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 © 2009 Red Hat, Inc.
32 * Copyright © 2011 Codethink Limited
33 * Copyright © 2011,2012 Google, Inc.
34 *
35 * This is part of HarfBuzz, a text shaping library.
36 *
37 * Permission is hereby granted, without written agreement and without
38 * license or royalty fees, to use, copy, modify, and distribute this
39 * software and its documentation for any purpose, provided that the
40 * above copyright notice and the following two paragraphs appear in
41 * all copies of this software.
42 *
43 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
44 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
45 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
46 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
47 * DAMAGE.
48 *
49 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
50 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
51 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
52 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
53 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
54 *
55 * Red Hat Author(s): Behdad Esfahbod
56 * Codethink Author(s): Ryan Lortie
57 * Google Author(s): Behdad Esfahbod
58 */
59
60 #ifndef HB_H_IN
61 #error "Include <hb.h> instead."
62 #endif
63
64 #ifndef HB_UNICODE_H
65 #define HB_UNICODE_H
66
67 #include "hb-common.h"
68
69 HB_BEGIN_DECLS
70
71
72 /* hb_unicode_general_category_t */
73
74 /* Unicode Character Database property: General_Category (gc) */
75 typedef enum
76 {
77 HB_UNICODE_GENERAL_CATEGORY_CONTROL, /* Cc */
78 HB_UNICODE_GENERAL_CATEGORY_FORMAT, /* Cf */
79 HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED, /* Cn */
80 HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE, /* Co */
81 HB_UNICODE_GENERAL_CATEGORY_SURROGATE, /* Cs */
82 HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER, /* Ll */
83 HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER, /* Lm */
84 HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER, /* Lo */
85 HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER, /* Lt */
86 HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER, /* Lu */
87 HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK, /* Mc */
88 HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK, /* Me */
89 HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK, /* Mn */
90 HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER, /* Nd */
91 HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER, /* Nl */
92 HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER, /* No */
93 HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION, /* Pc */
94 HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION, /* Pd */
95 HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION, /* Pe */
96 HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION, /* Pf */
97 HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION, /* Pi */
98 HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION, /* Po */
99 HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION, /* Ps */
100 HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL, /* Sc */
101 HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL, /* Sk */
102 HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL, /* Sm */
103 HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL, /* So */
104 HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR, /* Zl */
105 HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR, /* Zp */
106 HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR /* Zs */
107 } hb_unicode_general_category_t;
108
109 /* hb_unicode_combining_class_t */
110
111 /* Note: newer versions of Unicode may add new values. Clients should be ready to handle
112 * any value in the 0..254 range being returned from hb_unicode_combining_class().
113 */
114
115 /* Unicode Character Database property: Canonical_Combining_Class (ccc) */
116 typedef enum
117 {
118 HB_UNICODE_COMBINING_CLASS_NOT_REORDERED = 0,
119 HB_UNICODE_COMBINING_CLASS_OVERLAY = 1,
120 HB_UNICODE_COMBINING_CLASS_NUKTA = 7,
121 HB_UNICODE_COMBINING_CLASS_KANA_VOICING = 8,
122 HB_UNICODE_COMBINING_CLASS_VIRAMA = 9,
123
124 /* Hebrew */
125 HB_UNICODE_COMBINING_CLASS_CCC10 = 10,
126 HB_UNICODE_COMBINING_CLASS_CCC11 = 11,
127 HB_UNICODE_COMBINING_CLASS_CCC12 = 12,
128 HB_UNICODE_COMBINING_CLASS_CCC13 = 13,
129 HB_UNICODE_COMBINING_CLASS_CCC14 = 14,
130 HB_UNICODE_COMBINING_CLASS_CCC15 = 15,
131 HB_UNICODE_COMBINING_CLASS_CCC16 = 16,
132 HB_UNICODE_COMBINING_CLASS_CCC17 = 17,
133 HB_UNICODE_COMBINING_CLASS_CCC18 = 18,
134 HB_UNICODE_COMBINING_CLASS_CCC19 = 19,
135 HB_UNICODE_COMBINING_CLASS_CCC20 = 20,
136 HB_UNICODE_COMBINING_CLASS_CCC21 = 21,
137 HB_UNICODE_COMBINING_CLASS_CCC22 = 22,
138 HB_UNICODE_COMBINING_CLASS_CCC23 = 23,
139 HB_UNICODE_COMBINING_CLASS_CCC24 = 24,
140 HB_UNICODE_COMBINING_CLASS_CCC25 = 25,
141 HB_UNICODE_COMBINING_CLASS_CCC26 = 26,
142
143 /* Arabic */
144 HB_UNICODE_COMBINING_CLASS_CCC27 = 27,
145 HB_UNICODE_COMBINING_CLASS_CCC28 = 28,
146 HB_UNICODE_COMBINING_CLASS_CCC29 = 29,
147 HB_UNICODE_COMBINING_CLASS_CCC30 = 30,
148 HB_UNICODE_COMBINING_CLASS_CCC31 = 31,
149 HB_UNICODE_COMBINING_CLASS_CCC32 = 32,
150 HB_UNICODE_COMBINING_CLASS_CCC33 = 33,
151 HB_UNICODE_COMBINING_CLASS_CCC34 = 34,
152 HB_UNICODE_COMBINING_CLASS_CCC35 = 35,
153
154 /* Syriac */
155 HB_UNICODE_COMBINING_CLASS_CCC36 = 36,
156
157 /* Telugu */
158 HB_UNICODE_COMBINING_CLASS_CCC84 = 84,
159 HB_UNICODE_COMBINING_CLASS_CCC91 = 91,
160
161 /* Thai */
162 HB_UNICODE_COMBINING_CLASS_CCC103 = 103,
163 HB_UNICODE_COMBINING_CLASS_CCC107 = 107,
164
165 /* Lao */
166 HB_UNICODE_COMBINING_CLASS_CCC118 = 118,
167 HB_UNICODE_COMBINING_CLASS_CCC122 = 122,
168
169 /* Tibetan */
170 HB_UNICODE_COMBINING_CLASS_CCC129 = 129,
171 HB_UNICODE_COMBINING_CLASS_CCC130 = 130,
172 HB_UNICODE_COMBINING_CLASS_CCC133 = 132,
173
174
175 HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW_LEFT = 200,
176 HB_UNICODE_COMBINING_CLASS_ATTACHED_BELOW = 202,
177 HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE = 214,
178 HB_UNICODE_COMBINING_CLASS_ATTACHED_ABOVE_RIGHT = 216,
179 HB_UNICODE_COMBINING_CLASS_BELOW_LEFT = 218,
180 HB_UNICODE_COMBINING_CLASS_BELOW = 220,
181 HB_UNICODE_COMBINING_CLASS_BELOW_RIGHT = 222,
182 HB_UNICODE_COMBINING_CLASS_LEFT = 224,
183 HB_UNICODE_COMBINING_CLASS_RIGHT = 226,
184 HB_UNICODE_COMBINING_CLASS_ABOVE_LEFT = 228,
185 HB_UNICODE_COMBINING_CLASS_ABOVE = 230,
186 HB_UNICODE_COMBINING_CLASS_ABOVE_RIGHT = 232,
187 HB_UNICODE_COMBINING_CLASS_DOUBLE_BELOW = 233,
188 HB_UNICODE_COMBINING_CLASS_DOUBLE_ABOVE = 234,
189
190 HB_UNICODE_COMBINING_CLASS_IOTA_SUBSCRIPT = 240,
191
192 HB_UNICODE_COMBINING_CLASS_INVALID = 255
193 } hb_unicode_combining_class_t;
194
195
196 /*
197 * hb_unicode_funcs_t
198 */
199
200 typedef struct hb_unicode_funcs_t hb_unicode_funcs_t;
201
202
203 /*
204 * just give me the best implementation you've got there.
205 */
206 hb_unicode_funcs_t *
207 hb_unicode_funcs_get_default (void);
208
209
210 hb_unicode_funcs_t *
211 hb_unicode_funcs_create (hb_unicode_funcs_t *parent);
212
213 hb_unicode_funcs_t *
214 hb_unicode_funcs_get_empty (void);
215
216 hb_unicode_funcs_t *
217 hb_unicode_funcs_reference (hb_unicode_funcs_t *ufuncs);
218
219 void
220 hb_unicode_funcs_destroy (hb_unicode_funcs_t *ufuncs);
221
222 hb_bool_t
223 hb_unicode_funcs_set_user_data (hb_unicode_funcs_t *ufuncs,
224 hb_user_data_key_t *key,
225 void * data,
226 hb_destroy_func_t destroy,
227 hb_bool_t replace);
228
229
230 void *
231 hb_unicode_funcs_get_user_data (hb_unicode_funcs_t *ufuncs,
232 hb_user_data_key_t *key);
233
234
235 void
236 hb_unicode_funcs_make_immutable (hb_unicode_funcs_t *ufuncs);
237
238 hb_bool_t
239 hb_unicode_funcs_is_immutable (hb_unicode_funcs_t *ufuncs);
240
241 hb_unicode_funcs_t *
242 hb_unicode_funcs_get_parent (hb_unicode_funcs_t *ufuncs);
243
244
245 /*
246 * funcs
247 */
248
249 /* typedefs */
250
251 typedef hb_unicode_combining_class_t (*hb_unicode_combining_class_func_t) (hb_unicode_funcs_t *ufuncs,
252 hb_codepoint_t unicode,
253 void *user_data);
254 typedef unsigned int (*hb_unicode_eastasian_width_func_t) (hb_unicode_funcs_t *ufuncs,
255 hb_codepoint_t unicode,
256 void *user_data);
257 typedef hb_unicode_general_category_t (*hb_unicode_general_category_func_t) (hb_unicode_funcs_t *ufuncs,
258 hb_codepoint_t unicode,
259 void *user_data);
260 typedef hb_codepoint_t (*hb_unicode_mirroring_func_t) (hb_unicode_funcs_t *ufuncs,
261 hb_codepoint_t unicode,
262 void *user_data);
263 typedef hb_script_t (*hb_unicode_script_func_t) (hb_unicode_funcs_t *ufuncs,
264 hb_codepoint_t unicode,
265 void *user_data);
266
267 typedef hb_bool_t (*hb_unicode_compose_func_t) (hb_unicode_funcs_t *ufuncs,
268 hb_codepoint_t a,
269 hb_codepoint_t b,
270 hb_codepoint_t *ab,
271 void *user_data);
272 typedef hb_bool_t (*hb_unicode_decompose_func_t) (hb_unicode_funcs_t *ufuncs,
273 hb_codepoint_t ab,
274 hb_codepoint_t *a,
275 hb_codepoint_t *b,
276 void *user_data);
277
278 /**
279 * hb_unicode_decompose_compatibility_func_t:
280 * @ufuncs: a Unicode function structure
281 * @u: codepoint to decompose
282 * @decomposed: address of codepoint array (of length %HB_UNICODE_MAX_DECOMPOSITION_LEN) to write decomposition into
283 * @user_data: user data pointer as passed to hb_unicode_funcs_set_decompose_compatibility_func()
284 *
285 * Fully decompose @u to its Unicode compatibility decomposition. The codepoints of the decomposition will be written to @decomposed.
286 * The complete length of the decomposition will be returned.
287 *
288 * If @u has no compatibility decomposition, zero should be returned.
289 *
290 * The Unicode standard guarantees that a buffer of length %HB_UNICODE_MAX_DECOMPOSITION_LEN codepoints will always be sufficient for any
291 * compatibility decomposition plus an terminating value of 0. Consequently, @decompose must be allocated by the caller to be at least this length. Implementations
292 * of this function type must ensure that they do not write past the provided array.
293 *
294 * Return value: number of codepoints in the full compatibility decomposition of @u, or 0 if no decomposition available.
295 */
296 typedef unsigned int (*hb_unicode_decompose_compatibility_func_t) (hb_unicode_funcs_t *ufuncs,
297 hb_codepoint_t u,
298 hb_codepoint_t *decomposed,
299 void *user_data);
300
301 /* See Unicode 6.1 for details on the maximum decomposition length. */
302 #define HB_UNICODE_MAX_DECOMPOSITION_LEN (18+1) /* codepoints */
303
304 /* setters */
305
306 /**
307 * hb_unicode_funcs_set_combining_class_func:
308 * @ufuncs: a Unicode function structure
309 * @func: (closure user_data) (destroy destroy) (scope notified):
310 * @user_data:
311 * @destroy:
312 *
313 *
314 *
315 * Since: 0.9.2
316 **/
317 void
318 hb_unicode_funcs_set_combining_class_func (hb_unicode_funcs_t *ufuncs,
319 hb_unicode_combining_class_func_t func,
320 void *user_data, hb_destroy_func_t destroy);
321
322 /**
323 * hb_unicode_funcs_set_eastasian_width_func:
324 * @ufuncs: a Unicode function structure
325 * @func: (closure user_data) (destroy destroy) (scope notified):
326 * @user_data:
327 * @destroy:
328 *
329 *
330 *
331 * Since: 0.9.2
332 **/
333 void
334 hb_unicode_funcs_set_eastasian_width_func (hb_unicode_funcs_t *ufuncs,
335 hb_unicode_eastasian_width_func_t func,
336 void *user_data, hb_destroy_func_t destroy);
337
338 /**
339 * hb_unicode_funcs_set_general_category_func:
340 * @ufuncs: a Unicode function structure
341 * @func: (closure user_data) (destroy destroy) (scope notified):
342 * @user_data:
343 * @destroy:
344 *
345 *
346 *
347 * Since: 0.9.2
348 **/
349 void
350 hb_unicode_funcs_set_general_category_func (hb_unicode_funcs_t *ufuncs,
351 hb_unicode_general_category_func_t func,
352 void *user_data, hb_destroy_func_t destroy);
353
354 /**
355 * hb_unicode_funcs_set_mirroring_func:
356 * @ufuncs: a Unicode function structure
357 * @func: (closure user_data) (destroy destroy) (scope notified):
358 * @user_data:
359 * @destroy:
360 *
361 *
362 *
363 * Since: 0.9.2
364 **/
365 void
366 hb_unicode_funcs_set_mirroring_func (hb_unicode_funcs_t *ufuncs,
367 hb_unicode_mirroring_func_t func,
368 void *user_data, hb_destroy_func_t destroy);
369
370 /**
371 * hb_unicode_funcs_set_script_func:
372 * @ufuncs: a Unicode function structure
373 * @func: (closure user_data) (destroy destroy) (scope notified):
374 * @user_data:
375 * @destroy:
376 *
377 *
378 *
379 * Since: 0.9.2
380 **/
381 void
382 hb_unicode_funcs_set_script_func (hb_unicode_funcs_t *ufuncs,
383 hb_unicode_script_func_t func,
384 void *user_data, hb_destroy_func_t destroy);
385
386 /**
387 * hb_unicode_funcs_set_compose_func:
388 * @ufuncs: a Unicode function structure
389 * @func: (closure user_data) (destroy destroy) (scope notified):
390 * @user_data:
391 * @destroy:
392 *
393 *
394 *
395 * Since: 0.9.2
396 **/
397 void
398 hb_unicode_funcs_set_compose_func (hb_unicode_funcs_t *ufuncs,
399 hb_unicode_compose_func_t func,
400 void *user_data, hb_destroy_func_t destroy);
401
402 /**
403 * hb_unicode_funcs_set_decompose_func:
404 * @ufuncs: a Unicode function structure
405 * @func: (closure user_data) (destroy destroy) (scope notified):
406 * @user_data:
407 * @destroy:
408 *
409 *
410 *
411 * Since: 0.9.2
412 **/
413 void
414 hb_unicode_funcs_set_decompose_func (hb_unicode_funcs_t *ufuncs,
415 hb_unicode_decompose_func_t func,
416 void *user_data, hb_destroy_func_t destroy);
417
418 /**
419 * hb_unicode_funcs_set_decompose_compatibility_func:
420 * @ufuncs: a Unicode function structure
421 * @func: (closure user_data) (destroy destroy) (scope notified):
422 * @user_data:
423 * @destroy:
424 *
425 *
426 *
427 * Since: 0.9.2
428 **/
429 void
430 hb_unicode_funcs_set_decompose_compatibility_func (hb_unicode_funcs_t *ufuncs,
431 hb_unicode_decompose_compatibility_func_t func,
432 void *user_data, hb_destroy_func_t destroy);
433
434 /* accessors */
435
436 /**
437 * Since: 0.9.2
438 **/
439 hb_unicode_combining_class_t
440 hb_unicode_combining_class (hb_unicode_funcs_t *ufuncs,
441 hb_codepoint_t unicode);
442
443 /**
444 * Since: 0.9.2
445 **/
446 unsigned int
447 hb_unicode_eastasian_width (hb_unicode_funcs_t *ufuncs,
448 hb_codepoint_t unicode);
449
450 /**
451 * Since: 0.9.2
452 **/
453 hb_unicode_general_category_t
454 hb_unicode_general_category (hb_unicode_funcs_t *ufuncs,
455 hb_codepoint_t unicode);
456
457 /**
458 * Since: 0.9.2
459 **/
460 hb_codepoint_t
461 hb_unicode_mirroring (hb_unicode_funcs_t *ufuncs,
462 hb_codepoint_t unicode);
463
464 /**
465 * Since: 0.9.2
466 **/
467 hb_script_t
468 hb_unicode_script (hb_unicode_funcs_t *ufuncs,
469 hb_codepoint_t unicode);
470
471 /**
472 * Since: 0.9.2
473 **/
474 hb_bool_t
475 hb_unicode_compose (hb_unicode_funcs_t *ufuncs,
476 hb_codepoint_t a,
477 hb_codepoint_t b,
478 hb_codepoint_t *ab);
479
480 /**
481 * Since: 0.9.2
482 **/
483 hb_bool_t
484 hb_unicode_decompose (hb_unicode_funcs_t *ufuncs,
485 hb_codepoint_t ab,
486 hb_codepoint_t *a,
487 hb_codepoint_t *b);
488
489 /**
490 * Since: 0.9.2
491 **/
492 unsigned int
493 hb_unicode_decompose_compatibility (hb_unicode_funcs_t *ufuncs,
494 hb_codepoint_t u,
495 hb_codepoint_t *decomposed);
496
497 HB_END_DECLS
498
499 #endif /* HB_UNICODE_H */