1 /*
2 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
3 *
4 * This code is free software; you can redistribute it and/or modify it
5 * under the terms of the GNU General Public License version 2 only, as
6 * published by the Free Software Foundation. Oracle designates this
7 * particular file as subject to the "Classpath" exception as provided
8 * by Oracle in the LICENSE file that accompanied this code.
9 *
10 * This code is distributed in the hope that it will be useful, but WITHOUT
11 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
12 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
13 * version 2 for more details (a copy is included in the LICENSE file that
14 * accompanied this code).
15 *
16 * You should have received a copy of the GNU General Public License version
17 * 2 along with this work; if not, write to the Free Software Foundation,
18 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
19 *
20 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
21 * or visit www.oracle.com if you need additional information or have any
22 * questions.
23 */
24
25 // This file is available under and governed by the GNU General Public
26 // License version 2 only, as published by the Free Software Foundation.
27 // However, the following notice accompanied the original version of this
28 // file:
29 //
30 /*
31 * Copyright © 2011,2012,2013 Google, Inc.
32 *
33 * This is part of HarfBuzz, a text shaping library.
34 *
35 * Permission is hereby granted, without written agreement and without
36 * license or royalty fees, to use, copy, modify, and distribute this
37 * software and its documentation for any purpose, provided that the
38 * above copyright notice and the following two paragraphs appear in
39 * all copies of this software.
40 *
41 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
42 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
43 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
44 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
45 * DAMAGE.
46 *
47 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
48 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
49 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
50 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
51 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
52 *
53 * Google Author(s): Behdad Esfahbod
54 */
55
56 #include "hb-ot-shape-complex-indic-private.hh"
57
58 /* buffer var allocations */
59 #define myanmar_category() complex_var_u8_0() /* myanmar_category_t */
60 #define myanmar_position() complex_var_u8_1() /* myanmar_position_t */
61
62
63 /*
64 * Myanmar shaper.
65 */
66
67 static const hb_tag_t
68 basic_features[] =
69 {
70 /*
71 * Basic features.
72 * These features are applied in order, one at a time, after initial_reordering.
73 */
74 HB_TAG('r','p','h','f'),
75 HB_TAG('p','r','e','f'),
76 HB_TAG('b','l','w','f'),
77 HB_TAG('p','s','t','f'),
78 };
79 static const hb_tag_t
80 other_features[] =
81 {
82 /*
83 * Other features.
84 * These features are applied all at once, after final_reordering.
85 */
86 HB_TAG('p','r','e','s'),
87 HB_TAG('a','b','v','s'),
88 HB_TAG('b','l','w','s'),
89 HB_TAG('p','s','t','s'),
90 /* Positioning features, though we don't care about the types. */
91 HB_TAG('d','i','s','t'),
92 /* Pre-release version of Windows 8 Myanmar font had abvm,blwm
93 * features. The released Windows 8 version of the font (as well
94 * as the released spec) used 'mark' instead. The Windows 8
95 * shaper however didn't apply 'mark' but did apply 'mkmk'.
96 * Perhaps it applied abvm/blwm. This was fixed in a Windows 8
97 * update, so now it applies mark/mkmk. We are guessing that
98 * it still applies abvm/blwm too.
99 */
100 HB_TAG('a','b','v','m'),
101 HB_TAG('b','l','w','m'),
102 };
103
104 static void
105 setup_syllables (const hb_ot_shape_plan_t *plan,
106 hb_font_t *font,
107 hb_buffer_t *buffer);
108 static void
109 initial_reordering (const hb_ot_shape_plan_t *plan,
110 hb_font_t *font,
111 hb_buffer_t *buffer);
112 static void
113 final_reordering (const hb_ot_shape_plan_t *plan,
114 hb_font_t *font,
115 hb_buffer_t *buffer);
116
117 static void
118 collect_features_myanmar (hb_ot_shape_planner_t *plan)
119 {
120 hb_ot_map_builder_t *map = &plan->map;
121
122 /* Do this before any lookups have been applied. */
123 map->add_gsub_pause (setup_syllables);
124
125 map->add_global_bool_feature (HB_TAG('l','o','c','l'));
126 /* The Indic specs do not require ccmp, but we apply it here since if
127 * there is a use of it, it's typically at the beginning. */
128 map->add_global_bool_feature (HB_TAG('c','c','m','p'));
129
130
131 map->add_gsub_pause (initial_reordering);
132 for (unsigned int i = 0; i < ARRAY_LENGTH (basic_features); i++)
133 {
134 map->add_feature (basic_features[i], 1, F_GLOBAL | F_MANUAL_ZWJ);
135 map->add_gsub_pause (NULL);
136 }
137 map->add_gsub_pause (final_reordering);
138 for (unsigned int i = 0; i < ARRAY_LENGTH (other_features); i++)
139 map->add_feature (other_features[i], 1, F_GLOBAL | F_MANUAL_ZWJ);
140 }
141
142 static void
143 override_features_myanmar (hb_ot_shape_planner_t *plan)
144 {
145 plan->map.add_feature (HB_TAG('l','i','g','a'), 0, F_GLOBAL);
146 }
147
148
149 enum syllable_type_t {
150 consonant_syllable,
151 punctuation_cluster,
152 broken_cluster,
153 non_myanmar_cluster,
154 };
155
156 #include "hb-ot-shape-complex-myanmar-machine.hh"
157
158
159 /* Note: This enum is duplicated in the -machine.rl source file.
160 * Not sure how to avoid duplication. */
161 enum myanmar_category_t {
162 OT_As = 18, /* Asat */
163 OT_D = 19, /* Digits except zero */
164 OT_D0 = 20, /* Digit zero */
165 OT_DB = OT_N, /* Dot below */
166 OT_GB = OT_PLACEHOLDER,
167 OT_MH = 21, /* Various consonant medial types */
168 OT_MR = 22, /* Various consonant medial types */
169 OT_MW = 23, /* Various consonant medial types */
170 OT_MY = 24, /* Various consonant medial types */
171 OT_PT = 25, /* Pwo and other tones */
172 OT_VAbv = 26,
173 OT_VBlw = 27,
174 OT_VPre = 28,
175 OT_VPst = 29,
176 OT_VS = 30, /* Variation selectors */
177 OT_P = 31 /* Punctuation */
178 };
179
180
181 static inline bool
182 is_one_of (const hb_glyph_info_t &info, unsigned int flags)
183 {
184 /* If it ligated, all bets are off. */
185 if (_hb_glyph_info_ligated (&info)) return false;
186 return !!(FLAG_SAFE (info.myanmar_category()) & flags);
187 }
188
189 static inline bool
190 is_consonant (const hb_glyph_info_t &info)
191 {
192 return is_one_of (info, CONSONANT_FLAGS);
193 }
194
195
196 static inline void
197 set_myanmar_properties (hb_glyph_info_t &info)
198 {
199 hb_codepoint_t u = info.codepoint;
200 unsigned int type = hb_indic_get_categories (u);
201 indic_category_t cat = (indic_category_t) (type & 0x7Fu);
202 indic_position_t pos = (indic_position_t) (type >> 8);
203
204 /* Myanmar
205 * http://www.microsoft.com/typography/OpenTypeDev/myanmar/intro.htm#analyze
206 */
207 if (unlikely (hb_in_range (u, 0xFE00u, 0xFE0Fu)))
208 cat = (indic_category_t) OT_VS;
209
210 switch (u)
211 {
212 case 0x104Eu:
213 cat = (indic_category_t) OT_C; /* The spec says C, IndicSyllableCategory doesn't have. */
214 break;
215
216 case 0x002Du: case 0x00A0u: case 0x00D7u: case 0x2012u:
217 case 0x2013u: case 0x2014u: case 0x2015u: case 0x2022u:
218 case 0x25CCu: case 0x25FBu: case 0x25FCu: case 0x25FDu:
219 case 0x25FEu:
220 cat = (indic_category_t) OT_GB;
221 break;
222
223 case 0x1004u: case 0x101Bu: case 0x105Au:
224 cat = (indic_category_t) OT_Ra;
225 break;
226
227 case 0x1032u: case 0x1036u:
228 cat = (indic_category_t) OT_A;
229 break;
230
231 case 0x103Au:
232 cat = (indic_category_t) OT_As;
233 break;
234
235 case 0x1041u: case 0x1042u: case 0x1043u: case 0x1044u:
236 case 0x1045u: case 0x1046u: case 0x1047u: case 0x1048u:
237 case 0x1049u: case 0x1090u: case 0x1091u: case 0x1092u:
238 case 0x1093u: case 0x1094u: case 0x1095u: case 0x1096u:
239 case 0x1097u: case 0x1098u: case 0x1099u:
240 cat = (indic_category_t) OT_D;
241 break;
242
243 case 0x1040u:
244 cat = (indic_category_t) OT_D; /* XXX The spec says D0, but Uniscribe doesn't seem to do. */
245 break;
246
247 case 0x103Eu: case 0x1060u:
248 cat = (indic_category_t) OT_MH;
249 break;
250
251 case 0x103Cu:
252 cat = (indic_category_t) OT_MR;
253 break;
254
255 case 0x103Du: case 0x1082u:
256 cat = (indic_category_t) OT_MW;
257 break;
258
259 case 0x103Bu: case 0x105Eu: case 0x105Fu:
260 cat = (indic_category_t) OT_MY;
261 break;
262
263 case 0x1063u: case 0x1064u: case 0x1069u: case 0x106Au:
264 case 0x106Bu: case 0x106Cu: case 0x106Du: case 0xAA7Bu:
265 cat = (indic_category_t) OT_PT;
266 break;
267
268 case 0x1038u: case 0x1087u: case 0x1088u: case 0x1089u:
269 case 0x108Au: case 0x108Bu: case 0x108Cu: case 0x108Du:
270 case 0x108Fu: case 0x109Au: case 0x109Bu: case 0x109Cu:
271 cat = (indic_category_t) OT_SM;
272 break;
273
274 case 0x104Au: case 0x104Bu:
275 cat = (indic_category_t) OT_P;
276 break;
277 }
278
279 if (cat == OT_M)
280 {
281 switch ((int) pos)
282 {
283 case POS_PRE_C: cat = (indic_category_t) OT_VPre;
284 pos = POS_PRE_M; break;
285 case POS_ABOVE_C: cat = (indic_category_t) OT_VAbv; break;
286 case POS_BELOW_C: cat = (indic_category_t) OT_VBlw; break;
287 case POS_POST_C: cat = (indic_category_t) OT_VPst; break;
288 }
289 }
290
291 info.myanmar_category() = (myanmar_category_t) cat;
292 info.myanmar_position() = pos;
293 }
294
295
296
297 static void
298 setup_masks_myanmar (const hb_ot_shape_plan_t *plan HB_UNUSED,
299 hb_buffer_t *buffer,
300 hb_font_t *font HB_UNUSED)
301 {
302 HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_category);
303 HB_BUFFER_ALLOCATE_VAR (buffer, myanmar_position);
304
305 /* We cannot setup masks here. We save information about characters
306 * and setup masks later on in a pause-callback. */
307
308 unsigned int count = buffer->len;
309 hb_glyph_info_t *info = buffer->info;
310 for (unsigned int i = 0; i < count; i++)
311 set_myanmar_properties (info[i]);
312 }
313
314 static void
315 setup_syllables (const hb_ot_shape_plan_t *plan HB_UNUSED,
316 hb_font_t *font HB_UNUSED,
317 hb_buffer_t *buffer)
318 {
319 find_syllables (buffer);
320 }
321
322 static int
323 compare_myanmar_order (const hb_glyph_info_t *pa, const hb_glyph_info_t *pb)
324 {
325 int a = pa->myanmar_position();
326 int b = pb->myanmar_position();
327
328 return a < b ? -1 : a == b ? 0 : +1;
329 }
330
331
332 /* Rules from:
333 * http://www.microsoft.com/typography/OpenTypeDev/myanmar/intro.htm */
334
335 static void
336 initial_reordering_consonant_syllable (hb_buffer_t *buffer,
337 unsigned int start, unsigned int end)
338 {
339 hb_glyph_info_t *info = buffer->info;
340
341 unsigned int base = end;
342 bool has_reph = false;
343
344 {
345 unsigned int limit = start;
346 if (start + 3 <= end &&
347 info[start ].myanmar_category() == OT_Ra &&
348 info[start+1].myanmar_category() == OT_As &&
349 info[start+2].myanmar_category() == OT_H)
350 {
351 limit += 3;
352 base = start;
353 has_reph = true;
354 }
355
356 {
357 if (!has_reph)
358 base = limit;
359
360 for (unsigned int i = limit; i < end; i++)
361 if (is_consonant (info[i]))
362 {
363 base = i;
364 break;
365 }
366 }
367 }
368
369 /* Reorder! */
370 {
371 unsigned int i = start;
372 for (; i < start + (has_reph ? 3 : 0); i++)
373 info[i].myanmar_position() = POS_AFTER_MAIN;
374 for (; i < base; i++)
375 info[i].myanmar_position() = POS_PRE_C;
376 if (i < end)
377 {
378 info[i].myanmar_position() = POS_BASE_C;
379 i++;
380 }
381 indic_position_t pos = POS_AFTER_MAIN;
382 /* The following loop may be ugly, but it implements all of
383 * Myanmar reordering! */
384 for (; i < end; i++)
385 {
386 if (info[i].myanmar_category() == OT_MR) /* Pre-base reordering */
387 {
388 info[i].myanmar_position() = POS_PRE_C;
389 continue;
390 }
391 if (info[i].myanmar_position() < POS_BASE_C) /* Left matra */
392 {
393 continue;
394 }
395
396 if (pos == POS_AFTER_MAIN && info[i].myanmar_category() == OT_VBlw)
397 {
398 pos = POS_BELOW_C;
399 info[i].myanmar_position() = pos;
400 continue;
401 }
402
403 if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_A)
404 {
405 info[i].myanmar_position() = POS_BEFORE_SUB;
406 continue;
407 }
408 if (pos == POS_BELOW_C && info[i].myanmar_category() == OT_VBlw)
409 {
410 info[i].myanmar_position() = pos;
411 continue;
412 }
413 if (pos == POS_BELOW_C && info[i].myanmar_category() != OT_A)
414 {
415 pos = POS_AFTER_SUB;
416 info[i].myanmar_position() = pos;
417 continue;
418 }
419 info[i].myanmar_position() = pos;
420 }
421 }
422
423 /* Sit tight, rock 'n roll! */
424 buffer->sort (start, end, compare_myanmar_order);
425 }
426
427 static void
428 initial_reordering_syllable (const hb_ot_shape_plan_t *plan,
429 hb_face_t *face,
430 hb_buffer_t *buffer,
431 unsigned int start, unsigned int end)
432 {
433 syllable_type_t syllable_type = (syllable_type_t) (buffer->info[start].syllable() & 0x0F);
434 switch (syllable_type) {
435
436 case broken_cluster: /* We already inserted dotted-circles, so just call the consonant_syllable. */
437 case consonant_syllable:
438 initial_reordering_consonant_syllable (buffer, start, end);
439 break;
440
441 case punctuation_cluster:
442 case non_myanmar_cluster:
443 break;
444 }
445 }
446
447 static inline void
448 insert_dotted_circles (const hb_ot_shape_plan_t *plan HB_UNUSED,
449 hb_font_t *font,
450 hb_buffer_t *buffer)
451 {
452 /* Note: This loop is extra overhead, but should not be measurable. */
453 bool has_broken_syllables = false;
454 unsigned int count = buffer->len;
455 hb_glyph_info_t *info = buffer->info;
456 for (unsigned int i = 0; i < count; i++)
457 if ((info[i].syllable() & 0x0F) == broken_cluster)
458 {
459 has_broken_syllables = true;
460 break;
461 }
462 if (likely (!has_broken_syllables))
463 return;
464
465
466 hb_codepoint_t dottedcircle_glyph;
467 if (!font->get_glyph (0x25CCu, 0, &dottedcircle_glyph))
468 return;
469
470 hb_glyph_info_t dottedcircle = {0};
471 dottedcircle.codepoint = 0x25CCu;
472 set_myanmar_properties (dottedcircle);
473 dottedcircle.codepoint = dottedcircle_glyph;
474
475 buffer->clear_output ();
476
477 buffer->idx = 0;
478 unsigned int last_syllable = 0;
479 while (buffer->idx < buffer->len)
480 {
481 unsigned int syllable = buffer->cur().syllable();
482 syllable_type_t syllable_type = (syllable_type_t) (syllable & 0x0F);
483 if (unlikely (last_syllable != syllable && syllable_type == broken_cluster))
484 {
485 last_syllable = syllable;
486
487 hb_glyph_info_t info = dottedcircle;
488 info.cluster = buffer->cur().cluster;
489 info.mask = buffer->cur().mask;
490 info.syllable() = buffer->cur().syllable();
491
492 buffer->output_info (info);
493 }
494 else
495 buffer->next_glyph ();
496 }
497
498 buffer->swap_buffers ();
499 }
500
501 static void
502 initial_reordering (const hb_ot_shape_plan_t *plan,
503 hb_font_t *font,
504 hb_buffer_t *buffer)
505 {
506 insert_dotted_circles (plan, font, buffer);
507
508 foreach_syllable (buffer, start, end)
509 initial_reordering_syllable (plan, font->face, buffer, start, end);
510 }
511
512 static void
513 final_reordering (const hb_ot_shape_plan_t *plan,
514 hb_font_t *font HB_UNUSED,
515 hb_buffer_t *buffer)
516 {
517 hb_glyph_info_t *info = buffer->info;
518 unsigned int count = buffer->len;
519
520 /* Zero syllables now... */
521 for (unsigned int i = 0; i < count; i++)
522 info[i].syllable() = 0;
523
524 HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_category);
525 HB_BUFFER_DEALLOCATE_VAR (buffer, myanmar_position);
526 }
527
528
529 /* Uniscribe seems to have a shaper for 'mymr' that is like the
530 * generic shaper, except that it zeros mark advances GDEF_LATE. */
531 const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar_old =
532 {
533 "default",
534 NULL, /* collect_features */
535 NULL, /* override_features */
536 NULL, /* data_create */
537 NULL, /* data_destroy */
538 NULL, /* preprocess_text */
539 HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
540 NULL, /* decompose */
541 NULL, /* compose */
542 NULL, /* setup_masks */
543 HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
544 true, /* fallback_position */
545 };
546
547 const hb_ot_complex_shaper_t _hb_ot_complex_shaper_myanmar =
548 {
549 "myanmar",
550 collect_features_myanmar,
551 override_features_myanmar,
552 NULL, /* data_create */
553 NULL, /* data_destroy */
554 NULL, /* preprocess_text */
555 HB_OT_SHAPE_NORMALIZATION_MODE_COMPOSED_DIACRITICS_NO_SHORT_CIRCUIT,
556 NULL, /* decompose */
557 NULL, /* compose */
558 setup_masks_myanmar,
559 HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_EARLY,
560 false, /* fallback_position */
561 };