad Esfahbod 67ea29af49 [Indic] Add example of different Uniscribe behavior 2012-05-11 16:51:23 +02:00
Behdad Esfahbod ebe29733d4 [Indic] Add runtime Uniscribe bug compatibility mode!
Enable by setting envvar:

  HB_OT_INDIC_OPTIONS=uniscribe-bug-compatible

Plus, LeftMatra+Halant "feature".
2012-05-11 16:43:12 +02:00
Behdad Esfahbod 616e692e29 [Indic] Add #define UNISCRIBE_BUG_COMPATIBLE 1 2012-05-11 16:25:02 +02:00
Behdad Esfahbod 6782bdae3b [Indic] Fix Left Matra + Halant reordering
As can be seen in: U+092B,U+093F,U+094D
2012-05-11 16:23:43 +02:00
Behdad Esfahbod 3c2ea9481b Minor 2012-05-11 16:23:38 +02:00
Behdad Esfahbod 203d71069c [GSUB/GPOS] Check all glyph masks when matching input 2012-05-11 16:01:44 +02:00
Behdad Esfahbod 668c6046c1 [Indic] Apply Reph mask to all POS_REPH glyphs
Needed for upcoming changes to GSUB/GPOS mask matching.
2012-05-11 15:34:13 +02:00
Behdad Esfahbod 4be46bade2 [Indic] Fix state machine to backtrack 2012-05-11 14:39:01 +02:00
Behdad Esfahbod cee7187447 [Indic] Move syllable tracking from Indic to generic layer
This is to incorporate it into GSUB/GPOS processing.
2012-05-11 11:41:39 +02:00
Behdad Esfahbod 3bf27a9f0e [Indic] Disable conjuncts when a ZWJ happens
Not that the code makes any difference since the presence of ZWJ itself
causes the ligature to fail to match anyway.
2012-05-11 11:17:23 +02:00
Behdad Esfahbod c6d904d67d [Indic] Fix bitops typo!
Another 1000 down!
2012-05-11 11:07:40 +02:00
Behdad Esfahbod 55fe2cf79b Make APPLY debug output print current index and codepoint
Yay!
2012-05-11 03:56:33 +02:00
Behdad Esfahbod 7bd2b04fea Minor 2012-05-11 03:40:58 +02:00
Behdad Esfahbod cf26510dbb Some more...
Done.  I promise.
2012-05-11 03:35:08 +02:00
Behdad Esfahbod 9659523ca3 More beauty in debug output! 2012-05-11 03:33:36 +02:00
Behdad Esfahbod cf26e88a5a Finish off debug output beautification 2012-05-11 03:16:57 +02:00
Behdad Esfahbod d7bba01a35 Only print class name in debug output if there's one available 2012-05-11 02:46:26 +02:00
Behdad Esfahbod 85f73fa8da Only printout class name in tracing, if one is available
Makes debug output much more pleasant.
2012-05-11 02:40:42 +02:00
Behdad Esfahbod 98619ce4fa Minor 2012-05-11 02:34:06 +02:00
Behdad Esfahbod acea183e98 Add return annotation for APPLY 2012-05-11 02:33:11 +02:00
Behdad Esfahbod 5ccfe8e215 /Minor/ 2012-05-11 02:19:41 +02:00
Behdad Esfahbod 0ab8c86217 Annotate SANITIZE return values
More to come, for APPLY, CLOSURE, etc.
2012-05-11 02:11:52 +02:00
Behdad Esfahbod 829e814ff3 Minor 2012-05-11 00:52:16 +02:00
Behdad Esfahbod 6eec6f406d Code reshuffling 2012-05-11 00:50:38 +02:00
Behdad Esfahbod 1e08830b4f Beautify debug output 2012-05-11 00:43:57 +02:00
Behdad Esfahbod 6f45538017 More massaging trace messaging 2012-05-10 23:24:43 +02:00
Behdad Esfahbod b5fa37cb69 Minor 2012-05-10 23:09:48 +02:00
Behdad Esfahbod 208109703c Better trace message support infrastructure
We have varargs in the trace interface now.  To be used soon...
2012-05-10 23:06:58 +02:00
Behdad Esfahbod 02b2922fbf [Indic] Towards better Reph positioning
Fixed for Deva cases with two full-form consonants.  Failures **way** down.
Not much left to go :-).
2012-05-10 21:44:50 +02:00
Behdad Esfahbod 74e54cf446 [Indic] Add Ra back for scripts without Reph
We now check that the 'rphp' table exists before forming Reph, so
we don't need to comment out Ra for those scripts.
2012-05-10 21:22:58 +02:00
Behdad Esfahbod 2b70df5cc0 [Indic] Add note re Uniscribe clusters 2012-05-10 18:38:22 +02:00
Behdad Esfahbod 21d2803133 [Indic] Do clustering like Uniscribe does
Hindi Wikipedia failures down to 6639 (0.938381%)!
2012-05-10 18:34:34 +02:00
Behdad Esfahbod 8df5636968 [Indic] Reorder Reph to before the Halant after Matras
Uniscribe doesn't do it, but we want to do as it gives the Reph the
opportunity to interact with the Matras.  Test with mangal for example.
Sequence: <0930,094d,0915,094b,094d>
In test suite already.
2012-05-10 15:41:04 +02:00
Behdad Esfahbod daf3234bdc [Indic] Don't clear the mask for Reph
This was removing the mandatory global 1 bit in the mask and hence
disabling GPOS for Reph!
2012-05-10 15:28:27 +02:00
Behdad Esfahbod 7708ee23cb [Indic] Improve Left Matra repositioning
Move its dependents too.
2012-05-10 14:48:25 +02:00