| 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 | #ifdef HAVE_GLIB |
| 32 | |
| 33 | #include "hb-glib.h" |
| 34 | |
| 35 | #include "hb-machinery.hh" |
| 36 | |
| 37 | |
| 38 | /** |
| 39 | * SECTION:hb-glib |
| 40 | * @title: hb-glib |
| 41 | * @short_description: GLib integration |
| 42 | * @include: hb-glib.h |
| 43 | * |
| 44 | * Functions for using HarfBuzz with the GLib library to provide Unicode data. |
| 45 | **/ |
| 46 | |
| 47 | |
| 48 | #if !GLIB_CHECK_VERSION(2,29,14) |
| 49 | static const hb_script_t |
| 50 | glib_script_to_script[] = |
| 51 | { |
| 52 | HB_SCRIPT_COMMON, |
| 53 | HB_SCRIPT_INHERITED, |
| 54 | HB_SCRIPT_ARABIC, |
| 55 | HB_SCRIPT_ARMENIAN, |
| 56 | HB_SCRIPT_BENGALI, |
| 57 | HB_SCRIPT_BOPOMOFO, |
| 58 | HB_SCRIPT_CHEROKEE, |
| 59 | HB_SCRIPT_COPTIC, |
| 60 | HB_SCRIPT_CYRILLIC, |
| 61 | HB_SCRIPT_DESERET, |
| 62 | HB_SCRIPT_DEVANAGARI, |
| 63 | HB_SCRIPT_ETHIOPIC, |
| 64 | HB_SCRIPT_GEORGIAN, |
| 65 | HB_SCRIPT_GOTHIC, |
| 66 | HB_SCRIPT_GREEK, |
| 67 | HB_SCRIPT_GUJARATI, |
| 68 | HB_SCRIPT_GURMUKHI, |
| 69 | HB_SCRIPT_HAN, |
| 70 | HB_SCRIPT_HANGUL, |
| 71 | HB_SCRIPT_HEBREW, |
| 72 | HB_SCRIPT_HIRAGANA, |
| 73 | HB_SCRIPT_KANNADA, |
| 74 | HB_SCRIPT_KATAKANA, |
| 75 | HB_SCRIPT_KHMER, |
| 76 | HB_SCRIPT_LAO, |
| 77 | HB_SCRIPT_LATIN, |
| 78 | HB_SCRIPT_MALAYALAM, |
| 79 | HB_SCRIPT_MONGOLIAN, |
| 80 | HB_SCRIPT_MYANMAR, |
| 81 | HB_SCRIPT_OGHAM, |
| 82 | HB_SCRIPT_OLD_ITALIC, |
| 83 | HB_SCRIPT_ORIYA, |
| 84 | HB_SCRIPT_RUNIC, |
| 85 | HB_SCRIPT_SINHALA, |
| 86 | HB_SCRIPT_SYRIAC, |
| 87 | HB_SCRIPT_TAMIL, |
| 88 | HB_SCRIPT_TELUGU, |
| 89 | HB_SCRIPT_THAANA, |
| 90 | HB_SCRIPT_THAI, |
| 91 | HB_SCRIPT_TIBETAN, |
| 92 | HB_SCRIPT_CANADIAN_SYLLABICS, |
| 93 | HB_SCRIPT_YI, |
| 94 | HB_SCRIPT_TAGALOG, |
| 95 | HB_SCRIPT_HANUNOO, |
| 96 | HB_SCRIPT_BUHID, |
| 97 | HB_SCRIPT_TAGBANWA, |
| 98 | |
| 99 | /* Unicode-4.0 additions */ |
| 100 | HB_SCRIPT_BRAILLE, |
| 101 | HB_SCRIPT_CYPRIOT, |
| 102 | HB_SCRIPT_LIMBU, |
| 103 | HB_SCRIPT_OSMANYA, |
| 104 | HB_SCRIPT_SHAVIAN, |
| 105 | HB_SCRIPT_LINEAR_B, |
| 106 | HB_SCRIPT_TAI_LE, |
| 107 | HB_SCRIPT_UGARITIC, |
| 108 | |
| 109 | /* Unicode-4.1 additions */ |
| 110 | HB_SCRIPT_NEW_TAI_LUE, |
| 111 | HB_SCRIPT_BUGINESE, |
| 112 | HB_SCRIPT_GLAGOLITIC, |
| 113 | HB_SCRIPT_TIFINAGH, |
| 114 | HB_SCRIPT_SYLOTI_NAGRI, |
| 115 | HB_SCRIPT_OLD_PERSIAN, |
| 116 | HB_SCRIPT_KHAROSHTHI, |
| 117 | |
| 118 | /* Unicode-5.0 additions */ |
| 119 | HB_SCRIPT_UNKNOWN, |
| 120 | HB_SCRIPT_BALINESE, |
| 121 | HB_SCRIPT_CUNEIFORM, |
| 122 | HB_SCRIPT_PHOENICIAN, |
| 123 | HB_SCRIPT_PHAGS_PA, |
| 124 | HB_SCRIPT_NKO, |
| 125 | |
| 126 | /* Unicode-5.1 additions */ |
| 127 | HB_SCRIPT_KAYAH_LI, |
| 128 | HB_SCRIPT_LEPCHA, |
| 129 | HB_SCRIPT_REJANG, |
| 130 | HB_SCRIPT_SUNDANESE, |
| 131 | HB_SCRIPT_SAURASHTRA, |
| 132 | HB_SCRIPT_CHAM, |
| 133 | HB_SCRIPT_OL_CHIKI, |
| 134 | HB_SCRIPT_VAI, |
| 135 | HB_SCRIPT_CARIAN, |
| 136 | HB_SCRIPT_LYCIAN, |
| 137 | HB_SCRIPT_LYDIAN, |
| 138 | |
| 139 | /* Unicode-5.2 additions */ |
| 140 | HB_SCRIPT_AVESTAN, |
| 141 | HB_SCRIPT_BAMUM, |
| 142 | HB_SCRIPT_EGYPTIAN_HIEROGLYPHS, |
| 143 | HB_SCRIPT_IMPERIAL_ARAMAIC, |
| 144 | HB_SCRIPT_INSCRIPTIONAL_PAHLAVI, |
| 145 | HB_SCRIPT_INSCRIPTIONAL_PARTHIAN, |
| 146 | HB_SCRIPT_JAVANESE, |
| 147 | HB_SCRIPT_KAITHI, |
| 148 | HB_SCRIPT_TAI_THAM, |
| 149 | HB_SCRIPT_LISU, |
| 150 | HB_SCRIPT_MEETEI_MAYEK, |
| 151 | HB_SCRIPT_OLD_SOUTH_ARABIAN, |
| 152 | HB_SCRIPT_OLD_TURKIC, |
| 153 | HB_SCRIPT_SAMARITAN, |
| 154 | HB_SCRIPT_TAI_VIET, |
| 155 | |
| 156 | /* Unicode-6.0 additions */ |
| 157 | HB_SCRIPT_BATAK, |
| 158 | HB_SCRIPT_BRAHMI, |
| 159 | HB_SCRIPT_MANDAIC, |
| 160 | |
| 161 | /* Unicode-6.1 additions */ |
| 162 | HB_SCRIPT_CHAKMA, |
| 163 | HB_SCRIPT_MEROITIC_CURSIVE, |
| 164 | HB_SCRIPT_MEROITIC_HIEROGLYPHS, |
| 165 | HB_SCRIPT_MIAO, |
| 166 | HB_SCRIPT_SHARADA, |
| 167 | HB_SCRIPT_SORA_SOMPENG, |
| 168 | HB_SCRIPT_TAKRI |
| 169 | }; |
| 170 | #endif |
| 171 | |
| 172 | hb_script_t |
| 173 | hb_glib_script_to_script (GUnicodeScript script) |
| 174 | { |
| 175 | #if GLIB_CHECK_VERSION(2,29,14) |
| 176 | return (hb_script_t) g_unicode_script_to_iso15924 (script); |
| 177 | #else |
| 178 | if (likely ((unsigned int) script < ARRAY_LENGTH (glib_script_to_script))) |
| 179 | return glib_script_to_script[script]; |
| 180 | |
| 181 | if (unlikely (script == G_UNICODE_SCRIPT_INVALID_CODE)) |
| 182 | return HB_SCRIPT_INVALID; |
| 183 | |
| 184 | return HB_SCRIPT_UNKNOWN; |
| 185 | #endif |
| 186 | } |
| 187 | |
| 188 | GUnicodeScript |
| 189 | hb_glib_script_from_script (hb_script_t script) |
| 190 | { |
| 191 | #if GLIB_CHECK_VERSION(2,29,14) |
| 192 | return g_unicode_script_from_iso15924 (script); |
| 193 | #else |
| 194 | unsigned int count = ARRAY_LENGTH (glib_script_to_script); |
| 195 | for (unsigned int i = 0; i < count; i++) |
| 196 | if (glib_script_to_script[i] == script) |
| 197 | return (GUnicodeScript) i; |
| 198 | |
| 199 | if (unlikely (script == HB_SCRIPT_INVALID)) |
| 200 | return G_UNICODE_SCRIPT_INVALID_CODE; |
| 201 | |
| 202 | return G_UNICODE_SCRIPT_UNKNOWN; |
| 203 | #endif |
| 204 | } |
| 205 | |
| 206 | |
| 207 | static hb_unicode_combining_class_t |
| 208 | hb_glib_unicode_combining_class (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 209 | hb_codepoint_t unicode, |
| 210 | void *user_data HB_UNUSED) |
| 211 | |
| 212 | { |
| 213 | return (hb_unicode_combining_class_t) g_unichar_combining_class (unicode); |
| 214 | } |
| 215 | |
| 216 | static hb_unicode_general_category_t |
| 217 | hb_glib_unicode_general_category (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 218 | hb_codepoint_t unicode, |
| 219 | void *user_data HB_UNUSED) |
| 220 | |
| 221 | { |
| 222 | /* hb_unicode_general_category_t and GUnicodeType are identical */ |
| 223 | return (hb_unicode_general_category_t) g_unichar_type (unicode); |
| 224 | } |
| 225 | |
| 226 | static hb_codepoint_t |
| 227 | hb_glib_unicode_mirroring (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 228 | hb_codepoint_t unicode, |
| 229 | void *user_data HB_UNUSED) |
| 230 | { |
| 231 | g_unichar_get_mirror_char (unicode, &unicode); |
| 232 | return unicode; |
| 233 | } |
| 234 | |
| 235 | static hb_script_t |
| 236 | hb_glib_unicode_script (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 237 | hb_codepoint_t unicode, |
| 238 | void *user_data HB_UNUSED) |
| 239 | { |
| 240 | return hb_glib_script_to_script (g_unichar_get_script (unicode)); |
| 241 | } |
| 242 | |
| 243 | static hb_bool_t |
| 244 | hb_glib_unicode_compose (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 245 | hb_codepoint_t a, |
| 246 | hb_codepoint_t b, |
| 247 | hb_codepoint_t *ab, |
| 248 | void *user_data HB_UNUSED) |
| 249 | { |
| 250 | #if GLIB_CHECK_VERSION(2,29,12) |
| 251 | return g_unichar_compose (a, b, ab); |
| 252 | #endif |
| 253 | |
| 254 | /* We don't ifdef-out the fallback code such that compiler always |
| 255 | * sees it and makes sure it's compilable. */ |
| 256 | |
| 257 | gchar utf8[12]; |
| 258 | gchar *normalized; |
| 259 | int len; |
| 260 | hb_bool_t ret; |
| 261 | |
| 262 | len = g_unichar_to_utf8 (a, utf8); |
| 263 | len += g_unichar_to_utf8 (b, utf8 + len); |
| 264 | normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFC); |
| 265 | len = g_utf8_strlen (normalized, -1); |
| 266 | if (unlikely (!len)) |
| 267 | return false; |
| 268 | |
| 269 | if (len == 1) { |
| 270 | *ab = g_utf8_get_char (normalized); |
| 271 | ret = true; |
| 272 | } else { |
| 273 | ret = false; |
| 274 | } |
| 275 | |
| 276 | g_free (normalized); |
| 277 | return ret; |
| 278 | } |
| 279 | |
| 280 | static hb_bool_t |
| 281 | hb_glib_unicode_decompose (hb_unicode_funcs_t *ufuncs HB_UNUSED, |
| 282 | hb_codepoint_t ab, |
| 283 | hb_codepoint_t *a, |
| 284 | hb_codepoint_t *b, |
| 285 | void *user_data HB_UNUSED) |
| 286 | { |
| 287 | #if GLIB_CHECK_VERSION(2,29,12) |
| 288 | return g_unichar_decompose (ab, a, b); |
| 289 | #endif |
| 290 | |
| 291 | /* We don't ifdef-out the fallback code such that compiler always |
| 292 | * sees it and makes sure it's compilable. */ |
| 293 | |
| 294 | gchar utf8[6]; |
| 295 | gchar *normalized; |
| 296 | int len; |
| 297 | hb_bool_t ret; |
| 298 | |
| 299 | len = g_unichar_to_utf8 (ab, utf8); |
| 300 | normalized = g_utf8_normalize (utf8, len, G_NORMALIZE_NFD); |
| 301 | len = g_utf8_strlen (normalized, -1); |
| 302 | if (unlikely (!len)) |
| 303 | return false; |
| 304 | |
| 305 | if (len == 1) { |
| 306 | *a = g_utf8_get_char (normalized); |
| 307 | *b = 0; |
| 308 | ret = *a != ab; |
| 309 | } else if (len == 2) { |
| 310 | *a = g_utf8_get_char (normalized); |
| 311 | *b = g_utf8_get_char (g_utf8_next_char (normalized)); |
| 312 | /* Here's the ugly part: if ab decomposes to a single character and |
| 313 | * that character decomposes again, we have to detect that and undo |
| 314 | * the second part :-(. */ |
| 315 | gchar *recomposed = g_utf8_normalize (normalized, -1, G_NORMALIZE_NFC); |
| 316 | hb_codepoint_t c = g_utf8_get_char (recomposed); |
| 317 | if (c != ab && c != *a) { |
| 318 | *a = c; |
| 319 | *b = 0; |
| 320 | } |
| 321 | g_free (recomposed); |
| 322 | ret = true; |
| 323 | } else { |
| 324 | /* If decomposed to more than two characters, take the last one, |
| 325 | * and recompose the rest to get the first component. */ |
| 326 | gchar *end = g_utf8_offset_to_pointer (normalized, len - 1); |
| 327 | gchar *recomposed; |
| 328 | *b = g_utf8_get_char (end); |
| 329 | recomposed = g_utf8_normalize (normalized, end - normalized, G_NORMALIZE_NFC); |
| 330 | /* We expect that recomposed has exactly one character now. */ |
| 331 | *a = g_utf8_get_char (recomposed); |
| 332 | g_free (recomposed); |
| 333 | ret = true; |
| 334 | } |
| 335 | |
| 336 | g_free (normalized); |
| 337 | return ret; |
| 338 | } |
| 339 | |
| 340 | |
| 341 | #if HB_USE_ATEXIT |
| 342 | static void free_static_glib_funcs (); |
| 343 | #endif |
| 344 | |
| 345 | static struct hb_glib_unicode_funcs_lazy_loader_t : hb_unicode_funcs_lazy_loader_t<hb_glib_unicode_funcs_lazy_loader_t> |
| 346 | { |
| 347 | static hb_unicode_funcs_t *create () |
| 348 | { |
| 349 | hb_unicode_funcs_t *funcs = hb_unicode_funcs_create (nullptr); |
| 350 | |
| 351 | hb_unicode_funcs_set_combining_class_func (funcs, hb_glib_unicode_combining_class, nullptr, nullptr); |
| 352 | hb_unicode_funcs_set_general_category_func (funcs, hb_glib_unicode_general_category, nullptr, nullptr); |
| 353 | hb_unicode_funcs_set_mirroring_func (funcs, hb_glib_unicode_mirroring, nullptr, nullptr); |
| 354 | hb_unicode_funcs_set_script_func (funcs, hb_glib_unicode_script, nullptr, nullptr); |
| 355 | hb_unicode_funcs_set_compose_func (funcs, hb_glib_unicode_compose, nullptr, nullptr); |
| 356 | hb_unicode_funcs_set_decompose_func (funcs, hb_glib_unicode_decompose, nullptr, nullptr); |
| 357 | |
| 358 | hb_unicode_funcs_make_immutable (funcs); |
| 359 | |
| 360 | #if HB_USE_ATEXIT |
| 361 | atexit (free_static_glib_funcs); |
| 362 | #endif |
| 363 | |
| 364 | return funcs; |
| 365 | } |
| 366 | } static_glib_funcs; |
| 367 | |
| 368 | #if HB_USE_ATEXIT |
| 369 | static |
| 370 | void free_static_glib_funcs () |
| 371 | { |
| 372 | static_glib_funcs.free_instance (); |
| 373 | } |
| 374 | #endif |
| 375 | |
| 376 | hb_unicode_funcs_t * |
| 377 | hb_glib_get_unicode_funcs () |
| 378 | { |
| 379 | return static_glib_funcs.get_unconst (); |
| 380 | } |
| 381 | |
| 382 | |
| 383 | |
| 384 | #if GLIB_CHECK_VERSION(2,31,10) |
| 385 | |
| 386 | static void |
| 387 | _hb_g_bytes_unref (void *data) |
| 388 | { |
| 389 | g_bytes_unref ((GBytes *) data); |
| 390 | } |
| 391 | |
| 392 | /** |
| 393 | * hb_glib_blob_create: |
| 394 | * |
| 395 | * Since: 0.9.38 |
| 396 | **/ |
| 397 | hb_blob_t * |
| 398 | hb_glib_blob_create (GBytes *gbytes) |
| 399 | { |
| 400 | gsize size = 0; |
| 401 | gconstpointer data = g_bytes_get_data (gbytes, &size); |
| 402 | return hb_blob_create ((const char *) data, |
| 403 | size, |
| 404 | HB_MEMORY_MODE_READONLY, |
| 405 | g_bytes_ref (gbytes), |
| 406 | _hb_g_bytes_unref); |
| 407 | } |
| 408 | #endif |
| 409 | |
| 410 | |
| 411 | #endif |
| 412 | |