2013-02-11 19:36:23 +01:00
|
|
|
/*
|
|
|
|
* Copyright © 2011,2012,2013 Google, Inc.
|
|
|
|
*
|
|
|
|
* This is part of HarfBuzz, a text shaping library.
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, without written agreement and without
|
|
|
|
* license or royalty fees, to use, copy, modify, and distribute this
|
|
|
|
* software and its documentation for any purpose, provided that the
|
|
|
|
* above copyright notice and the following two paragraphs appear in
|
|
|
|
* all copies of this software.
|
|
|
|
*
|
|
|
|
* IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
|
|
|
|
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
|
|
|
|
* ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
|
|
|
|
* IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
|
|
|
|
* DAMAGE.
|
|
|
|
*
|
|
|
|
* THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
|
|
|
|
* BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
|
|
|
|
* ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
|
|
|
|
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
|
|
|
|
*
|
|
|
|
* Google Author(s): Behdad Esfahbod
|
|
|
|
*/
|
|
|
|
|
2013-02-12 22:13:56 +01:00
|
|
|
#include "hb-ot-shape-complex-indic-private.hh"
|
2013-02-11 19:36:23 +01:00
|
|
|
|
|
|
|
/* buffer var allocations */
|
|
|
|
#define myanmar_category() complex_var_u8_0() /* myanmar_category_t */
|
2013-02-12 18:14:10 +01:00
|
|
|
#define myanmar_position() complex_var_u8_1() /* myanmar_position_t */
|
2013-02-11 19:36:23 +01:00
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Myanmar shaper.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static const hb_tag_t
|
2013-02-15 12:22:26 +01:00
|
|
|
basic_features[] =
|
2013-02-11 19:36:23 +01:00
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Basic features.
|
|
|
|
* These features are applied in order, one at a time, after initial_reordering.
|
|
|
|
*/
|
|
|
|
HB_TAG('r','p','h','f'),
|
|
|
|
HB_TAG('p','r','e','f'),
|
|
|
|
HB_TAG('b','l','w','f'),
|
|
|
|
HB_TAG('p','s','t','f'),
|
2013-02-15 12:22:26 +01:00
|
|
|
};
|
|
|
|
static const hb_tag_t
|
|
|
|
other_features[] =
|
|
|
|
{
|
2013-02-11 19:36:23 +01:00
|
|
|
/*
|
|
|
|
* Other features.
|
|
|
|
* These features are applied all at once, after final_reordering.
|
|
|
|
*/
|
|
|
|
HB_TAG('p','r','e','s'),
|
|
|
|
HB_TAG('a','b','v','s'),
|
|
|
|
HB_TAG('b','l','w','s'),
|
|
|
|
HB_TAG('p','s','t','s'),
|
|
|
|
/* Positioning features, though we don't care about the types. */
|
|
|
|
HB_TAG('d','i','s','t'),
|
2013-10-15 12:28:23 +02:00
|
|
|
/* Pre-release version of Windows 8 Myanmar font had abvm,blwm
|
|
|
|
* features. The released Windows 8 version of the font (as well
|
|
|
|
* as the released spec) used 'mark' instead. The Windows 8
|
|
|
|
* shaper however didn't apply 'mark' but did apply 'mkmk'.
|
|
|
|
* Perhaps it applied abvm/blwm. This was fixed in a Windows 8
|
|
|
|
* update, so now it applies mark/mkmk. We are guessing that
|
|
|
|
* it still applies abvm/blwm too.
|
|
|
|
*/
|
|
|
|
HB_TAG('a','b','v','m'),
|
|
|
|
HB_TAG('b','l','w','m'),
|
2013-02-11 19:36:23 +01:00
|
|
|
};
|
|
|
|
|
|
|
|
static void
|
|
|
|
setup_syllables (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_font_t *font,
|
|
|
|
hb_buffer_t *buffer);
|
|
|
|
static void
|
|
|
|
initial_reordering (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_font_t *font,
|
|
|
|
hb_buffer_t *buffer);
|
|
|
|
static void
|
|
|
|
final_reordering (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_font_t *font,
|
|
|
|
hb_buffer_t *buffer);
|
|
|
|
|
|
|
|
static void
|
|
|
|
collect_features_myanmar (hb_ot_shape_planner_t *plan)
|
|
|
|
{
|
|
|
|
hb_ot_map_builder_t *map = &plan->map;
|
|
|
|
|
|
|
|
/* Do this before any lookups have been applied. */
|
|
|
|
map->add_gsub_pause (setup_syllables);
|
|
|
|
|
2013-02-14 17:05:56 +01:00
|
|
|
map->add_global_bool_feature (HB_TAG('l','o','c','l'));
|
2013-02-11 19:36:23 +01:00
|
|
|
/* The Indic specs do not require ccmp, but we apply it here since if
|
|
|
|
* there is a use of it, it's typically at the beginning. */
|
2013-02-14 17:05:56 +01:00
|
|
|
map->add_global_bool_feature (HB_TAG('c','c','m','p'));
|
2013-02-11 19:36:23 +01:00
|
|
|
|
|
|
|
|
|
|
|
map->add_gsub_pause (initial_reordering);
|
2013-02-15 12:22:26 +01:00
|
|
|
for (unsigned int i = 0; i < ARRAY_LENGTH (basic_features); i++)
|
|
|
|
{
|
[Indic] Futher adjust ZWJ handling in Indic-like shapers
After the Ngapi hackfest work, we were assuming that fonts
won't use presentation features to choose specific forms
(eg. conjuncts). As such, we were using auto-joiner behavior
for such features. It proved to be troublesome as many fonts
used presentation forms ('pres') for example to form conjuncts,
which need to be disabled when a ZWJ is inserted.
Two examples:
U+0D2F,U+200D,U+0D4D,U+0D2F with kartika.ttf
U+0995,U+09CD,U+200D,U+09B7 with vrinda.ttf
What we do now is to never do magic to ZWJ during GSUB's main input
match for Indic-style shapers. Note that backtrack/lookahead are still
matched liberally, as is GPOS. This seems to be an acceptable
compromise.
As to the bug that initially started this work, that one needs to
be fixed differently:
Bug 58714 - Kannada u+0cb0 u+200d u+0ccd u+0c95 u+0cbe does not
provide same results as Windows8
https://bugs.freedesktop.org/show_bug.cgi?id=58714
New numbers:
BENGALI: 353689 out of 354188 tests passed. 499 failed (0.140886%)
DEVANAGARI: 707305 out of 707394 tests passed. 89 failed (0.0125814%)
GUJARATI: 366349 out of 366457 tests passed. 108 failed (0.0294714%)
GURMUKHI: 60706 out of 60747 tests passed. 41 failed (0.067493%)
KANNADA: 951030 out of 951913 tests passed. 883 failed (0.0927606%)
KHMER: 299070 out of 299124 tests passed. 54 failed (0.0180527%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1048102 out of 1048334 tests passed. 232 failed (0.0221304%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271666 out of 271847 tests passed. 181 failed (0.0665816%)
TAMIL: 1091753 out of 1091754 tests passed. 1 failed (9.15957e-05%)
TELUGU: 970555 out of 970573 tests passed. 18 failed (0.00185457%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2013-03-19 10:53:26 +01:00
|
|
|
map->add_feature (basic_features[i], 1, F_GLOBAL | F_MANUAL_ZWJ);
|
2017-10-15 12:11:08 +02:00
|
|
|
map->add_gsub_pause (nullptr);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
map->add_gsub_pause (final_reordering);
|
2013-02-15 12:22:26 +01:00
|
|
|
for (unsigned int i = 0; i < ARRAY_LENGTH (other_features); i++)
|
[Indic] Futher adjust ZWJ handling in Indic-like shapers
After the Ngapi hackfest work, we were assuming that fonts
won't use presentation features to choose specific forms
(eg. conjuncts). As such, we were using auto-joiner behavior
for such features. It proved to be troublesome as many fonts
used presentation forms ('pres') for example to form conjuncts,
which need to be disabled when a ZWJ is inserted.
Two examples:
U+0D2F,U+200D,U+0D4D,U+0D2F with kartika.ttf
U+0995,U+09CD,U+200D,U+09B7 with vrinda.ttf
What we do now is to never do magic to ZWJ during GSUB's main input
match for Indic-style shapers. Note that backtrack/lookahead are still
matched liberally, as is GPOS. This seems to be an acceptable
compromise.
As to the bug that initially started this work, that one needs to
be fixed differently:
Bug 58714 - Kannada u+0cb0 u+200d u+0ccd u+0c95 u+0cbe does not
provide same results as Windows8
https://bugs.freedesktop.org/show_bug.cgi?id=58714
New numbers:
BENGALI: 353689 out of 354188 tests passed. 499 failed (0.140886%)
DEVANAGARI: 707305 out of 707394 tests passed. 89 failed (0.0125814%)
GUJARATI: 366349 out of 366457 tests passed. 108 failed (0.0294714%)
GURMUKHI: 60706 out of 60747 tests passed. 41 failed (0.067493%)
KANNADA: 951030 out of 951913 tests passed. 883 failed (0.0927606%)
KHMER: 299070 out of 299124 tests passed. 54 failed (0.0180527%)
LAO: 53611 out of 53644 tests passed. 33 failed (0.0615167%)
MALAYALAM: 1048102 out of 1048334 tests passed. 232 failed (0.0221304%)
ORIYA: 42320 out of 42329 tests passed. 9 failed (0.021262%)
SINHALA: 271666 out of 271847 tests passed. 181 failed (0.0665816%)
TAMIL: 1091753 out of 1091754 tests passed. 1 failed (9.15957e-05%)
TELUGU: 970555 out of 970573 tests passed. 18 failed (0.00185457%)
TIBETAN: 208469 out of 208469 tests passed. 0 failed (0%)
2013-03-19 10:53:26 +01:00
|
|
|
map->add_feature (other_features[i], 1, F_GLOBAL | F_MANUAL_ZWJ);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
override_features_myanmar (hb_ot_shape_planner_t *plan)
|
|
|
|
{
|
2013-02-14 17:25:10 +01:00
|
|
|
plan->map.add_feature (HB_TAG('l','i','g','a'), 0, F_GLOBAL);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
enum syllable_type_t {
|
|
|
|
consonant_syllable,
|
2013-11-26 00:10:38 +01:00
|
|
|
punctuation_cluster,
|
2013-02-11 19:36:23 +01:00
|
|
|
broken_cluster,
|
|
|
|
non_myanmar_cluster,
|
|
|
|
};
|
|
|
|
|
|
|
|
#include "hb-ot-shape-complex-myanmar-machine.hh"
|
|
|
|
|
|
|
|
|
|
|
|
/* Note: This enum is duplicated in the -machine.rl source file.
|
|
|
|
* Not sure how to avoid duplication. */
|
|
|
|
enum myanmar_category_t {
|
2017-10-03 15:20:07 +02:00
|
|
|
OT_As = 18, /* Asat */
|
2013-02-11 19:36:23 +01:00
|
|
|
OT_D0 = 20, /* Digit zero */
|
|
|
|
OT_DB = OT_N, /* Dot below */
|
2014-05-27 23:53:37 +02:00
|
|
|
OT_GB = OT_PLACEHOLDER,
|
2013-02-11 19:36:23 +01:00
|
|
|
OT_MH = 21, /* Various consonant medial types */
|
|
|
|
OT_MR = 22, /* Various consonant medial types */
|
|
|
|
OT_MW = 23, /* Various consonant medial types */
|
|
|
|
OT_MY = 24, /* Various consonant medial types */
|
|
|
|
OT_PT = 25, /* Pwo and other tones */
|
|
|
|
OT_VAbv = 26,
|
|
|
|
OT_VBlw = 27,
|
|
|
|
OT_VPre = 28,
|
|
|
|
OT_VPst = 29,
|
2013-11-26 00:10:38 +01:00
|
|
|
OT_VS = 30, /* Variation selectors */
|
2017-10-03 15:20:07 +02:00
|
|
|
OT_P = 31, /* Punctuation */
|
|
|
|
OT_D = 32, /* Digits except zero */
|
2013-02-11 19:36:23 +01:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
static inline bool
|
|
|
|
is_one_of (const hb_glyph_info_t &info, unsigned int flags)
|
|
|
|
{
|
|
|
|
/* If it ligated, all bets are off. */
|
2013-10-18 01:09:08 +02:00
|
|
|
if (_hb_glyph_info_ligated (&info)) return false;
|
2015-07-21 16:50:02 +02:00
|
|
|
return !!(FLAG_SAFE (info.myanmar_category()) & flags);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool
|
|
|
|
is_consonant (const hb_glyph_info_t &info)
|
|
|
|
{
|
|
|
|
return is_one_of (info, CONSONANT_FLAGS);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static inline void
|
|
|
|
set_myanmar_properties (hb_glyph_info_t &info)
|
|
|
|
{
|
|
|
|
hb_codepoint_t u = info.codepoint;
|
2013-02-12 22:13:56 +01:00
|
|
|
unsigned int type = hb_indic_get_categories (u);
|
2014-07-11 20:54:42 +02:00
|
|
|
indic_category_t cat = (indic_category_t) (type & 0x7Fu);
|
2013-02-12 22:13:56 +01:00
|
|
|
indic_position_t pos = (indic_position_t) (type >> 8);
|
2013-02-11 19:36:23 +01:00
|
|
|
|
|
|
|
/* Myanmar
|
|
|
|
* http://www.microsoft.com/typography/OpenTypeDev/myanmar/intro.htm#analyze
|
|
|
|
*/
|
2017-06-19 12:17:09 +02:00
|
|
|
if (unlikely (hb_in_range<hb_codepoint_t> (u, 0xFE00u, 0xFE0Fu)))
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_VS;
|
|
|
|
|
2013-02-11 19:36:23 +01:00
|
|
|
switch (u)
|
|
|
|
{
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x104Eu:
|
2013-11-26 00:03:34 +01:00
|
|
|
cat = (indic_category_t) OT_C; /* The spec says C, IndicSyllableCategory doesn't have. */
|
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x002Du: case 0x00A0u: case 0x00D7u: case 0x2012u:
|
|
|
|
case 0x2013u: case 0x2014u: case 0x2015u: case 0x2022u:
|
|
|
|
case 0x25CCu: case 0x25FBu: case 0x25FCu: case 0x25FDu:
|
|
|
|
case 0x25FEu:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_GB;
|
2013-02-11 22:06:02 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1004u: case 0x101Bu: case 0x105Au:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_Ra;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1032u: case 0x1036u:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_A;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2015-12-18 12:05:11 +01:00
|
|
|
case 0x1039u:
|
|
|
|
cat = (indic_category_t) OT_H;
|
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x103Au:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_As;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1041u: case 0x1042u: case 0x1043u: case 0x1044u:
|
|
|
|
case 0x1045u: case 0x1046u: case 0x1047u: case 0x1048u:
|
|
|
|
case 0x1049u: case 0x1090u: case 0x1091u: case 0x1092u:
|
|
|
|
case 0x1093u: case 0x1094u: case 0x1095u: case 0x1096u:
|
|
|
|
case 0x1097u: case 0x1098u: case 0x1099u:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_D;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1040u:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_D; /* XXX The spec says D0, but Uniscribe doesn't seem to do. */
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x103Eu: case 0x1060u:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_MH;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x103Cu:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_MR;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x103Du: case 0x1082u:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_MW;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x103Bu: case 0x105Eu: case 0x105Fu:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_MY;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1063u: case 0x1064u: case 0x1069u: case 0x106Au:
|
|
|
|
case 0x106Bu: case 0x106Cu: case 0x106Du: case 0xAA7Bu:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_PT;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x1038u: case 0x1087u: case 0x1088u: case 0x1089u:
|
|
|
|
case 0x108Au: case 0x108Bu: case 0x108Cu: case 0x108Du:
|
|
|
|
case 0x108Fu: case 0x109Au: case 0x109Bu: case 0x109Cu:
|
2013-02-12 22:13:56 +01:00
|
|
|
cat = (indic_category_t) OT_SM;
|
2013-02-11 19:36:23 +01:00
|
|
|
break;
|
2013-11-26 00:10:38 +01:00
|
|
|
|
2014-07-11 20:54:42 +02:00
|
|
|
case 0x104Au: case 0x104Bu:
|
2013-11-26 00:10:38 +01:00
|
|
|
cat = (indic_category_t) OT_P;
|
|
|
|
break;
|
2016-05-06 18:56:07 +02:00
|
|
|
|
|
|
|
case 0xAA74u: case 0xAA75u: case 0xAA76u:
|
|
|
|
/* https://github.com/roozbehp/unicode-data/issues/3 */
|
|
|
|
cat = (indic_category_t) OT_C;
|
|
|
|
break;
|
2013-02-12 22:13:56 +01:00
|
|
|
}
|
2013-02-11 19:36:23 +01:00
|
|
|
|
2013-02-12 22:13:56 +01:00
|
|
|
if (cat == OT_M)
|
|
|
|
{
|
|
|
|
switch ((int) pos)
|
|
|
|
{
|
|
|
|
case POS_PRE_C: cat = (indic_category_t) OT_VPre;
|
|
|
|
pos = POS_PRE_M; break;
|
|
|
|
case POS_ABOVE_C: cat = (indic_category_t) OT_VAbv; break;
|
|
|
|
case POS_BELOW_C: cat = (indic_category_t) OT_VBlw; break;
|
|
|
|
case POS_POST_C: cat = (indic_category_t) OT_VPst; break;
|
|
|
|
}
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
2013-02-12 22:13:56 +01:00
|
|
|
info.myanmar_category() = (myanmar_category_t) cat;
|
2013-02-11 19:36:23 +01:00
|
|
|
info.myanmar_position() = pos;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
setup_masks_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
|
|
|
|
hb_buffer_t *buffer,
|
|
|
|
hb_font_t *font HB_UNUSED)
|
|
|
|
{
|
|
|
|
HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_category);
|
|
|
|
HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_position);
|
|
|
|
|
|
|
|
/* We cannot setup masks here. We save information about characters
|
|
|
|
* and setup masks later on in a pause-callback. */
|
|
|
|
|
|
|
|
unsigned int count = buffer->len;
|
2014-07-17 20:22:11 +02:00
|
|
|
hb_glyph_info_t *info = buffer->info;
|
2013-02-11 19:36:23 +01:00
|
|
|
for (unsigned int i = 0; i < count; i++)
|
2014-07-17 20:22:11 +02:00
|
|
|
set_myanmar_properties (info[i]);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
setup_syllables (const hb_ot_shape_plan_t *plan HB_UNUSED,
|
|
|
|
hb_font_t *font HB_UNUSED,
|
|
|
|
hb_buffer_t *buffer)
|
|
|
|
{
|
|
|
|
find_syllables (buffer);
|
2017-08-11 03:45:33 +02:00
|
|
|
foreach_syllable (buffer, start, end)
|
|
|
|
buffer->unsafe_to_break (start, end);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
compare_myanmar_order (const hb_glyph_info_t *pa, const hb_glyph_info_t *pb)
|
|
|
|
{
|
|
|
|
int a = pa->myanmar_position();
|
|
|
|
int b = pb->myanmar_position();
|
|
|
|
|
|
|
|
return a < b ? -1 : a == b ? 0 : +1;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* Rules from:
|
|
|
|
* http://www.microsoft.com/typography/OpenTypeDev/myanmar/intro.htm */
|
|
|
|
|
|
|
|
static void
|
2015-07-22 12:58:11 +02:00
|
|
|
initial_reordering_consonant_syllable (hb_buffer_t *buffer,
|
2013-02-11 19:36:23 +01:00
|
|
|
unsigned int start, unsigned int end)
|
|
|
|
{
|
|
|
|
hb_glyph_info_t *info = buffer->info;
|
|
|
|
|
|
|
|
unsigned int base = end;
|
|
|
|
bool has_reph = false;
|
|
|
|
|
|
|
|
{
|
|
|
|
unsigned int limit = start;
|
|
|
|
if (start + 3 <= end &&
|
|
|
|
info[start ].myanmar_category() == OT_Ra &&
|
|
|
|
info[start+1].myanmar_category() == OT_As &&
|
|
|
|
info[start+2].myanmar_category() == OT_H)
|
|
|
|
{
|
|
|
|
limit += 3;
|
|
|
|
base = start;
|
|
|
|
has_reph = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
{
|
|
|
|
if (!has_reph)
|
|
|
|
base = limit;
|
|
|
|
|
|
|
|
for (unsigned int i = limit; i < end; i++)
|
|
|
|
if (is_consonant (info[i]))
|
|
|
|
{
|
|
|
|
base = i;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Reorder! */
|
|
|
|
{
|
|
|
|
unsigned int i = start;
|
|
|
|
for (; i < start + (has_reph ? 3 : 0); i++)
|
|
|
|
info[i].myanmar_position() = POS_AFTER_MAIN;
|
|
|
|
for (; i < base; i++)
|
|
|
|
info[i].myanmar_position() = POS_PRE_C;
|
|
|
|
if (i < end)
|
|
|
|
{
|
|
|
|
info[i].myanmar_position() = POS_BASE_C;
|
|
|
|
i++;
|
|
|
|
}
|
2013-02-12 22:13:56 +01:00
|
|
|
indic_position_t pos = POS_AFTER_MAIN;
|
2013-02-11 19:36:23 +01:00
|
|
|
/* The following loop may be ugly, but it implements all of
|
|
|
|
* Myanmar reordering! */
|
|
|
|
for (; i < end; i++)
|
|
|
|
{
|
|
|
|
if (info[i].myanmar_category() == OT_MR) /* Pre-base reordering */
|
|
|
|
{
|
|
|
|
info[i].myanmar_position() = POS_PRE_C;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (info[i].myanmar_position() < POS_BASE_C) /* Left matra */
|
|
|
|
{
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (pos == POS_AFTER_MAIN && info[i].myanmar_category() == OT_VBlw)
|
|
|
|
{
|
|
|
|
pos = POS_BELOW_C;
|
|
|
|
info[i].myanmar_position() = pos;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_A)
|
|
|
|
{
|
|
|
|
info[i].myanmar_position() = POS_BEFORE_SUB;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_VBlw)
|
|
|
|
{
|
|
|
|
info[i].myanmar_position() = pos;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (pos == POS_BELOW_C && info[i].myanmar_category() != OT_A)
|
|
|
|
{
|
|
|
|
pos = POS_AFTER_SUB;
|
|
|
|
info[i].myanmar_position() = pos;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
info[i].myanmar_position() = pos;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Sit tight, rock 'n roll! */
|
2015-09-01 17:15:25 +02:00
|
|
|
buffer->sort (start, end, compare_myanmar_order);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
initial_reordering_syllable (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_face_t *face,
|
|
|
|
hb_buffer_t *buffer,
|
|
|
|
unsigned int start, unsigned int end)
|
|
|
|
{
|
|
|
|
syllable_type_t syllable_type = (syllable_type_t) (buffer->info[start].syllable() & 0x0F);
|
|
|
|
switch (syllable_type) {
|
2015-07-22 12:58:11 +02:00
|
|
|
|
|
|
|
case broken_cluster: /* We already inserted dotted-circles, so just call the consonant_syllable. */
|
|
|
|
case consonant_syllable:
|
|
|
|
initial_reordering_consonant_syllable (buffer, start, end);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case punctuation_cluster:
|
|
|
|
case non_myanmar_cluster:
|
|
|
|
break;
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void
|
|
|
|
insert_dotted_circles (const hb_ot_shape_plan_t *plan HB_UNUSED,
|
|
|
|
hb_font_t *font,
|
|
|
|
hb_buffer_t *buffer)
|
|
|
|
{
|
|
|
|
/* Note: This loop is extra overhead, but should not be measurable. */
|
|
|
|
bool has_broken_syllables = false;
|
|
|
|
unsigned int count = buffer->len;
|
2014-07-17 20:22:11 +02:00
|
|
|
hb_glyph_info_t *info = buffer->info;
|
2013-02-11 19:36:23 +01:00
|
|
|
for (unsigned int i = 0; i < count; i++)
|
2014-07-17 20:22:11 +02:00
|
|
|
if ((info[i].syllable() & 0x0F) == broken_cluster)
|
|
|
|
{
|
2013-02-11 19:36:23 +01:00
|
|
|
has_broken_syllables = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (likely (!has_broken_syllables))
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
|
|
hb_codepoint_t dottedcircle_glyph;
|
2016-02-24 11:05:23 +01:00
|
|
|
if (!font->get_nominal_glyph (0x25CCu, &dottedcircle_glyph))
|
2013-02-11 19:36:23 +01:00
|
|
|
return;
|
|
|
|
|
|
|
|
hb_glyph_info_t dottedcircle = {0};
|
2014-07-11 20:54:42 +02:00
|
|
|
dottedcircle.codepoint = 0x25CCu;
|
2013-02-11 19:36:23 +01:00
|
|
|
set_myanmar_properties (dottedcircle);
|
|
|
|
dottedcircle.codepoint = dottedcircle_glyph;
|
|
|
|
|
|
|
|
buffer->clear_output ();
|
|
|
|
|
|
|
|
buffer->idx = 0;
|
|
|
|
unsigned int last_syllable = 0;
|
2016-02-02 09:39:19 +01:00
|
|
|
while (buffer->idx < buffer->len && !buffer->in_error)
|
2013-02-11 19:36:23 +01:00
|
|
|
{
|
|
|
|
unsigned int syllable = buffer->cur().syllable();
|
|
|
|
syllable_type_t syllable_type = (syllable_type_t) (syllable & 0x0F);
|
|
|
|
if (unlikely (last_syllable != syllable && syllable_type == broken_cluster))
|
|
|
|
{
|
|
|
|
last_syllable = syllable;
|
|
|
|
|
2015-10-21 15:16:49 +02:00
|
|
|
hb_glyph_info_t ginfo = dottedcircle;
|
|
|
|
ginfo.cluster = buffer->cur().cluster;
|
|
|
|
ginfo.mask = buffer->cur().mask;
|
|
|
|
ginfo.syllable() = buffer->cur().syllable();
|
2013-02-11 19:36:23 +01:00
|
|
|
|
2015-10-21 15:16:49 +02:00
|
|
|
buffer->output_info (ginfo);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
else
|
|
|
|
buffer->next_glyph ();
|
|
|
|
}
|
|
|
|
|
|
|
|
buffer->swap_buffers ();
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
initial_reordering (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_font_t *font,
|
|
|
|
hb_buffer_t *buffer)
|
|
|
|
{
|
|
|
|
insert_dotted_circles (plan, font, buffer);
|
|
|
|
|
2015-07-22 12:58:11 +02:00
|
|
|
foreach_syllable (buffer, start, end)
|
|
|
|
initial_reordering_syllable (plan, font->face, buffer, start, end);
|
2013-02-11 19:36:23 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
final_reordering (const hb_ot_shape_plan_t *plan,
|
|
|
|
hb_font_t *font HB_UNUSED,
|
|
|
|
hb_buffer_t *buffer)
|
|
|
|
{
|
|
|
|
hb_glyph_info_t *info = buffer->info;
|
|
|
|
unsigned int count = buffer->len;
|
|
|
|
|
|
|
|
/* Zero syllables now... */
|
|
|
|
for (unsigned int i = 0; i < count; i++)
|
|
|
|
info[i].syllable() = 0;
|
|
|
|
|
|
|
|
HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_category);
|
|
|
|
HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_position);
|
|
|
|
}
|
|
|
|
|
2013-10-28 00:20:59 +01:00
|
|
|
|
2014-07-27 01:17:44 +02:00
|
|
|
/* Uniscribe seems to have a shaper for 'mymr' that is like the
|
|
|
|
* generic shaper, except that it zeros mark advances GDEF_LATE. */
|
|
|
|
const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar_old =
|
|
|
|
{
|
|
|
|
"default",
|
2017-10-15 12:11:08 +02:00
|
|
|
nullptr, /* collect_features */
|
|
|
|
nullptr, /* override_features */
|
|
|
|
nullptr, /* data_create */
|
|
|
|
nullptr, /* data_destroy */
|
|
|
|
nullptr, /* preprocess_text */
|
|
|
|
nullptr, /* postprocess_glyphs */
|
2014-07-27 01:17:44 +02:00
|
|
|
HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
|
2017-10-15 12:11:08 +02:00
|
|
|
nullptr, /* decompose */
|
|
|
|
nullptr, /* compose */
|
|
|
|
nullptr, /* setup_masks */
|
|
|
|
nullptr, /* disable_otl */
|
|
|
|
nullptr, /* reorder_marks */
|
2014-07-27 01:17:44 +02:00
|
|
|
HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
|
|
|
|
true, /* fallback_position */
|
|
|
|
};
|
|
|
|
|
2013-02-11 19:36:23 +01:00
|
|
|
const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar =
|
|
|
|
{
|
|
|
|
"myanmar",
|
|
|
|
collect_features_myanmar,
|
|
|
|
override_features_myanmar,
|
2017-10-15 12:11:08 +02:00
|
|
|
nullptr, /* data_create */
|
|
|
|
nullptr, /* data_destroy */
|
|
|
|
nullptr, /* preprocess_text */
|
|
|
|
nullptr, /* postprocess_glyphs */
|
2013-12-31 09:04:35 +01:00
|
|
|
HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
|
2017-10-15 12:11:08 +02:00
|
|
|
nullptr, /* decompose */
|
|
|
|
nullptr, /* compose */
|
2013-02-11 19:36:23 +01:00
|
|
|
setup_masks_myanmar,
|
2017-10-15 12:11:08 +02:00
|
|
|
nullptr, /* disable_otl */
|
|
|
|
nullptr, /* reorder_marks */
|
2013-10-28 00:20:59 +01:00
|
|
|
HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY,
|
2013-02-11 19:36:23 +01:00
|
|
|
false, /* fallback_position */
|
|
|
|
};
|