| 1 | /* |
| 2 | * Copyright © 2009,2010 Red Hat, Inc. |
| 3 | * Copyright © 2011,2012 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-machinery.hh" |
| 32 | |
| 33 | #include <locale.h> |
| 34 | #ifdef HAVE_XLOCALE_H |
| 35 | #include <xlocale.h> |
| 36 | #endif |
| 37 | |
| 38 | |
| 39 | /** |
| 40 | * SECTION:hb-common |
| 41 | * @title: hb-common |
| 42 | * @short_description: Common data types |
| 43 | * @include: hb.h |
| 44 | * |
| 45 | * Common data types used across HarfBuzz are defined here. |
| 46 | **/ |
| 47 | |
| 48 | |
| 49 | /* hb_options_t */ |
| 50 | |
| 51 | hb_atomic_int_t _hb_options; |
| 52 | |
| 53 | void |
| 54 | _hb_options_init () |
| 55 | { |
| 56 | hb_options_union_t u; |
| 57 | u.i = 0; |
| 58 | u.opts.initialized = true; |
| 59 | |
| 60 | const char *c = getenv ("HB_OPTIONS" ); |
| 61 | if (c) |
| 62 | { |
| 63 | while (*c) |
| 64 | { |
| 65 | const char *p = strchr (c, ':'); |
| 66 | if (!p) |
| 67 | p = c + strlen (c); |
| 68 | |
| 69 | #define OPTION(name, symbol) \ |
| 70 | if (0 == strncmp (c, name, p - c) && strlen (name) == p - c) u.opts.symbol = true; |
| 71 | |
| 72 | OPTION ("uniscribe-bug-compatible" , uniscribe_bug_compatible); |
| 73 | OPTION ("aat" , aat); |
| 74 | |
| 75 | #undef OPTION |
| 76 | |
| 77 | c = *p ? p + 1 : p; |
| 78 | } |
| 79 | |
| 80 | } |
| 81 | |
| 82 | /* This is idempotent and threadsafe. */ |
| 83 | _hb_options.set_relaxed (u.i); |
| 84 | } |
| 85 | |
| 86 | |
| 87 | /* hb_tag_t */ |
| 88 | |
| 89 | /** |
| 90 | * hb_tag_from_string: |
| 91 | * @str: (array length=len) (element-type uint8_t): |
| 92 | * @len: |
| 93 | * |
| 94 | * |
| 95 | * |
| 96 | * Return value: |
| 97 | * |
| 98 | * Since: 0.9.2 |
| 99 | **/ |
| 100 | hb_tag_t |
| 101 | hb_tag_from_string (const char *str, int len) |
| 102 | { |
| 103 | char tag[4]; |
| 104 | unsigned int i; |
| 105 | |
| 106 | if (!str || !len || !*str) |
| 107 | return HB_TAG_NONE; |
| 108 | |
| 109 | if (len < 0 || len > 4) |
| 110 | len = 4; |
| 111 | for (i = 0; i < (unsigned) len && str[i]; i++) |
| 112 | tag[i] = str[i]; |
| 113 | for (; i < 4; i++) |
| 114 | tag[i] = ' '; |
| 115 | |
| 116 | return HB_TAG (tag[0], tag[1], tag[2], tag[3]); |
| 117 | } |
| 118 | |
| 119 | /** |
| 120 | * hb_tag_to_string: |
| 121 | * @tag: |
| 122 | * @buf: (out caller-allocates) (array fixed-size=4) (element-type uint8_t): |
| 123 | * |
| 124 | * |
| 125 | * |
| 126 | * Since: 0.9.5 |
| 127 | **/ |
| 128 | void |
| 129 | hb_tag_to_string (hb_tag_t tag, char *buf) |
| 130 | { |
| 131 | buf[0] = (char) (uint8_t) (tag >> 24); |
| 132 | buf[1] = (char) (uint8_t) (tag >> 16); |
| 133 | buf[2] = (char) (uint8_t) (tag >> 8); |
| 134 | buf[3] = (char) (uint8_t) (tag >> 0); |
| 135 | } |
| 136 | |
| 137 | |
| 138 | /* hb_direction_t */ |
| 139 | |
| 140 | const char direction_strings[][4] = { |
| 141 | "ltr" , |
| 142 | "rtl" , |
| 143 | "ttb" , |
| 144 | "btt" |
| 145 | }; |
| 146 | |
| 147 | /** |
| 148 | * hb_direction_from_string: |
| 149 | * @str: (array length=len) (element-type uint8_t): |
| 150 | * @len: |
| 151 | * |
| 152 | * |
| 153 | * |
| 154 | * Return value: |
| 155 | * |
| 156 | * Since: 0.9.2 |
| 157 | **/ |
| 158 | hb_direction_t |
| 159 | hb_direction_from_string (const char *str, int len) |
| 160 | { |
| 161 | if (unlikely (!str || !len || !*str)) |
| 162 | return HB_DIRECTION_INVALID; |
| 163 | |
| 164 | /* Lets match loosely: just match the first letter, such that |
| 165 | * all of "ltr", "left-to-right", etc work! |
| 166 | */ |
| 167 | char c = TOLOWER (str[0]); |
| 168 | for (unsigned int i = 0; i < ARRAY_LENGTH (direction_strings); i++) |
| 169 | if (c == direction_strings[i][0]) |
| 170 | return (hb_direction_t) (HB_DIRECTION_LTR + i); |
| 171 | |
| 172 | return HB_DIRECTION_INVALID; |
| 173 | } |
| 174 | |
| 175 | /** |
| 176 | * hb_direction_to_string: |
| 177 | * @direction: |
| 178 | * |
| 179 | * |
| 180 | * |
| 181 | * Return value: (transfer none): |
| 182 | * |
| 183 | * Since: 0.9.2 |
| 184 | **/ |
| 185 | const char * |
| 186 | hb_direction_to_string (hb_direction_t direction) |
| 187 | { |
| 188 | if (likely ((unsigned int) (direction - HB_DIRECTION_LTR) |
| 189 | < ARRAY_LENGTH (direction_strings))) |
| 190 | return direction_strings[direction - HB_DIRECTION_LTR]; |
| 191 | |
| 192 | return "invalid" ; |
| 193 | } |
| 194 | |
| 195 | |
| 196 | /* hb_language_t */ |
| 197 | |
| 198 | struct hb_language_impl_t { |
| 199 | const char s[1]; |
| 200 | }; |
| 201 | |
| 202 | static const char canon_map[256] = { |
| 203 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 204 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
| 205 | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, '-', 0, 0, |
| 206 | '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 0, 0, 0, 0, 0, 0, |
| 207 | 0, 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', |
| 208 | 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 0, 0, 0, 0, '-', |
| 209 | 0, 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', |
| 210 | 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 0, 0, 0, 0, 0 |
| 211 | }; |
| 212 | |
| 213 | static bool |
| 214 | lang_equal (hb_language_t v1, |
| 215 | const void *v2) |
| 216 | { |
| 217 | const unsigned char *p1 = (const unsigned char *) v1; |
| 218 | const unsigned char *p2 = (const unsigned char *) v2; |
| 219 | |
| 220 | while (*p1 && *p1 == canon_map[*p2]) { |
| 221 | p1++; |
| 222 | p2++; |
| 223 | } |
| 224 | |
| 225 | return *p1 == canon_map[*p2]; |
| 226 | } |
| 227 | |
| 228 | #if 0 |
| 229 | static unsigned int |
| 230 | lang_hash (const void *key) |
| 231 | { |
| 232 | const unsigned char *p = key; |
| 233 | unsigned int h = 0; |
| 234 | while (canon_map[*p]) |
| 235 | { |
| 236 | h = (h << 5) - h + canon_map[*p]; |
| 237 | p++; |
| 238 | } |
| 239 | |
| 240 | return h; |
| 241 | } |
| 242 | #endif |
| 243 | |
| 244 | |
| 245 | struct hb_language_item_t { |
| 246 | |
| 247 | struct hb_language_item_t *next; |
| 248 | hb_language_t lang; |
| 249 | |
| 250 | bool operator == (const char *s) const |
| 251 | { return lang_equal (lang, s); } |
| 252 | |
| 253 | hb_language_item_t & operator = (const char *s) { |
| 254 | /* If a custom allocated is used calling strdup() pairs |
| 255 | badly with a call to the custom free() in fini() below. |
| 256 | Therefore don't call strdup(), implement its behavior. |
| 257 | */ |
| 258 | size_t len = strlen(s) + 1; |
| 259 | lang = (hb_language_t) malloc(len); |
| 260 | if (likely (lang)) |
| 261 | { |
| 262 | memcpy((unsigned char *) lang, s, len); |
| 263 | for (unsigned char *p = (unsigned char *) lang; *p; p++) |
| 264 | *p = canon_map[*p]; |
| 265 | } |
| 266 | |
| 267 | return *this; |
| 268 | } |
| 269 | |
| 270 | void fini () { free ((void *) lang); } |
| 271 | }; |
| 272 | |
| 273 | |
| 274 | /* Thread-safe lock-free language list */ |
| 275 | |
| 276 | static hb_atomic_ptr_t <hb_language_item_t> langs; |
| 277 | |
| 278 | #if HB_USE_ATEXIT |
| 279 | static void |
| 280 | free_langs () |
| 281 | { |
| 282 | retry: |
| 283 | hb_language_item_t *first_lang = langs; |
| 284 | if (unlikely (!langs.cmpexch (first_lang, nullptr))) |
| 285 | goto retry; |
| 286 | |
| 287 | while (first_lang) { |
| 288 | hb_language_item_t *next = first_lang->next; |
| 289 | first_lang->fini (); |
| 290 | free (first_lang); |
| 291 | first_lang = next; |
| 292 | } |
| 293 | } |
| 294 | #endif |
| 295 | |
| 296 | static hb_language_item_t * |
| 297 | lang_find_or_insert (const char *key) |
| 298 | { |
| 299 | retry: |
| 300 | hb_language_item_t *first_lang = langs; |
| 301 | |
| 302 | for (hb_language_item_t *lang = first_lang; lang; lang = lang->next) |
| 303 | if (*lang == key) |
| 304 | return lang; |
| 305 | |
| 306 | /* Not found; allocate one. */ |
| 307 | hb_language_item_t *lang = (hb_language_item_t *) calloc (1, sizeof (hb_language_item_t)); |
| 308 | if (unlikely (!lang)) |
| 309 | return nullptr; |
| 310 | lang->next = first_lang; |
| 311 | *lang = key; |
| 312 | if (unlikely (!lang->lang)) |
| 313 | { |
| 314 | free (lang); |
| 315 | return nullptr; |
| 316 | } |
| 317 | |
| 318 | if (unlikely (!langs.cmpexch (first_lang, lang))) |
| 319 | { |
| 320 | lang->fini (); |
| 321 | free (lang); |
| 322 | goto retry; |
| 323 | } |
| 324 | |
| 325 | #if HB_USE_ATEXIT |
| 326 | if (!first_lang) |
| 327 | atexit (free_langs); /* First person registers atexit() callback. */ |
| 328 | #endif |
| 329 | |
| 330 | return lang; |
| 331 | } |
| 332 | |
| 333 | |
| 334 | /** |
| 335 | * hb_language_from_string: |
| 336 | * @str: (array length=len) (element-type uint8_t): a string representing |
| 337 | * a BCP 47 language tag |
| 338 | * @len: length of the @str, or -1 if it is %NULL-terminated. |
| 339 | * |
| 340 | * Converts @str representing a BCP 47 language tag to the corresponding |
| 341 | * #hb_language_t. |
| 342 | * |
| 343 | * Return value: (transfer none): |
| 344 | * The #hb_language_t corresponding to the BCP 47 language tag. |
| 345 | * |
| 346 | * Since: 0.9.2 |
| 347 | **/ |
| 348 | hb_language_t |
| 349 | hb_language_from_string (const char *str, int len) |
| 350 | { |
| 351 | if (!str || !len || !*str) |
| 352 | return HB_LANGUAGE_INVALID; |
| 353 | |
| 354 | hb_language_item_t *item = nullptr; |
| 355 | if (len >= 0) |
| 356 | { |
| 357 | /* NUL-terminate it. */ |
| 358 | char strbuf[64]; |
| 359 | len = MIN (len, (int) sizeof (strbuf) - 1); |
| 360 | memcpy (strbuf, str, len); |
| 361 | strbuf[len] = '\0'; |
| 362 | item = lang_find_or_insert (strbuf); |
| 363 | } |
| 364 | else |
| 365 | item = lang_find_or_insert (str); |
| 366 | |
| 367 | return likely (item) ? item->lang : HB_LANGUAGE_INVALID; |
| 368 | } |
| 369 | |
| 370 | /** |
| 371 | * hb_language_to_string: |
| 372 | * @language: an #hb_language_t to convert. |
| 373 | * |
| 374 | * See hb_language_from_string(). |
| 375 | * |
| 376 | * Return value: (transfer none): |
| 377 | * A %NULL-terminated string representing the @language. Must not be freed by |
| 378 | * the caller. |
| 379 | * |
| 380 | * Since: 0.9.2 |
| 381 | **/ |
| 382 | const char * |
| 383 | hb_language_to_string (hb_language_t language) |
| 384 | { |
| 385 | /* This is actually nullptr-safe! */ |
| 386 | return language->s; |
| 387 | } |
| 388 | |
| 389 | /** |
| 390 | * hb_language_get_default: |
| 391 | * |
| 392 | * Get default language from current locale. |
| 393 | * |
| 394 | * Note that the first time this function is called, it calls |
| 395 | * "setlocale (LC_CTYPE, nullptr)" to fetch current locale. The underlying |
| 396 | * setlocale function is, in many implementations, NOT threadsafe. To avoid |
| 397 | * problems, call this function once before multiple threads can call it. |
| 398 | * This function is only used from hb_buffer_guess_segment_properties() by |
| 399 | * HarfBuzz itself. |
| 400 | * |
| 401 | * Return value: (transfer none): |
| 402 | * |
| 403 | * Since: 0.9.2 |
| 404 | **/ |
| 405 | hb_language_t |
| 406 | hb_language_get_default () |
| 407 | { |
| 408 | static hb_atomic_ptr_t <hb_language_t> default_language; |
| 409 | |
| 410 | hb_language_t language = default_language; |
| 411 | if (unlikely (language == HB_LANGUAGE_INVALID)) |
| 412 | { |
| 413 | language = hb_language_from_string (setlocale (LC_CTYPE, nullptr), -1); |
| 414 | (void) default_language.cmpexch (HB_LANGUAGE_INVALID, language); |
| 415 | } |
| 416 | |
| 417 | return language; |
| 418 | } |
| 419 | |
| 420 | |
| 421 | /* hb_script_t */ |
| 422 | |
| 423 | /** |
| 424 | * hb_script_from_iso15924_tag: |
| 425 | * @tag: an #hb_tag_t representing an ISO 15924 tag. |
| 426 | * |
| 427 | * Converts an ISO 15924 script tag to a corresponding #hb_script_t. |
| 428 | * |
| 429 | * Return value: |
| 430 | * An #hb_script_t corresponding to the ISO 15924 tag. |
| 431 | * |
| 432 | * Since: 0.9.2 |
| 433 | **/ |
| 434 | hb_script_t |
| 435 | hb_script_from_iso15924_tag (hb_tag_t tag) |
| 436 | { |
| 437 | if (unlikely (tag == HB_TAG_NONE)) |
| 438 | return HB_SCRIPT_INVALID; |
| 439 | |
| 440 | /* Be lenient, adjust case (one capital letter followed by three small letters) */ |
| 441 | tag = (tag & 0xDFDFDFDFu) | 0x00202020u; |
| 442 | |
| 443 | switch (tag) { |
| 444 | |
| 445 | /* These graduated from the 'Q' private-area codes, but |
| 446 | * the old code is still aliased by Unicode, and the Qaai |
| 447 | * one in use by ICU. */ |
| 448 | case HB_TAG('Q','a','a','i'): return HB_SCRIPT_INHERITED; |
| 449 | case HB_TAG('Q','a','a','c'): return HB_SCRIPT_COPTIC; |
| 450 | |
| 451 | /* Script variants from https://unicode.org/iso15924/ */ |
| 452 | case HB_TAG('C','y','r','s'): return HB_SCRIPT_CYRILLIC; |
| 453 | case HB_TAG('L','a','t','f'): return HB_SCRIPT_LATIN; |
| 454 | case HB_TAG('L','a','t','g'): return HB_SCRIPT_LATIN; |
| 455 | case HB_TAG('S','y','r','e'): return HB_SCRIPT_SYRIAC; |
| 456 | case HB_TAG('S','y','r','j'): return HB_SCRIPT_SYRIAC; |
| 457 | case HB_TAG('S','y','r','n'): return HB_SCRIPT_SYRIAC; |
| 458 | } |
| 459 | |
| 460 | /* If it looks right, just use the tag as a script */ |
| 461 | if (((uint32_t) tag & 0xE0E0E0E0u) == 0x40606060u) |
| 462 | return (hb_script_t) tag; |
| 463 | |
| 464 | /* Otherwise, return unknown */ |
| 465 | return HB_SCRIPT_UNKNOWN; |
| 466 | } |
| 467 | |
| 468 | /** |
| 469 | * hb_script_from_string: |
| 470 | * @str: (array length=len) (element-type uint8_t): a string representing an |
| 471 | * ISO 15924 tag. |
| 472 | * @len: length of the @str, or -1 if it is %NULL-terminated. |
| 473 | * |
| 474 | * Converts a string @str representing an ISO 15924 script tag to a |
| 475 | * corresponding #hb_script_t. Shorthand for hb_tag_from_string() then |
| 476 | * hb_script_from_iso15924_tag(). |
| 477 | * |
| 478 | * Return value: |
| 479 | * An #hb_script_t corresponding to the ISO 15924 tag. |
| 480 | * |
| 481 | * Since: 0.9.2 |
| 482 | **/ |
| 483 | hb_script_t |
| 484 | hb_script_from_string (const char *str, int len) |
| 485 | { |
| 486 | return hb_script_from_iso15924_tag (hb_tag_from_string (str, len)); |
| 487 | } |
| 488 | |
| 489 | /** |
| 490 | * hb_script_to_iso15924_tag: |
| 491 | * @script: an #hb_script_ to convert. |
| 492 | * |
| 493 | * See hb_script_from_iso15924_tag(). |
| 494 | * |
| 495 | * Return value: |
| 496 | * An #hb_tag_t representing an ISO 15924 script tag. |
| 497 | * |
| 498 | * Since: 0.9.2 |
| 499 | **/ |
| 500 | hb_tag_t |
| 501 | hb_script_to_iso15924_tag (hb_script_t script) |
| 502 | { |
| 503 | return (hb_tag_t) script; |
| 504 | } |
| 505 | |
| 506 | /** |
| 507 | * hb_script_get_horizontal_direction: |
| 508 | * @script: |
| 509 | * |
| 510 | * |
| 511 | * |
| 512 | * Return value: |
| 513 | * |
| 514 | * Since: 0.9.2 |
| 515 | **/ |
| 516 | hb_direction_t |
| 517 | hb_script_get_horizontal_direction (hb_script_t script) |
| 518 | { |
| 519 | /* https://docs.google.com/spreadsheets/d/1Y90M0Ie3MUJ6UVCRDOypOtijlMDLNNyyLk36T6iMu0o */ |
| 520 | switch ((hb_tag_t) script) |
| 521 | { |
| 522 | /* Unicode-1.1 additions */ |
| 523 | case HB_SCRIPT_ARABIC: |
| 524 | case HB_SCRIPT_HEBREW: |
| 525 | |
| 526 | /* Unicode-3.0 additions */ |
| 527 | case HB_SCRIPT_SYRIAC: |
| 528 | case HB_SCRIPT_THAANA: |
| 529 | |
| 530 | /* Unicode-4.0 additions */ |
| 531 | case HB_SCRIPT_CYPRIOT: |
| 532 | |
| 533 | /* Unicode-4.1 additions */ |
| 534 | case HB_SCRIPT_KHAROSHTHI: |
| 535 | |
| 536 | /* Unicode-5.0 additions */ |
| 537 | case HB_SCRIPT_PHOENICIAN: |
| 538 | case HB_SCRIPT_NKO: |
| 539 | |
| 540 | /* Unicode-5.1 additions */ |
| 541 | case HB_SCRIPT_LYDIAN: |
| 542 | |
| 543 | /* Unicode-5.2 additions */ |
| 544 | case HB_SCRIPT_AVESTAN: |
| 545 | case HB_SCRIPT_IMPERIAL_ARAMAIC: |
| 546 | case HB_SCRIPT_INSCRIPTIONAL_PAHLAVI: |
| 547 | case HB_SCRIPT_INSCRIPTIONAL_PARTHIAN: |
| 548 | case HB_SCRIPT_OLD_SOUTH_ARABIAN: |
| 549 | case HB_SCRIPT_OLD_TURKIC: |
| 550 | case HB_SCRIPT_SAMARITAN: |
| 551 | |
| 552 | /* Unicode-6.0 additions */ |
| 553 | case HB_SCRIPT_MANDAIC: |
| 554 | |
| 555 | /* Unicode-6.1 additions */ |
| 556 | case HB_SCRIPT_MEROITIC_CURSIVE: |
| 557 | case HB_SCRIPT_MEROITIC_HIEROGLYPHS: |
| 558 | |
| 559 | /* Unicode-7.0 additions */ |
| 560 | case HB_SCRIPT_MANICHAEAN: |
| 561 | case HB_SCRIPT_MENDE_KIKAKUI: |
| 562 | case HB_SCRIPT_NABATAEAN: |
| 563 | case HB_SCRIPT_OLD_NORTH_ARABIAN: |
| 564 | case HB_SCRIPT_PALMYRENE: |
| 565 | case HB_SCRIPT_PSALTER_PAHLAVI: |
| 566 | |
| 567 | /* Unicode-8.0 additions */ |
| 568 | case HB_SCRIPT_HATRAN: |
| 569 | |
| 570 | /* Unicode-9.0 additions */ |
| 571 | case HB_SCRIPT_ADLAM: |
| 572 | |
| 573 | /* Unicode-11.0 additions */ |
| 574 | case HB_SCRIPT_HANIFI_ROHINGYA: |
| 575 | case HB_SCRIPT_OLD_SOGDIAN: |
| 576 | case HB_SCRIPT_SOGDIAN: |
| 577 | |
| 578 | return HB_DIRECTION_RTL; |
| 579 | |
| 580 | |
| 581 | /* https://github.com/harfbuzz/harfbuzz/issues/1000 */ |
| 582 | case HB_SCRIPT_OLD_HUNGARIAN: |
| 583 | case HB_SCRIPT_OLD_ITALIC: |
| 584 | case HB_SCRIPT_RUNIC: |
| 585 | |
| 586 | return HB_DIRECTION_INVALID; |
| 587 | } |
| 588 | |
| 589 | return HB_DIRECTION_LTR; |
| 590 | } |
| 591 | |
| 592 | |
| 593 | /* hb_user_data_array_t */ |
| 594 | |
| 595 | bool |
| 596 | hb_user_data_array_t::set (hb_user_data_key_t *key, |
| 597 | void * data, |
| 598 | hb_destroy_func_t destroy, |
| 599 | hb_bool_t replace) |
| 600 | { |
| 601 | if (!key) |
| 602 | return false; |
| 603 | |
| 604 | if (replace) { |
| 605 | if (!data && !destroy) { |
| 606 | items.remove (key, lock); |
| 607 | return true; |
| 608 | } |
| 609 | } |
| 610 | hb_user_data_item_t item = {key, data, destroy}; |
| 611 | bool ret = !!items.replace_or_insert (item, lock, (bool) replace); |
| 612 | |
| 613 | return ret; |
| 614 | } |
| 615 | |
| 616 | void * |
| 617 | hb_user_data_array_t::get (hb_user_data_key_t *key) |
| 618 | { |
| 619 | hb_user_data_item_t item = {nullptr, nullptr, nullptr}; |
| 620 | |
| 621 | return items.find (key, &item, lock) ? item.data : nullptr; |
| 622 | } |
| 623 | |
| 624 | |
| 625 | /* hb_version */ |
| 626 | |
| 627 | |
| 628 | /** |
| 629 | * SECTION:hb-version |
| 630 | * @title: hb-version |
| 631 | * @short_description: Information about the version of HarfBuzz in use |
| 632 | * @include: hb.h |
| 633 | * |
| 634 | * These functions and macros allow accessing version of the HarfBuzz |
| 635 | * library used at compile- as well as run-time, and to direct code |
| 636 | * conditionally based on those versions, again, at compile- or run-time. |
| 637 | **/ |
| 638 | |
| 639 | |
| 640 | /** |
| 641 | * hb_version: |
| 642 | * @major: (out): Library major version component. |
| 643 | * @minor: (out): Library minor version component. |
| 644 | * @micro: (out): Library micro version component. |
| 645 | * |
| 646 | * Returns library version as three integer components. |
| 647 | * |
| 648 | * Since: 0.9.2 |
| 649 | **/ |
| 650 | void |
| 651 | hb_version (unsigned int *major, |
| 652 | unsigned int *minor, |
| 653 | unsigned int *micro) |
| 654 | { |
| 655 | *major = HB_VERSION_MAJOR; |
| 656 | *minor = HB_VERSION_MINOR; |
| 657 | *micro = HB_VERSION_MICRO; |
| 658 | } |
| 659 | |
| 660 | /** |
| 661 | * hb_version_string: |
| 662 | * |
| 663 | * Returns library version as a string with three components. |
| 664 | * |
| 665 | * Return value: library version string. |
| 666 | * |
| 667 | * Since: 0.9.2 |
| 668 | **/ |
| 669 | const char * |
| 670 | hb_version_string () |
| 671 | { |
| 672 | return HB_VERSION_STRING; |
| 673 | } |
| 674 | |
| 675 | /** |
| 676 | * hb_version_atleast: |
| 677 | * @major: |
| 678 | * @minor: |
| 679 | * @micro: |
| 680 | * |
| 681 | * |
| 682 | * |
| 683 | * Return value: |
| 684 | * |
| 685 | * Since: 0.9.30 |
| 686 | **/ |
| 687 | hb_bool_t |
| 688 | hb_version_atleast (unsigned int major, |
| 689 | unsigned int minor, |
| 690 | unsigned int micro) |
| 691 | { |
| 692 | return HB_VERSION_ATLEAST (major, minor, micro); |
| 693 | } |
| 694 | |
| 695 | |
| 696 | |
| 697 | /* hb_feature_t and hb_variation_t */ |
| 698 | |
| 699 | static bool |
| 700 | parse_space (const char **pp, const char *end) |
| 701 | { |
| 702 | while (*pp < end && ISSPACE (**pp)) |
| 703 | (*pp)++; |
| 704 | return true; |
| 705 | } |
| 706 | |
| 707 | static bool |
| 708 | parse_char (const char **pp, const char *end, char c) |
| 709 | { |
| 710 | parse_space (pp, end); |
| 711 | |
| 712 | if (*pp == end || **pp != c) |
| 713 | return false; |
| 714 | |
| 715 | (*pp)++; |
| 716 | return true; |
| 717 | } |
| 718 | |
| 719 | static bool |
| 720 | parse_uint (const char **pp, const char *end, unsigned int *pv) |
| 721 | { |
| 722 | char buf[32]; |
| 723 | unsigned int len = MIN (ARRAY_LENGTH (buf) - 1, (unsigned int) (end - *pp)); |
| 724 | strncpy (buf, *pp, len); |
| 725 | buf[len] = '\0'; |
| 726 | |
| 727 | char *p = buf; |
| 728 | char *pend = p; |
| 729 | unsigned int v; |
| 730 | |
| 731 | /* Intentionally use strtol instead of strtoul, such that |
| 732 | * -1 turns into "big number"... */ |
| 733 | errno = 0; |
| 734 | v = strtol (p, &pend, 0); |
| 735 | if (errno || p == pend) |
| 736 | return false; |
| 737 | |
| 738 | *pv = v; |
| 739 | *pp += pend - p; |
| 740 | return true; |
| 741 | } |
| 742 | |
| 743 | static bool |
| 744 | parse_uint32 (const char **pp, const char *end, uint32_t *pv) |
| 745 | { |
| 746 | char buf[32]; |
| 747 | unsigned int len = MIN (ARRAY_LENGTH (buf) - 1, (unsigned int) (end - *pp)); |
| 748 | strncpy (buf, *pp, len); |
| 749 | buf[len] = '\0'; |
| 750 | |
| 751 | char *p = buf; |
| 752 | char *pend = p; |
| 753 | unsigned int v; |
| 754 | |
| 755 | /* Intentionally use strtol instead of strtoul, such that |
| 756 | * -1 turns into "big number"... */ |
| 757 | errno = 0; |
| 758 | v = strtol (p, &pend, 0); |
| 759 | if (errno || p == pend) |
| 760 | return false; |
| 761 | |
| 762 | *pv = v; |
| 763 | *pp += pend - p; |
| 764 | return true; |
| 765 | } |
| 766 | |
| 767 | #if defined (HAVE_NEWLOCALE) && defined (HAVE_STRTOD_L) |
| 768 | #define USE_XLOCALE 1 |
| 769 | #define HB_LOCALE_T locale_t |
| 770 | #define HB_CREATE_LOCALE(locName) newlocale (LC_ALL_MASK, locName, nullptr) |
| 771 | #define HB_FREE_LOCALE(loc) freelocale (loc) |
| 772 | #elif defined(_MSC_VER) |
| 773 | #define USE_XLOCALE 1 |
| 774 | #define HB_LOCALE_T _locale_t |
| 775 | #define HB_CREATE_LOCALE(locName) _create_locale (LC_ALL, locName) |
| 776 | #define HB_FREE_LOCALE(loc) _free_locale (loc) |
| 777 | #define strtod_l(a, b, c) _strtod_l ((a), (b), (c)) |
| 778 | #endif |
| 779 | |
| 780 | #ifdef USE_XLOCALE |
| 781 | |
| 782 | #if HB_USE_ATEXIT |
| 783 | static void free_static_C_locale (); |
| 784 | #endif |
| 785 | |
| 786 | static struct hb_C_locale_lazy_loader_t : hb_lazy_loader_t<hb_remove_pointer (HB_LOCALE_T), |
| 787 | hb_C_locale_lazy_loader_t> |
| 788 | { |
| 789 | static HB_LOCALE_T create () |
| 790 | { |
| 791 | HB_LOCALE_T C_locale = HB_CREATE_LOCALE ("C" ); |
| 792 | |
| 793 | #if HB_USE_ATEXIT |
| 794 | atexit (free_static_C_locale); |
| 795 | #endif |
| 796 | |
| 797 | return C_locale; |
| 798 | } |
| 799 | static void destroy (HB_LOCALE_T p) |
| 800 | { |
| 801 | HB_FREE_LOCALE (p); |
| 802 | } |
| 803 | static HB_LOCALE_T get_null () |
| 804 | { |
| 805 | return nullptr; |
| 806 | } |
| 807 | } static_C_locale; |
| 808 | |
| 809 | #if HB_USE_ATEXIT |
| 810 | static |
| 811 | void free_static_C_locale () |
| 812 | { |
| 813 | static_C_locale.free_instance (); |
| 814 | } |
| 815 | #endif |
| 816 | |
| 817 | static HB_LOCALE_T |
| 818 | get_C_locale () |
| 819 | { |
| 820 | return static_C_locale.get_unconst (); |
| 821 | } |
| 822 | #endif /* USE_XLOCALE */ |
| 823 | |
| 824 | static bool |
| 825 | parse_float (const char **pp, const char *end, float *pv) |
| 826 | { |
| 827 | char buf[32]; |
| 828 | unsigned int len = MIN (ARRAY_LENGTH (buf) - 1, (unsigned int) (end - *pp)); |
| 829 | strncpy (buf, *pp, len); |
| 830 | buf[len] = '\0'; |
| 831 | |
| 832 | char *p = buf; |
| 833 | char *pend = p; |
| 834 | float v; |
| 835 | |
| 836 | errno = 0; |
| 837 | #ifdef USE_XLOCALE |
| 838 | v = strtod_l (p, &pend, get_C_locale ()); |
| 839 | #else |
| 840 | v = strtod (p, &pend); |
| 841 | #endif |
| 842 | if (errno || p == pend) |
| 843 | return false; |
| 844 | |
| 845 | *pv = v; |
| 846 | *pp += pend - p; |
| 847 | return true; |
| 848 | } |
| 849 | |
| 850 | static bool |
| 851 | parse_bool (const char **pp, const char *end, uint32_t *pv) |
| 852 | { |
| 853 | parse_space (pp, end); |
| 854 | |
| 855 | const char *p = *pp; |
| 856 | while (*pp < end && ISALPHA(**pp)) |
| 857 | (*pp)++; |
| 858 | |
| 859 | /* CSS allows on/off as aliases 1/0. */ |
| 860 | if (*pp - p == 2 && 0 == strncmp (p, "on" , 2)) |
| 861 | *pv = 1; |
| 862 | else if (*pp - p == 3 && 0 == strncmp (p, "off" , 3)) |
| 863 | *pv = 0; |
| 864 | else |
| 865 | return false; |
| 866 | |
| 867 | return true; |
| 868 | } |
| 869 | |
| 870 | /* hb_feature_t */ |
| 871 | |
| 872 | static bool |
| 873 | parse_feature_value_prefix (const char **pp, const char *end, hb_feature_t *feature) |
| 874 | { |
| 875 | if (parse_char (pp, end, '-')) |
| 876 | feature->value = 0; |
| 877 | else { |
| 878 | parse_char (pp, end, '+'); |
| 879 | feature->value = 1; |
| 880 | } |
| 881 | |
| 882 | return true; |
| 883 | } |
| 884 | |
| 885 | static bool |
| 886 | parse_tag (const char **pp, const char *end, hb_tag_t *tag) |
| 887 | { |
| 888 | parse_space (pp, end); |
| 889 | |
| 890 | char quote = 0; |
| 891 | |
| 892 | if (*pp < end && (**pp == '\'' || **pp == '"')) |
| 893 | { |
| 894 | quote = **pp; |
| 895 | (*pp)++; |
| 896 | } |
| 897 | |
| 898 | const char *p = *pp; |
| 899 | while (*pp < end && (ISALNUM(**pp) || **pp == '_')) |
| 900 | (*pp)++; |
| 901 | |
| 902 | if (p == *pp || *pp - p > 4) |
| 903 | return false; |
| 904 | |
| 905 | *tag = hb_tag_from_string (p, *pp - p); |
| 906 | |
| 907 | if (quote) |
| 908 | { |
| 909 | /* CSS expects exactly four bytes. And we only allow quotations for |
| 910 | * CSS compatibility. So, enforce the length. */ |
| 911 | if (*pp - p != 4) |
| 912 | return false; |
| 913 | if (*pp == end || **pp != quote) |
| 914 | return false; |
| 915 | (*pp)++; |
| 916 | } |
| 917 | |
| 918 | return true; |
| 919 | } |
| 920 | |
| 921 | static bool |
| 922 | parse_feature_indices (const char **pp, const char *end, hb_feature_t *feature) |
| 923 | { |
| 924 | parse_space (pp, end); |
| 925 | |
| 926 | bool has_start; |
| 927 | |
| 928 | feature->start = HB_FEATURE_GLOBAL_START; |
| 929 | feature->end = HB_FEATURE_GLOBAL_END; |
| 930 | |
| 931 | if (!parse_char (pp, end, '[')) |
| 932 | return true; |
| 933 | |
| 934 | has_start = parse_uint (pp, end, &feature->start); |
| 935 | |
| 936 | if (parse_char (pp, end, ':') || parse_char (pp, end, ';')) { |
| 937 | parse_uint (pp, end, &feature->end); |
| 938 | } else { |
| 939 | if (has_start) |
| 940 | feature->end = feature->start + 1; |
| 941 | } |
| 942 | |
| 943 | return parse_char (pp, end, ']'); |
| 944 | } |
| 945 | |
| 946 | static bool |
| 947 | parse_feature_value_postfix (const char **pp, const char *end, hb_feature_t *feature) |
| 948 | { |
| 949 | bool had_equal = parse_char (pp, end, '='); |
| 950 | bool had_value = parse_uint32 (pp, end, &feature->value) || |
| 951 | parse_bool (pp, end, &feature->value); |
| 952 | /* CSS doesn't use equal-sign between tag and value. |
| 953 | * If there was an equal-sign, then there *must* be a value. |
| 954 | * A value without an equal-sign is ok, but not required. */ |
| 955 | return !had_equal || had_value; |
| 956 | } |
| 957 | |
| 958 | static bool |
| 959 | parse_one_feature (const char **pp, const char *end, hb_feature_t *feature) |
| 960 | { |
| 961 | return parse_feature_value_prefix (pp, end, feature) && |
| 962 | parse_tag (pp, end, &feature->tag) && |
| 963 | parse_feature_indices (pp, end, feature) && |
| 964 | parse_feature_value_postfix (pp, end, feature) && |
| 965 | parse_space (pp, end) && |
| 966 | *pp == end; |
| 967 | } |
| 968 | |
| 969 | /** |
| 970 | * hb_feature_from_string: |
| 971 | * @str: (array length=len) (element-type uint8_t): a string to parse |
| 972 | * @len: length of @str, or -1 if string is %NULL terminated |
| 973 | * @feature: (out): the #hb_feature_t to initialize with the parsed values |
| 974 | * |
| 975 | * Parses a string into a #hb_feature_t. |
| 976 | * |
| 977 | * TODO: document the syntax here. |
| 978 | * |
| 979 | * Return value: |
| 980 | * %true if @str is successfully parsed, %false otherwise. |
| 981 | * |
| 982 | * Since: 0.9.5 |
| 983 | **/ |
| 984 | hb_bool_t |
| 985 | hb_feature_from_string (const char *str, int len, |
| 986 | hb_feature_t *feature) |
| 987 | { |
| 988 | hb_feature_t feat; |
| 989 | |
| 990 | if (len < 0) |
| 991 | len = strlen (str); |
| 992 | |
| 993 | if (likely (parse_one_feature (&str, str + len, &feat))) |
| 994 | { |
| 995 | if (feature) |
| 996 | *feature = feat; |
| 997 | return true; |
| 998 | } |
| 999 | |
| 1000 | if (feature) |
| 1001 | memset (feature, 0, sizeof (*feature)); |
| 1002 | return false; |
| 1003 | } |
| 1004 | |
| 1005 | /** |
| 1006 | * hb_feature_to_string: |
| 1007 | * @feature: an #hb_feature_t to convert |
| 1008 | * @buf: (array length=size) (out): output string |
| 1009 | * @size: the allocated size of @buf |
| 1010 | * |
| 1011 | * Converts a #hb_feature_t into a %NULL-terminated string in the format |
| 1012 | * understood by hb_feature_from_string(). The client in responsible for |
| 1013 | * allocating big enough size for @buf, 128 bytes is more than enough. |
| 1014 | * |
| 1015 | * Since: 0.9.5 |
| 1016 | **/ |
| 1017 | void |
| 1018 | hb_feature_to_string (hb_feature_t *feature, |
| 1019 | char *buf, unsigned int size) |
| 1020 | { |
| 1021 | if (unlikely (!size)) return; |
| 1022 | |
| 1023 | char s[128]; |
| 1024 | unsigned int len = 0; |
| 1025 | if (feature->value == 0) |
| 1026 | s[len++] = '-'; |
| 1027 | hb_tag_to_string (feature->tag, s + len); |
| 1028 | len += 4; |
| 1029 | while (len && s[len - 1] == ' ') |
| 1030 | len--; |
| 1031 | if (feature->start != 0 || feature->end != (unsigned int) -1) |
| 1032 | { |
| 1033 | s[len++] = '['; |
| 1034 | if (feature->start) |
| 1035 | len += MAX (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u" , feature->start)); |
| 1036 | if (feature->end != feature->start + 1) { |
| 1037 | s[len++] = ':'; |
| 1038 | if (feature->end != (unsigned int) -1) |
| 1039 | len += MAX (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u" , feature->end)); |
| 1040 | } |
| 1041 | s[len++] = ']'; |
| 1042 | } |
| 1043 | if (feature->value > 1) |
| 1044 | { |
| 1045 | s[len++] = '='; |
| 1046 | len += MAX (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%u" , feature->value)); |
| 1047 | } |
| 1048 | assert (len < ARRAY_LENGTH (s)); |
| 1049 | len = MIN (len, size - 1); |
| 1050 | memcpy (buf, s, len); |
| 1051 | buf[len] = '\0'; |
| 1052 | } |
| 1053 | |
| 1054 | /* hb_variation_t */ |
| 1055 | |
| 1056 | static bool |
| 1057 | parse_variation_value (const char **pp, const char *end, hb_variation_t *variation) |
| 1058 | { |
| 1059 | parse_char (pp, end, '='); /* Optional. */ |
| 1060 | return parse_float (pp, end, &variation->value); |
| 1061 | } |
| 1062 | |
| 1063 | static bool |
| 1064 | parse_one_variation (const char **pp, const char *end, hb_variation_t *variation) |
| 1065 | { |
| 1066 | return parse_tag (pp, end, &variation->tag) && |
| 1067 | parse_variation_value (pp, end, variation) && |
| 1068 | parse_space (pp, end) && |
| 1069 | *pp == end; |
| 1070 | } |
| 1071 | |
| 1072 | /** |
| 1073 | * hb_variation_from_string: |
| 1074 | * |
| 1075 | * Since: 1.4.2 |
| 1076 | */ |
| 1077 | hb_bool_t |
| 1078 | hb_variation_from_string (const char *str, int len, |
| 1079 | hb_variation_t *variation) |
| 1080 | { |
| 1081 | hb_variation_t var; |
| 1082 | |
| 1083 | if (len < 0) |
| 1084 | len = strlen (str); |
| 1085 | |
| 1086 | if (likely (parse_one_variation (&str, str + len, &var))) |
| 1087 | { |
| 1088 | if (variation) |
| 1089 | *variation = var; |
| 1090 | return true; |
| 1091 | } |
| 1092 | |
| 1093 | if (variation) |
| 1094 | memset (variation, 0, sizeof (*variation)); |
| 1095 | return false; |
| 1096 | } |
| 1097 | |
| 1098 | /** |
| 1099 | * hb_variation_to_string: |
| 1100 | * |
| 1101 | * Since: 1.4.2 |
| 1102 | */ |
| 1103 | void |
| 1104 | hb_variation_to_string (hb_variation_t *variation, |
| 1105 | char *buf, unsigned int size) |
| 1106 | { |
| 1107 | if (unlikely (!size)) return; |
| 1108 | |
| 1109 | char s[128]; |
| 1110 | unsigned int len = 0; |
| 1111 | hb_tag_to_string (variation->tag, s + len); |
| 1112 | len += 4; |
| 1113 | while (len && s[len - 1] == ' ') |
| 1114 | len--; |
| 1115 | s[len++] = '='; |
| 1116 | len += MAX (0, snprintf (s + len, ARRAY_LENGTH (s) - len, "%g" , (double) variation->value)); |
| 1117 | |
| 1118 | assert (len < ARRAY_LENGTH (s)); |
| 1119 | len = MIN (len, size - 1); |
| 1120 | memcpy (buf, s, len); |
| 1121 | buf[len] = '\0'; |
| 1122 | } |
| 1123 | |
| 1124 | /* If there is no visibility control, then hb-static.cc will NOT |
| 1125 | * define anything. Instead, we get it to define one set in here |
| 1126 | * only, so only libharfbuzz.so defines them, not other libs. */ |
| 1127 | #ifdef HB_NO_VISIBILITY |
| 1128 | #undef HB_NO_VISIBILITY |
| 1129 | #include "hb-static.cc" |
| 1130 | #define HB_NO_VISIBILITY 1 |
| 1131 | #endif |
| 1132 | |