1 /*
2 * Copyright © 2009 Red Hat, Inc.
3 * Copyright © 2011 Google, Inc.
4 *
5 * This is part of HarfBuzz, a text shaping library.
6 *
7 * Permission is hereby granted, without written agreement and without
8 * license or royalty fees, to use, copy, modify, and distribute this
9 * software and its documentation for any purpose, provided that the
10 * above copyright notice and the following two paragraphs appear in
11 * all copies of this software.
12 *
13 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
14 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
15 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
16 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
17 * DAMAGE.
18 *
19 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
20 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
21 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
22 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
23 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
24 *
25 * Red Hat Author(s): Behdad Esfahbod
26 * Google Author(s): Behdad Esfahbod
27 */
28
29 #include "hb.hh"
30
31 #include "hb-glib.h"
32
33 #include "hb-machinery.hh"
34
35
36 /**
37 * SECTION:hb-glib
38 * @title: hb-glib
39 * @short_description: GLib integration
40 * @include: hb-glib.h
41 *
42 * Functions for using HarfBuzz with the GLib library to provide Unicode data.
43 **/
44
45
46 #if !GLIB_CHECK_VERSION(2,29,14)
47 static const hb_script_t
48 glib_script_to_script[] =
49 {
50 HB_SCRIPT_COMMON,
51 HB_SCRIPT_INHERITED,
52 HB_SCRIPT_ARABIC,
53 HB_SCRIPT_ARMENIAN,
54 HB_SCRIPT_BENGALI,
55 HB_SCRIPT_BOPOMOFO,
56 HB_SCRIPT_CHEROKEE,
57 HB_SCRIPT_COPTIC,
58 HB_SCRIPT_CYRILLIC,
59 HB_SCRIPT_DESERET,
60 HB_SCRIPT_DEVANAGARI,
61 HB_SCRIPT_ETHIOPIC,
62 HB_SCRIPT_GEORGIAN,
63 HB_SCRIPT_GOTHIC,
64 HB_SCRIPT_GREEK,
65 HB_SCRIPT_GUJARATI,
66 HB_SCRIPT_GURMUKHI,
67 HB_SCRIPT_HAN,
68 HB_SCRIPT_HANGUL,
69 HB_SCRIPT_HEBREW,
70 HB_SCRIPT_HIRAGANA,
71 HB_SCRIPT_KANNADA,
72 HB_SCRIPT_KATAKANA,
73 HB_SCRIPT_KHMER,
74 HB_SCRIPT_LAO,
75 HB_SCRIPT_LATIN,
76 HB_SCRIPT_MALAYALAM,
77 HB_SCRIPT_MONGOLIAN,
78 HB_SCRIPT_MYANMAR,
79 HB_SCRIPT_OGHAM,
80 HB_SCRIPT_OLD_ITALIC,
81 HB_SCRIPT_ORIYA,
82 HB_SCRIPT_RUNIC,
83 HB_SCRIPT_SINHALA,
84 HB_SCRIPT_SYRIAC,
85 HB_SCRIPT_TAMIL,
86 HB_SCRIPT_TELUGU,
87 HB_SCRIPT_THAANA,
88 HB_SCRIPT_THAI,
89 HB_SCRIPT_TIBETAN,
90 HB_SCRIPT_CANADIAN_SYLLABICS,
91 HB_SCRIPT_YI,
92 HB_SCRIPT_TAGALOG,
93 HB_SCRIPT_HANUNOO,
94 HB_SCRIPT_BUHID,
95 HB_SCRIPT_TAGBANWA,
96
97 /* Unicode-4.0 additions */
98 HB_SCRIPT_BRAILLE,
99 HB_SCRIPT_CYPRIOT,
100 HB_SCRIPT_LIMBU,
101 HB_SCRIPT_OSMANYA,
102 HB_SCRIPT_SHAVIAN,
103 HB_SCRIPT_LINEAR_B,
104 HB_SCRIPT_TAI_LE,
105 HB_SCRIPT_UGARITIC,
106
107 /* Unicode-4.1 additions */
108 HB_SCRIPT_NEW_TAI_LUE,
109 HB_SCRIPT_BUGINESE,
110 HB_SCRIPT_GLAGOLITIC,
111 HB_SCRIPT_TIFINAGH,
112 HB_SCRIPT_SYLOTI_NAGRI,
113 HB_SCRIPT_OLD_PERSIAN,
114 HB_SCRIPT_KHAROSHTHI,
115
116 /* Unicode-5.0 additions */
117 HB_SCRIPT_UNKNOWN,
118 HB_SCRIPT_BALINESE,
119 HB_SCRIPT_CUNEIFORM,
120 HB_SCRIPT_PHOENICIAN,
121 HB_SCRIPT_PHAGS_PA,
122 HB_SCRIPT_NKO,
123
124 /* Unicode-5.1 additions */
125 HB_SCRIPT_KAYAH_LI,
126 HB_SCRIPT_LEPCHA,
127 HB_SCRIPT_REJANG,
128 HB_SCRIPT_SUNDANESE,
129 HB_SCRIPT_SAURASHTRA,
130 HB_SCRIPT_CHAM,
131 HB_SCRIPT_OL_CHIKI,
132 HB_SCRIPT_VAI,
133 HB_SCRIPT_CARIAN,
134 HB_SCRIPT_LYCIAN,
135 HB_SCRIPT_LYDIAN,
136
137 /* Unicode-5.2 additions */
138 HB_SCRIPT_AVESTAN,
139 HB_SCRIPT_BAMUM,
140 HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,
141 HB_SCRIPT_IMPERIAL_ARAMAIC,
142 HB_SCRIPT_INSCRIPTIONAL_PAHLAVI,
143 HB_SCRIPT_INSCRIPTIONAL_PARTHIAN,
144 HB_SCRIPT_JAVANESE,
145 HB_SCRIPT_KAITHI,
146 HB_SCRIPT_TAI_THAM,
147 HB_SCRIPT_LISU,
148 HB_SCRIPT_MEETEI_MAYEK,
149 HB_SCRIPT_OLD_SOUTH_ARABIAN,
150 HB_SCRIPT_OLD_TURKIC,
151 HB_SCRIPT_SAMARITAN,
152 HB_SCRIPT_TAI_VIET,
153
154 /* Unicode-6.0 additions */
155 HB_SCRIPT_BATAK,
156 HB_SCRIPT_BRAHMI,
157 HB_SCRIPT_MANDAIC,
158
159 /* Unicode-6.1 additions */
160 HB_SCRIPT_CHAKMA,
161 HB_SCRIPT_MEROITIC_CURSIVE,
162 HB_SCRIPT_MEROITIC_HIEROGLYPHS,
163 HB_SCRIPT_MIAO,
164 HB_SCRIPT_SHARADA,
165 HB_SCRIPT_SORA_SOMPENG,
166 HB_SCRIPT_TAKRI
167 };
168 #endif
169
170 hb_script_t
hb_glib_script_to_script(GUnicodeScript script)171 hb_glib_script_to_script (GUnicodeScript script)
172 {
173 #if GLIB_CHECK_VERSION(2,29,14)
174 return (hb_script_t) g_unicode_script_to_iso15924 (script);
175 #else
176 if (likely ((unsigned int) script < ARRAY_LENGTH (glib_script_to_script)))
177 return glib_script_to_script[script];
178
179 if (unlikely (script == G_UNICODE_SCRIPT_INVALID_CODE))
180 return HB_SCRIPT_INVALID;
181
182 return HB_SCRIPT_UNKNOWN;
183 #endif
184 }
185
186 GUnicodeScript
hb_glib_script_from_script(hb_script_t script)187 hb_glib_script_from_script (hb_script_t script)
188 {
189 #if GLIB_CHECK_VERSION(2,29,14)
190 return g_unicode_script_from_iso15924 (script);
191 #else
192 unsigned int count = ARRAY_LENGTH (glib_script_to_script);
193 for (unsigned int i = 0; i < count; i++)
194 if (glib_script_to_script[i] == script)
195 return (GUnicodeScript) i;
196
197 if (unlikely (script == HB_SCRIPT_INVALID))
198 return G_UNICODE_SCRIPT_INVALID_CODE;
199
200 return G_UNICODE_SCRIPT_UNKNOWN;
201 #endif
202 }
203
204
205 static hb_unicode_combining_class_t
hb_glib_unicode_combining_class(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t unicode,void * user_data HB_UNUSED)206 hb_glib_unicode_combining_class (hb_unicode_funcs_t *ufuncs HB_UNUSED,
207 hb_codepoint_t unicode,
208 void *user_data HB_UNUSED)
209
210 {
211 return (hb_unicode_combining_class_t) g_unichar_combining_class (unicode);
212 }
213
214 static hb_unicode_general_category_t
hb_glib_unicode_general_category(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t unicode,void * user_data HB_UNUSED)215 hb_glib_unicode_general_category (hb_unicode_funcs_t *ufuncs HB_UNUSED,
216 hb_codepoint_t unicode,
217 void *user_data HB_UNUSED)
218
219 {
220 /* hb_unicode_general_category_t and GUnicodeType are identical */
221 return (hb_unicode_general_category_t) g_unichar_type (unicode);
222 }
223
224 static hb_codepoint_t
hb_glib_unicode_mirroring(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t unicode,void * user_data HB_UNUSED)225 hb_glib_unicode_mirroring (hb_unicode_funcs_t *ufuncs HB_UNUSED,
226 hb_codepoint_t unicode,
227 void *user_data HB_UNUSED)
228 {
229 g_unichar_get_mirror_char (unicode, &unicode);
230 return unicode;
231 }
232
233 static hb_script_t
hb_glib_unicode_script(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t unicode,void * user_data HB_UNUSED)234 hb_glib_unicode_script (hb_unicode_funcs_t *ufuncs HB_UNUSED,
235 hb_codepoint_t unicode,
236 void *user_data HB_UNUSED)
237 {
238 return hb_glib_script_to_script (g_unichar_get_script (unicode));
239 }
240
241 static hb_bool_t
hb_glib_unicode_compose(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t a,hb_codepoint_t b,hb_codepoint_t * ab,void * user_data HB_UNUSED)242 hb_glib_unicode_compose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
243 hb_codepoint_t a,
244 hb_codepoint_t b,
245 hb_codepoint_t *ab,
246 void *user_data HB_UNUSED)
247 {
248 #if GLIB_CHECK_VERSION(2,29,12)
249 return g_unichar_compose (a, b, ab);
250 #endif
251
252 /* We don't ifdef-out the fallback code such that compiler always
253 * sees it and makes sure it's compilable. */
254
255 gchar utf8[12];
256 gchar *normalized;
257 int len;
258 hb_bool_t ret;
259
260 len = g_unichar_to_utf8 (a, utf8);
261 len += g_unichar_to_utf8 (b, utf8 + len);
262 normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFC);
263 len = g_utf8_strlen (normalized, -1);
264 if (unlikely (!len))
265 return false;
266
267 if (len == 1) {
268 *ab = g_utf8_get_char (normalized);
269 ret = true;
270 } else {
271 ret = false;
272 }
273
274 g_free (normalized);
275 return ret;
276 }
277
278 static hb_bool_t
hb_glib_unicode_decompose(hb_unicode_funcs_t * ufuncs HB_UNUSED,hb_codepoint_t ab,hb_codepoint_t * a,hb_codepoint_t * b,void * user_data HB_UNUSED)279 hb_glib_unicode_decompose (hb_unicode_funcs_t *ufuncs HB_UNUSED,
280 hb_codepoint_t ab,
281 hb_codepoint_t *a,
282 hb_codepoint_t *b,
283 void *user_data HB_UNUSED)
284 {
285 #if GLIB_CHECK_VERSION(2,29,12)
286 return g_unichar_decompose (ab, a, b);
287 #endif
288
289 /* We don't ifdef-out the fallback code such that compiler always
290 * sees it and makes sure it's compilable. */
291
292 gchar utf8[6];
293 gchar *normalized;
294 int len;
295 hb_bool_t ret;
296
297 len = g_unichar_to_utf8 (ab, utf8);
298 normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFD);
299 len = g_utf8_strlen (normalized, -1);
300 if (unlikely (!len))
301 return false;
302
303 if (len == 1) {
304 *a = g_utf8_get_char (normalized);
305 *b = 0;
306 ret = *a != ab;
307 } else if (len == 2) {
308 *a = g_utf8_get_char (normalized);
309 *b = g_utf8_get_char (g_utf8_next_char (normalized));
310 /* Here's the ugly part: if ab decomposes to a single character and
311 * that character decomposes again, we have to detect that and undo
312 * the second part :-(. */
313 gchar *recomposed = g_utf8_normalize (normalized, -1, G_NORMALIZE_NFC);
314 hb_codepoint_t c = g_utf8_get_char (recomposed);
315 if (c != ab && c != *a) {
316 *a = c;
317 *b = 0;
318 }
319 g_free (recomposed);
320 ret = true;
321 } else {
322 /* If decomposed to more than two characters, take the last one,
323 * and recompose the rest to get the first component. */
324 gchar *end = g_utf8_offset_to_pointer (normalized, len - 1);
325 gchar *recomposed;
326 *b = g_utf8_get_char (end);
327 recomposed = g_utf8_normalize (normalized, end - normalized, G_NORMALIZE_NFC);
328 /* We expect that recomposed has exactly one character now. */
329 *a = g_utf8_get_char (recomposed);
330 g_free (recomposed);
331 ret = true;
332 }
333
334 g_free (normalized);
335 return ret;
336 }
337
338
339 #if HB_USE_ATEXIT
340 static void free_static_glib_funcs ();
341 #endif
342
343 static struct hb_glib_unicode_funcs_lazy_loader_t : hb_unicode_funcs_lazy_loader_t<hb_glib_unicode_funcs_lazy_loader_t>
344 {
createhb_glib_unicode_funcs_lazy_loader_t345 static hb_unicode_funcs_t *create ()
346 {
347 hb_unicode_funcs_t *funcs = hb_unicode_funcs_create (nullptr);
348
349 hb_unicode_funcs_set_combining_class_func (funcs, hb_glib_unicode_combining_class, nullptr, nullptr);
350 hb_unicode_funcs_set_general_category_func (funcs, hb_glib_unicode_general_category, nullptr, nullptr);
351 hb_unicode_funcs_set_mirroring_func (funcs, hb_glib_unicode_mirroring, nullptr, nullptr);
352 hb_unicode_funcs_set_script_func (funcs, hb_glib_unicode_script, nullptr, nullptr);
353 hb_unicode_funcs_set_compose_func (funcs, hb_glib_unicode_compose, nullptr, nullptr);
354 hb_unicode_funcs_set_decompose_func (funcs, hb_glib_unicode_decompose, nullptr, nullptr);
355
356 hb_unicode_funcs_make_immutable (funcs);
357
358 #if HB_USE_ATEXIT
359 atexit (free_static_glib_funcs);
360 #endif
361
362 return funcs;
363 }
364 } static_glib_funcs;
365
366 #if HB_USE_ATEXIT
367 static
free_static_glib_funcs()368 void free_static_glib_funcs ()
369 {
370 static_glib_funcs.free_instance ();
371 }
372 #endif
373
374 hb_unicode_funcs_t *
hb_glib_get_unicode_funcs()375 hb_glib_get_unicode_funcs ()
376 {
377 return static_glib_funcs.get_unconst ();
378 }
379
380
381
382 #if GLIB_CHECK_VERSION(2,31,10)
383
384 static void
_hb_g_bytes_unref(void * data)385 _hb_g_bytes_unref (void *data)
386 {
387 g_bytes_unref ((GBytes *) data);
388 }
389
390 /**
391 * hb_glib_blob_create:
392 *
393 * Since: 0.9.38
394 **/
395 hb_blob_t *
hb_glib_blob_create(GBytes * gbytes)396 hb_glib_blob_create (GBytes *gbytes)
397 {
398 gsize size = 0;
399 gconstpointer data = g_bytes_get_data (gbytes, &size);
400 return hb_blob_create ((const char *) data,
401 size,
402 HB_MEMORY_MODE_READONLY,
403 g_bytes_ref (gbytes),
404 _hb_g_bytes_unref);
405 }
406 #endif
407