test-buffer.c revision 82ecaff736e245e117d70b6ec1497508c6eb08d2
1/* 2 * Copyright © 2011 Google, Inc. 3 * 4 * This is part of HarfBuzz, a text shaping library. 5 * 6 * Permission is hereby granted, without written agreement and without 7 * license or royalty fees, to use, copy, modify, and distribute this 8 * software and its documentation for any purpose, provided that the 9 * above copyright notice and the following two paragraphs appear in 10 * all copies of this software. 11 * 12 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR 13 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES 14 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN 15 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH 16 * DAMAGE. 17 * 18 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, 19 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND 20 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS 21 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO 22 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS. 23 * 24 * Google Author(s): Behdad Esfahbod 25 */ 26 27#include "hb-test.h" 28 29/* Unit tests for hb-buffer.h */ 30 31 32static const char utf8[10] = "ab\360\240\200\200defg"; 33static const uint16_t utf16[8] = {'a', 'b', 0xD840, 0xDC00, 'd', 'e', 'f', 'g'}; 34static const uint32_t utf32[7] = {'a', 'b', 0x20000, 'd', 'e', 'f', 'g'}; 35 36 37typedef enum { 38 BUFFER_EMPTY, 39 BUFFER_ONE_BY_ONE, 40 BUFFER_UTF32, 41 BUFFER_UTF16, 42 BUFFER_UTF8, 43 BUFFER_NUM_TYPES, 44} buffer_type_t; 45 46static const char *buffer_names[] = { 47 "empty", 48 "one-by-one", 49 "utf32", 50 "utf16", 51 "utf8" 52}; 53 54typedef struct 55{ 56 hb_buffer_t *buffer; 57} fixture_t; 58 59static void 60fixture_init (fixture_t *fixture, gconstpointer user_data) 61{ 62 hb_buffer_t *b; 63 unsigned int i; 64 65 b = fixture->buffer = hb_buffer_create (); 66 67 switch (GPOINTER_TO_INT (user_data)) 68 { 69 case BUFFER_EMPTY: 70 break; 71 72 case BUFFER_ONE_BY_ONE: 73 for (i = 1; i < G_N_ELEMENTS (utf32) - 1; i++) 74 hb_buffer_add (b, utf32[i], 1, i); 75 break; 76 77 case BUFFER_UTF32: 78 hb_buffer_add_utf32 (b, utf32, G_N_ELEMENTS (utf32), 1, G_N_ELEMENTS (utf32) - 2); 79 break; 80 81 case BUFFER_UTF16: 82 hb_buffer_add_utf16 (b, utf16, G_N_ELEMENTS (utf16), 1, G_N_ELEMENTS (utf16) - 2); 83 break; 84 85 case BUFFER_UTF8: 86 hb_buffer_add_utf8 (b, utf8, G_N_ELEMENTS (utf8), 1, G_N_ELEMENTS (utf8) - 2); 87 break; 88 89 default: 90 g_assert_not_reached (); 91 } 92} 93 94static void 95fixture_finish (fixture_t *fixture, gconstpointer user_data) 96{ 97 hb_buffer_destroy (fixture->buffer); 98} 99 100 101static void 102test_buffer_properties (fixture_t *fixture, gconstpointer user_data) 103{ 104 hb_buffer_t *b = fixture->buffer; 105 hb_unicode_funcs_t *ufuncs; 106 107 /* test default properties */ 108 109 g_assert (hb_buffer_get_unicode_funcs (b) == hb_unicode_funcs_get_default ()); 110 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_INVALID); 111 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_INVALID); 112 g_assert (hb_buffer_get_language (b) == NULL); 113 114 115 /* test property changes are retained */ 116 ufuncs = hb_unicode_funcs_create (NULL); 117 hb_buffer_set_unicode_funcs (b, ufuncs); 118 hb_unicode_funcs_destroy (ufuncs); 119 g_assert (hb_buffer_get_unicode_funcs (b) == ufuncs); 120 121 hb_buffer_set_direction (b, HB_DIRECTION_RTL); 122 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_RTL); 123 124 hb_buffer_set_script (b, HB_SCRIPT_ARABIC); 125 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_ARABIC); 126 127 hb_buffer_set_language (b, hb_language_from_string ("fa", -1)); 128 g_assert (hb_buffer_get_language (b) == hb_language_from_string ("Fa", -1)); 129 130 131 132 /* test clear clears all properties but unicode_funcs */ 133 134 hb_buffer_clear (b); 135 136 g_assert (hb_buffer_get_unicode_funcs (b) == ufuncs); 137 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_INVALID); 138 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_INVALID); 139 g_assert (hb_buffer_get_language (b) == NULL); 140 141 142 /* test reset clears all properties */ 143 144 hb_buffer_set_direction (b, HB_DIRECTION_RTL); 145 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_RTL); 146 147 hb_buffer_set_script (b, HB_SCRIPT_ARABIC); 148 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_ARABIC); 149 150 hb_buffer_set_language (b, hb_language_from_string ("fa", -1)); 151 g_assert (hb_buffer_get_language (b) == hb_language_from_string ("Fa", -1)); 152 153 hb_buffer_reset (b); 154 155 g_assert (hb_buffer_get_unicode_funcs (b) == hb_unicode_funcs_get_default ()); 156 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_INVALID); 157 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_INVALID); 158 g_assert (hb_buffer_get_language (b) == NULL); 159} 160 161static void 162test_buffer_contents (fixture_t *fixture, gconstpointer user_data) 163{ 164 hb_buffer_t *b = fixture->buffer; 165 unsigned int i, len, len2; 166 buffer_type_t buffer_type = GPOINTER_TO_INT (user_data); 167 hb_glyph_info_t *glyphs; 168 169 if (buffer_type == BUFFER_EMPTY) { 170 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 171 return; 172 } 173 174 len = hb_buffer_get_length (b); 175 hb_buffer_get_glyph_infos (b, NULL); /* test NULL */ 176 glyphs = hb_buffer_get_glyph_infos (b, &len2); 177 g_assert_cmpint (len, ==, len2); 178 g_assert_cmpint (len, ==, 5); 179 180 for (i = 0; i < len; i++) { 181 g_assert_cmphex (glyphs[i].mask, ==, 1); 182 g_assert_cmphex (glyphs[i].var1.u32, ==, 0); 183 g_assert_cmphex (glyphs[i].var2.u32, ==, 0); 184 } 185 186 for (i = 0; i < len; i++) { 187 unsigned int cluster; 188 cluster = 1+i; 189 if (i >= 2) { 190 if (buffer_type == BUFFER_UTF16) 191 cluster++; 192 else if (buffer_type == BUFFER_UTF8) 193 cluster += 3; 194 } 195 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 196 g_assert_cmphex (glyphs[i].cluster, ==, cluster); 197 } 198 199 /* reverse, test, and reverse back */ 200 201 hb_buffer_reverse (b); 202 for (i = 0; i < len; i++) 203 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]); 204 205 hb_buffer_reverse (b); 206 for (i = 0; i < len; i++) 207 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 208 209 /* reverse_clusters works same as reverse for now since each codepoint is 210 * in its own cluster */ 211 212 hb_buffer_reverse_clusters (b); 213 for (i = 0; i < len; i++) 214 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]); 215 216 hb_buffer_reverse_clusters (b); 217 for (i = 0; i < len; i++) 218 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 219 220 /* now form a cluster and test again */ 221 glyphs[2].cluster = glyphs[1].cluster; 222 223 /* reverse, test, and reverse back */ 224 225 hb_buffer_reverse (b); 226 for (i = 0; i < len; i++) 227 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]); 228 229 hb_buffer_reverse (b); 230 for (i = 0; i < len; i++) 231 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 232 233 /* reverse_clusters twice still should return the original string, 234 * but when applied once, the 1-2 cluster should be retained. */ 235 236 hb_buffer_reverse_clusters (b); 237 for (i = 0; i < len; i++) { 238 unsigned int j = len-1-i; 239 if (j == 1) 240 j = 2; 241 else if (j == 2) 242 j = 1; 243 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+j]); 244 } 245 246 hb_buffer_reverse_clusters (b); 247 for (i = 0; i < len; i++) 248 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 249 250 251 /* test setting length */ 252 253 /* enlarge */ 254 g_assert (hb_buffer_set_length (b, 10)); 255 glyphs = hb_buffer_get_glyph_infos (b, NULL); 256 g_assert_cmpint (hb_buffer_get_length (b), ==, 10); 257 for (i = 0; i < 5; i++) 258 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 259 for (i = 5; i < 10; i++) 260 g_assert_cmphex (glyphs[i].codepoint, ==, 0); 261 /* shrink */ 262 g_assert (hb_buffer_set_length (b, 3)); 263 glyphs = hb_buffer_get_glyph_infos (b, NULL); 264 g_assert_cmpint (hb_buffer_get_length (b), ==, 3); 265 for (i = 0; i < 3; i++) 266 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]); 267 268 269 g_assert (hb_buffer_allocation_successful (b)); 270 271 272 /* test reset clears content */ 273 274 hb_buffer_reset (b); 275 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 276} 277 278static void 279test_buffer_positions (fixture_t *fixture, gconstpointer user_data) 280{ 281 hb_buffer_t *b = fixture->buffer; 282 unsigned int i, len, len2; 283 hb_glyph_position_t *positions; 284 285 /* Without shaping, positions should all be zero */ 286 len = hb_buffer_get_length (b); 287 hb_buffer_get_glyph_positions (b, NULL); /* test NULL */ 288 positions = hb_buffer_get_glyph_positions (b, &len2); 289 g_assert_cmpint (len, ==, len2); 290 for (i = 0; i < len; i++) { 291 g_assert_cmpint (0, ==, positions[i].x_advance); 292 g_assert_cmpint (0, ==, positions[i].y_advance); 293 g_assert_cmpint (0, ==, positions[i].x_offset); 294 g_assert_cmpint (0, ==, positions[i].y_offset); 295 g_assert_cmpint (0, ==, positions[i].var.i32); 296 } 297 298 /* test reset clears content */ 299 hb_buffer_reset (b); 300 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 301} 302 303static void 304test_buffer_allocation (fixture_t *fixture, gconstpointer user_data) 305{ 306 hb_buffer_t *b = fixture->buffer; 307 308 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 309 310 g_assert (hb_buffer_pre_allocate (b, 100)); 311 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 312 g_assert (hb_buffer_allocation_successful (b)); 313 314 /* lets try a huge allocation, make sure it fails */ 315 g_assert (!hb_buffer_pre_allocate (b, (unsigned int) -1)); 316 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 317 g_assert (!hb_buffer_allocation_successful (b)); 318 319 /* small one again */ 320 g_assert (hb_buffer_pre_allocate (b, 50)); 321 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 322 g_assert (!hb_buffer_allocation_successful (b)); 323 324 hb_buffer_reset (b); 325 g_assert (hb_buffer_allocation_successful (b)); 326 327 /* all allocation and size */ 328 g_assert (!hb_buffer_pre_allocate (b, ((unsigned int) -1) / 20 + 1)); 329 g_assert (!hb_buffer_allocation_successful (b)); 330 331 hb_buffer_reset (b); 332 g_assert (hb_buffer_allocation_successful (b)); 333 334 /* technically, this one can actually pass on 64bit machines, but 335 * I'm doubtful that any malloc allows 4GB allocations at a time. 336 * But let's only enable it on a 32-bit machine. */ 337 if (sizeof (long) == 4) { 338 g_assert (!hb_buffer_pre_allocate (b, ((unsigned int) -1) / 20 - 1)); 339 g_assert (!hb_buffer_allocation_successful (b)); 340 } 341 342 hb_buffer_reset (b); 343 g_assert (hb_buffer_allocation_successful (b)); 344} 345 346 347typedef struct { 348 const char utf8[8]; 349 const uint32_t codepoints[8]; 350} utf8_conversion_test_t; 351 352/* note: we skip the first and last byte when adding to buffer */ 353static const utf8_conversion_test_t utf8_conversion_tests[] = { 354 {"a\303\207", {-1}}, 355 {"a\303\207b", {0xC7}}, 356 {"ab\303cd", {'b', -1, 'c'}}, 357 {"ab\303\302\301cd", {'b', -1, -1, -1, 'c'}} 358}; 359 360static void 361test_buffer_utf8_conversion (void) 362{ 363 hb_buffer_t *b; 364 hb_glyph_info_t *glyphs; 365 unsigned int bytes, chars, i, j, len; 366 367 b = hb_buffer_create (); 368 369 for (i = 0; i < G_N_ELEMENTS (utf8_conversion_tests); i++) 370 { 371 const utf8_conversion_test_t *test = &utf8_conversion_tests[i]; 372 char *escaped; 373 374 escaped = g_strescape (test->utf8, NULL); 375 g_test_message ("UTF-8 test #%d: %s", i, escaped); 376 g_free (escaped); 377 378 bytes = strlen (test->utf8); 379 for (chars = 0; test->codepoints[chars]; chars++) 380 ; 381 382 hb_buffer_reset (b); 383 hb_buffer_add_utf8 (b, test->utf8, bytes, 1, bytes - 2); 384 385 glyphs = hb_buffer_get_glyph_infos (b, &len); 386 g_assert_cmpint (len, ==, chars); 387 for (j = 0; j < chars; j++) 388 g_assert_cmphex (glyphs[j].codepoint, ==, test->codepoints[j]); 389 } 390 391 hb_buffer_destroy (b); 392} 393 394 395 396/* Following test table is adapted from glib/glib/tests/utf8-validate.c 397 * with relicensing permission from Matthias Clasen. */ 398 399typedef struct { 400 const char *utf8; 401 int max_len; 402 unsigned int offset; 403 gboolean valid; 404} utf8_validity_test_t; 405 406static const utf8_validity_test_t utf8_validity_tests[] = { 407 /* some tests to check max_len handling */ 408 /* length 1 */ 409 { "abcde", -1, 5, TRUE }, 410 { "abcde", 3, 3, TRUE }, 411 { "abcde", 5, 5, TRUE }, 412 /* length 2 */ 413 { "\xc2\xa9\xc2\xa9\xc2\xa9", -1, 6, TRUE }, 414 { "\xc2\xa9\xc2\xa9\xc2\xa9", 1, 0, FALSE }, 415 { "\xc2\xa9\xc2\xa9\xc2\xa9", 2, 2, TRUE }, 416 { "\xc2\xa9\xc2\xa9\xc2\xa9", 3, 2, FALSE }, 417 { "\xc2\xa9\xc2\xa9\xc2\xa9", 4, 4, TRUE }, 418 { "\xc2\xa9\xc2\xa9\xc2\xa9", 5, 4, FALSE }, 419 { "\xc2\xa9\xc2\xa9\xc2\xa9", 6, 6, TRUE }, 420 /* length 3 */ 421 { "\xe2\x89\xa0\xe2\x89\xa0", -1, 6, TRUE }, 422 { "\xe2\x89\xa0\xe2\x89\xa0", 1, 0, FALSE }, 423 { "\xe2\x89\xa0\xe2\x89\xa0", 2, 0, FALSE }, 424 { "\xe2\x89\xa0\xe2\x89\xa0", 3, 3, TRUE }, 425 { "\xe2\x89\xa0\xe2\x89\xa0", 4, 3, FALSE }, 426 { "\xe2\x89\xa0\xe2\x89\xa0", 5, 3, FALSE }, 427 { "\xe2\x89\xa0\xe2\x89\xa0", 6, 6, TRUE }, 428 429 /* examples from http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt */ 430 /* greek 'kosme' */ 431 { "\xce\xba\xe1\xbd\xb9\xcf\x83\xce\xbc\xce\xb5", -1, 11, TRUE }, 432 /* first sequence of each length */ 433 { "\x00", -1, 0, TRUE }, 434 { "\xc2\x80", -1, 2, TRUE }, 435 { "\xe0\xa0\x80", -1, 3, TRUE }, 436 { "\xf0\x90\x80\x80", -1, 4, TRUE }, 437 { "\xf8\x88\x80\x80\x80", -1, 0, FALSE }, 438 { "\xfc\x84\x80\x80\x80\x80", -1, 0, FALSE }, 439 /* last sequence of each length */ 440 { "\x7f", -1, 1, TRUE }, 441 { "\xdf\xbf", -1, 2, TRUE }, 442 { "\xef\xbf\xbf", -1, 0, TRUE }, 443 { "\xf7\xbf\xbf\xbf", -1, 0, TRUE }, 444 { "\xfb\xbf\xbf\xbf\xbf", -1, 0, FALSE }, 445 { "\xfd\xbf\xbf\xbf\xbf\xbf", -1, 0, FALSE }, 446 /* other boundary conditions */ 447 { "\xed\x9f\xbf", -1, 3, TRUE }, 448 { "\xee\x80\x80", -1, 3, TRUE }, 449 { "\xef\xbf\xbd", -1, 3, TRUE }, 450 { "\xf4\x8f\xbf\xbf", -1, 0, TRUE }, 451 /* malformed sequences */ 452 /* continuation bytes */ 453 { "\x80", -1, 0, FALSE }, 454 { "\xbf", -1, 0, FALSE }, 455 { "\x80\xbf", -1, 0, FALSE }, 456 { "\x80\xbf\x80", -1, 0, FALSE }, 457 { "\x80\xbf\x80\xbf", -1, 0, FALSE }, 458 { "\x80\xbf\x80\xbf\x80", -1, 0, FALSE }, 459 { "\x80\xbf\x80\xbf\x80\xbf", -1, 0, FALSE }, 460 { "\x80\xbf\x80\xbf\x80\xbf\x80", -1, 0, FALSE }, 461 462 /* all possible continuation byte */ 463 { "\x80", -1, 0, FALSE }, 464 { "\x81", -1, 0, FALSE }, 465 { "\x82", -1, 0, FALSE }, 466 { "\x83", -1, 0, FALSE }, 467 { "\x84", -1, 0, FALSE }, 468 { "\x85", -1, 0, FALSE }, 469 { "\x86", -1, 0, FALSE }, 470 { "\x87", -1, 0, FALSE }, 471 { "\x88", -1, 0, FALSE }, 472 { "\x89", -1, 0, FALSE }, 473 { "\x8a", -1, 0, FALSE }, 474 { "\x8b", -1, 0, FALSE }, 475 { "\x8c", -1, 0, FALSE }, 476 { "\x8d", -1, 0, FALSE }, 477 { "\x8e", -1, 0, FALSE }, 478 { "\x8f", -1, 0, FALSE }, 479 { "\x90", -1, 0, FALSE }, 480 { "\x91", -1, 0, FALSE }, 481 { "\x92", -1, 0, FALSE }, 482 { "\x93", -1, 0, FALSE }, 483 { "\x94", -1, 0, FALSE }, 484 { "\x95", -1, 0, FALSE }, 485 { "\x96", -1, 0, FALSE }, 486 { "\x97", -1, 0, FALSE }, 487 { "\x98", -1, 0, FALSE }, 488 { "\x99", -1, 0, FALSE }, 489 { "\x9a", -1, 0, FALSE }, 490 { "\x9b", -1, 0, FALSE }, 491 { "\x9c", -1, 0, FALSE }, 492 { "\x9d", -1, 0, FALSE }, 493 { "\x9e", -1, 0, FALSE }, 494 { "\x9f", -1, 0, FALSE }, 495 { "\xa0", -1, 0, FALSE }, 496 { "\xa1", -1, 0, FALSE }, 497 { "\xa2", -1, 0, FALSE }, 498 { "\xa3", -1, 0, FALSE }, 499 { "\xa4", -1, 0, FALSE }, 500 { "\xa5", -1, 0, FALSE }, 501 { "\xa6", -1, 0, FALSE }, 502 { "\xa7", -1, 0, FALSE }, 503 { "\xa8", -1, 0, FALSE }, 504 { "\xa9", -1, 0, FALSE }, 505 { "\xaa", -1, 0, FALSE }, 506 { "\xab", -1, 0, FALSE }, 507 { "\xac", -1, 0, FALSE }, 508 { "\xad", -1, 0, FALSE }, 509 { "\xae", -1, 0, FALSE }, 510 { "\xaf", -1, 0, FALSE }, 511 { "\xb0", -1, 0, FALSE }, 512 { "\xb1", -1, 0, FALSE }, 513 { "\xb2", -1, 0, FALSE }, 514 { "\xb3", -1, 0, FALSE }, 515 { "\xb4", -1, 0, FALSE }, 516 { "\xb5", -1, 0, FALSE }, 517 { "\xb6", -1, 0, FALSE }, 518 { "\xb7", -1, 0, FALSE }, 519 { "\xb8", -1, 0, FALSE }, 520 { "\xb9", -1, 0, FALSE }, 521 { "\xba", -1, 0, FALSE }, 522 { "\xbb", -1, 0, FALSE }, 523 { "\xbc", -1, 0, FALSE }, 524 { "\xbd", -1, 0, FALSE }, 525 { "\xbe", -1, 0, FALSE }, 526 { "\xbf", -1, 0, FALSE }, 527 /* lone start characters */ 528 { "\xc0\x20", -1, 0, FALSE }, 529 { "\xc1\x20", -1, 0, FALSE }, 530 { "\xc2\x20", -1, 0, FALSE }, 531 { "\xc3\x20", -1, 0, FALSE }, 532 { "\xc4\x20", -1, 0, FALSE }, 533 { "\xc5\x20", -1, 0, FALSE }, 534 { "\xc6\x20", -1, 0, FALSE }, 535 { "\xc7\x20", -1, 0, FALSE }, 536 { "\xc8\x20", -1, 0, FALSE }, 537 { "\xc9\x20", -1, 0, FALSE }, 538 { "\xca\x20", -1, 0, FALSE }, 539 { "\xcb\x20", -1, 0, FALSE }, 540 { "\xcc\x20", -1, 0, FALSE }, 541 { "\xcd\x20", -1, 0, FALSE }, 542 { "\xce\x20", -1, 0, FALSE }, 543 { "\xcf\x20", -1, 0, FALSE }, 544 { "\xd0\x20", -1, 0, FALSE }, 545 { "\xd1\x20", -1, 0, FALSE }, 546 { "\xd2\x20", -1, 0, FALSE }, 547 { "\xd3\x20", -1, 0, FALSE }, 548 { "\xd4\x20", -1, 0, FALSE }, 549 { "\xd5\x20", -1, 0, FALSE }, 550 { "\xd6\x20", -1, 0, FALSE }, 551 { "\xd7\x20", -1, 0, FALSE }, 552 { "\xd8\x20", -1, 0, FALSE }, 553 { "\xd9\x20", -1, 0, FALSE }, 554 { "\xda\x20", -1, 0, FALSE }, 555 { "\xdb\x20", -1, 0, FALSE }, 556 { "\xdc\x20", -1, 0, FALSE }, 557 { "\xdd\x20", -1, 0, FALSE }, 558 { "\xde\x20", -1, 0, FALSE }, 559 { "\xdf\x20", -1, 0, FALSE }, 560 { "\xe0\x20", -1, 0, FALSE }, 561 { "\xe1\x20", -1, 0, FALSE }, 562 { "\xe2\x20", -1, 0, FALSE }, 563 { "\xe3\x20", -1, 0, FALSE }, 564 { "\xe4\x20", -1, 0, FALSE }, 565 { "\xe5\x20", -1, 0, FALSE }, 566 { "\xe6\x20", -1, 0, FALSE }, 567 { "\xe7\x20", -1, 0, FALSE }, 568 { "\xe8\x20", -1, 0, FALSE }, 569 { "\xe9\x20", -1, 0, FALSE }, 570 { "\xea\x20", -1, 0, FALSE }, 571 { "\xeb\x20", -1, 0, FALSE }, 572 { "\xec\x20", -1, 0, FALSE }, 573 { "\xed\x20", -1, 0, FALSE }, 574 { "\xee\x20", -1, 0, FALSE }, 575 { "\xef\x20", -1, 0, FALSE }, 576 { "\xf0\x20", -1, 0, FALSE }, 577 { "\xf1\x20", -1, 0, FALSE }, 578 { "\xf2\x20", -1, 0, FALSE }, 579 { "\xf3\x20", -1, 0, FALSE }, 580 { "\xf4\x20", -1, 0, FALSE }, 581 { "\xf5\x20", -1, 0, FALSE }, 582 { "\xf6\x20", -1, 0, FALSE }, 583 { "\xf7\x20", -1, 0, FALSE }, 584 { "\xf8\x20", -1, 0, FALSE }, 585 { "\xf9\x20", -1, 0, FALSE }, 586 { "\xfa\x20", -1, 0, FALSE }, 587 { "\xfb\x20", -1, 0, FALSE }, 588 { "\xfc\x20", -1, 0, FALSE }, 589 { "\xfd\x20", -1, 0, FALSE }, 590 /* missing continuation bytes */ 591 { "\x20\xc0", -1, 1, FALSE }, 592 { "\x20\xe0\x80", -1, 1, FALSE }, 593 { "\x20\xf0\x80\x80", -1, 1, FALSE }, 594 { "\x20\xf8\x80\x80\x80", -1, 1, FALSE }, 595 { "\x20\xfc\x80\x80\x80\x80", -1, 1, FALSE }, 596 { "\x20\xdf", -1, 1, FALSE }, 597 { "\x20\xef\xbf", -1, 1, FALSE }, 598 { "\x20\xf7\xbf\xbf", -1, 1, FALSE }, 599 { "\x20\xfb\xbf\xbf\xbf", -1, 1, FALSE }, 600 { "\x20\xfd\xbf\xbf\xbf\xbf", -1, 1, FALSE }, 601 /* impossible bytes */ 602 { "\x20\xfe\x20", -1, 1, FALSE }, 603 { "\x20\xff\x20", -1, 1, FALSE }, 604#if 0 605 /* XXX fix these, or document that we don't detect them? */ 606 /* overlong sequences */ 607 { "\x20\xc0\xaf\x20", -1, 1, FALSE }, 608 { "\x20\xe0\x80\xaf\x20", -1, 1, FALSE }, 609 { "\x20\xf0\x80\x80\xaf\x20", -1, 1, FALSE }, 610 { "\x20\xf8\x80\x80\x80\xaf\x20", -1, 1, FALSE }, 611 { "\x20\xfc\x80\x80\x80\x80\xaf\x20", -1, 1, FALSE }, 612 { "\x20\xc1\xbf\x20", -1, 1, FALSE }, 613 { "\x20\xe0\x9f\xbf\x20", -1, 1, FALSE }, 614 { "\x20\xf0\x8f\xbf\xbf\x20", -1, 1, FALSE }, 615 { "\x20\xf8\x87\xbf\xbf\xbf\x20", -1, 1, FALSE }, 616 { "\x20\xfc\x83\xbf\xbf\xbf\xbf\x20", -1, 1, FALSE }, 617 { "\x20\xc0\x80\x20", -1, 1, FALSE }, 618 { "\x20\xe0\x80\x80\x20", -1, 1, FALSE }, 619 { "\x20\xf0\x80\x80\x80\x20", -1, 1, FALSE }, 620 { "\x20\xf8\x80\x80\x80\x80\x20", -1, 1, FALSE }, 621 { "\x20\xfc\x80\x80\x80\x80\x80\x20", -1, 1, FALSE }, 622 /* illegal code positions */ 623 { "\x20\xed\xa0\x80\x20", -1, 1, FALSE }, 624 { "\x20\xed\xad\xbf\x20", -1, 1, FALSE }, 625 { "\x20\xed\xae\x80\x20", -1, 1, FALSE }, 626 { "\x20\xed\xaf\xbf\x20", -1, 1, FALSE }, 627 { "\x20\xed\xb0\x80\x20", -1, 1, FALSE }, 628 { "\x20\xed\xbe\x80\x20", -1, 1, FALSE }, 629 { "\x20\xed\xbf\xbf\x20", -1, 1, FALSE }, 630 { "\x20\xed\xa0\x80\xed\xb0\x80\x20", -1, 1, FALSE }, 631 { "\x20\xed\xa0\x80\xed\xbf\xbf\x20", -1, 1, FALSE }, 632 { "\x20\xed\xad\xbf\xed\xb0\x80\x20", -1, 1, FALSE }, 633 { "\x20\xed\xad\xbf\xed\xbf\xbf\x20", -1, 1, FALSE }, 634 { "\x20\xed\xae\x80\xed\xb0\x80\x20", -1, 1, FALSE }, 635 { "\x20\xed\xae\x80\xed\xbf\xbf\x20", -1, 1, FALSE }, 636 { "\x20\xed\xaf\xbf\xed\xb0\x80\x20", -1, 1, FALSE }, 637 { "\x20\xed\xaf\xbf\xed\xbf\xbf\x20", -1, 1, FALSE }, 638 { "\x20\xef\xbf\xbe\x20", -1, 1, FALSE }, 639 { "\x20\xef\xbf\xbf\x20", -1, 1, FALSE }, 640#endif 641 { "", -1, 0, TRUE } 642}; 643 644static void 645test_buffer_utf8_validity (void) 646{ 647 hb_buffer_t *b; 648 unsigned int i; 649 650 b = hb_buffer_create (); 651 652 for (i = 0; i < G_N_ELEMENTS (utf8_validity_tests); i++) 653 { 654 const utf8_validity_test_t *test = &utf8_validity_tests[i]; 655 unsigned int text_bytes, segment_bytes, j, len; 656 hb_glyph_info_t *glyphs; 657 char *escaped; 658 659 escaped = g_strescape (test->utf8, NULL); 660 g_test_message ("UTF-8 test #%d: %s", i, escaped); 661 g_free (escaped); 662 663 text_bytes = strlen (test->utf8); 664 if (test->max_len == -1) 665 segment_bytes = text_bytes; 666 else 667 segment_bytes = test->max_len; 668 669 hb_buffer_reset (b); 670 hb_buffer_add_utf8 (b, test->utf8, text_bytes, 0, segment_bytes); 671 672 glyphs = hb_buffer_get_glyph_infos (b, &len); 673 for (j = 0; j < len; j++) 674 if (glyphs[j].codepoint == (hb_codepoint_t) -1) 675 break; 676 677 g_assert (test->valid ? j == len : j < len); 678 if (!test->valid) 679 g_assert (glyphs[j].cluster == test->offset); 680 } 681 682 hb_buffer_destroy (b); 683} 684 685 686typedef struct { 687 const uint16_t utf16[8]; 688 const uint32_t codepoints[8]; 689} utf16_conversion_test_t; 690 691/* note: we skip the first and last item from utf16 when adding to buffer */ 692static const utf16_conversion_test_t utf16_conversion_tests[] = { 693 {{0x41, 0x004D, 0x0430, 0x4E8C, 0xD800, 0xDF02, 0x61} , {0x004D, 0x0430, 0x4E8C, 0x10302}}, 694 {{0x41, 0xD800, 0xDF02, 0x61}, {0x10302}}, 695 {{0x41, 0xD800, 0xDF02}, {-1}}, 696 {{0x41, 0x61, 0xD800, 0xDF02}, {0x61, -1}}, 697 {{0x41, 0xD800, 0x61, 0xDF02}, {-1, 0x61}}, 698 {{0x41, 0x61}, {}} 699}; 700 701static void 702test_buffer_utf16_conversion (void) 703{ 704 hb_buffer_t *b; 705 unsigned int i; 706 707 b = hb_buffer_create (); 708 709 for (i = 0; i < G_N_ELEMENTS (utf16_conversion_tests); i++) 710 { 711 const utf16_conversion_test_t *test = &utf16_conversion_tests[i]; 712 unsigned int u_len, chars, j, len; 713 hb_glyph_info_t *glyphs; 714 715 g_test_message ("UTF-16 test #%d", i); 716 717 for (u_len = 0; test->utf16[u_len]; u_len++) 718 ; 719 for (chars = 0; test->codepoints[chars]; chars++) 720 ; 721 722 hb_buffer_reset (b); 723 hb_buffer_add_utf16 (b, test->utf16, u_len, 1, u_len - 2); 724 725 glyphs = hb_buffer_get_glyph_infos (b, &len); 726 g_assert_cmpint (len, ==, chars); 727 for (j = 0; j < chars; j++) 728 g_assert_cmphex (glyphs[j].codepoint, ==, test->codepoints[j]); 729 } 730 731 hb_buffer_destroy (b); 732} 733 734static void 735test_empty (hb_buffer_t *b) 736{ 737 g_assert_cmpint (hb_buffer_get_length (b), ==, 0); 738 g_assert (!hb_buffer_get_glyph_infos (b, NULL)); 739 g_assert (!hb_buffer_get_glyph_positions (b, NULL)); 740} 741 742static void 743test_buffer_empty (void) 744{ 745 hb_buffer_t *b = hb_buffer_get_empty (); 746 747 g_assert (hb_buffer_get_empty ()); 748 g_assert (hb_buffer_get_empty () == b); 749 750 g_assert (!hb_buffer_allocation_successful (b)); 751 752 test_empty (b); 753 754 hb_buffer_add_utf32 (b, utf32, G_N_ELEMENTS (utf32), 1, G_N_ELEMENTS (utf32) - 2); 755 756 test_empty (b); 757 758 hb_buffer_reverse (b); 759 hb_buffer_reverse_clusters (b); 760 761 g_assert (!hb_buffer_set_length (b, 10)); 762 763 test_empty (b); 764 765 g_assert (hb_buffer_set_length (b, 0)); 766 767 test_empty (b); 768 769 g_assert (!hb_buffer_allocation_successful (b)); 770 771 hb_buffer_reset (b); 772 773 test_empty (b); 774 775 g_assert (!hb_buffer_allocation_successful (b)); 776} 777 778int 779main (int argc, char **argv) 780{ 781 unsigned int i; 782 783 hb_test_init (&argc, &argv); 784 785 for (i = 0; i < BUFFER_NUM_TYPES; i++) 786 { 787 const void *buffer_type = GINT_TO_POINTER (i); 788 const char *buffer_name = buffer_names[i]; 789 790 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_properties); 791 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_contents); 792 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_positions); 793 } 794 795 hb_test_add_fixture (fixture, GINT_TO_POINTER (BUFFER_EMPTY), test_buffer_allocation); 796 797 hb_test_add (test_buffer_utf8_conversion); 798 hb_test_add (test_buffer_utf8_validity); 799 hb_test_add (test_buffer_utf16_conversion); 800 hb_test_add (test_buffer_empty); 801 802 return hb_test_run(); 803} 804