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
32 static const char utf8[10] = "ab\360\240\200\200defg";
33 static const uint16_t utf16[8] = {'a', 'b', 0xD840, 0xDC00, 'd', 'e', 'f', 'g'};
34 static const uint32_t utf32[7] = {'a', 'b', 0x20000, 'd', 'e', 'f', 'g'};
35
36
37 typedef 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
46 static const char *buffer_names[] = {
47 "empty",
48 "one-by-one",
49 "utf32",
50 "utf16",
51 "utf8"
52 };
53
54 typedef struct
55 {
56 hb_buffer_t *buffer;
57 } fixture_t;
58
59 static void
fixture_init(fixture_t * fixture,gconstpointer user_data)60 fixture_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], 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
94 static void
fixture_finish(fixture_t * fixture,gconstpointer user_data)95 fixture_finish (fixture_t *fixture, gconstpointer user_data)
96 {
97 hb_buffer_destroy (fixture->buffer);
98 }
99
100
101 static void
test_buffer_properties(fixture_t * fixture,gconstpointer user_data)102 test_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 g_assert (hb_buffer_get_flags (b) == HB_BUFFER_FLAG_DEFAULT);
114
115
116 /* test property changes are retained */
117 ufuncs = hb_unicode_funcs_create (NULL);
118 hb_buffer_set_unicode_funcs (b, ufuncs);
119 hb_unicode_funcs_destroy (ufuncs);
120 g_assert (hb_buffer_get_unicode_funcs (b) == ufuncs);
121
122 hb_buffer_set_direction (b, HB_DIRECTION_RTL);
123 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_RTL);
124
125 hb_buffer_set_script (b, HB_SCRIPT_ARABIC);
126 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_ARABIC);
127
128 hb_buffer_set_language (b, hb_language_from_string ("fa", -1));
129 g_assert (hb_buffer_get_language (b) == hb_language_from_string ("Fa", -1));
130
131 hb_buffer_set_flags (b, HB_BUFFER_FLAG_BOT);
132 g_assert (hb_buffer_get_flags (b) == HB_BUFFER_FLAG_BOT);
133
134
135
136 /* test clear clears all properties but unicode_funcs */
137
138 hb_buffer_clear_contents (b);
139
140 g_assert (hb_buffer_get_unicode_funcs (b) == ufuncs);
141 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_INVALID);
142 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_INVALID);
143 g_assert (hb_buffer_get_language (b) == NULL);
144 g_assert (hb_buffer_get_flags (b) == HB_BUFFER_FLAGS_DEFAULT);
145
146
147 /* test reset clears all properties */
148
149 hb_buffer_set_direction (b, HB_DIRECTION_RTL);
150 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_RTL);
151
152 hb_buffer_set_script (b, HB_SCRIPT_ARABIC);
153 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_ARABIC);
154
155 hb_buffer_set_language (b, hb_language_from_string ("fa", -1));
156 g_assert (hb_buffer_get_language (b) == hb_language_from_string ("Fa", -1));
157
158 hb_buffer_set_flags (b, HB_BUFFER_FLAG_BOT);
159 g_assert (hb_buffer_get_flags (b) == HB_BUFFER_FLAG_BOT);
160
161 hb_buffer_reset (b);
162
163 g_assert (hb_buffer_get_unicode_funcs (b) == hb_unicode_funcs_get_default ());
164 g_assert (hb_buffer_get_direction (b) == HB_DIRECTION_INVALID);
165 g_assert (hb_buffer_get_script (b) == HB_SCRIPT_INVALID);
166 g_assert (hb_buffer_get_language (b) == NULL);
167 g_assert (hb_buffer_get_flags (b) == HB_BUFFER_FLAGS_DEFAULT);
168 }
169
170 static void
test_buffer_contents(fixture_t * fixture,gconstpointer user_data)171 test_buffer_contents (fixture_t *fixture, gconstpointer user_data)
172 {
173 hb_buffer_t *b = fixture->buffer;
174 unsigned int i, len, len2;
175 buffer_type_t buffer_type = GPOINTER_TO_INT (user_data);
176 hb_glyph_info_t *glyphs;
177
178 if (buffer_type == BUFFER_EMPTY) {
179 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
180 return;
181 }
182
183 len = hb_buffer_get_length (b);
184 hb_buffer_get_glyph_infos (b, NULL); /* test NULL */
185 glyphs = hb_buffer_get_glyph_infos (b, &len2);
186 g_assert_cmpint (len, ==, len2);
187 g_assert_cmpint (len, ==, 5);
188
189 for (i = 0; i < len; i++) {
190 g_assert_cmphex (glyphs[i].mask, ==, 1);
191 g_assert_cmphex (glyphs[i].var1.u32, ==, 0);
192 g_assert_cmphex (glyphs[i].var2.u32, ==, 0);
193 }
194
195 for (i = 0; i < len; i++) {
196 unsigned int cluster;
197 cluster = 1+i;
198 if (i >= 2) {
199 if (buffer_type == BUFFER_UTF16)
200 cluster++;
201 else if (buffer_type == BUFFER_UTF8)
202 cluster += 3;
203 }
204 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
205 g_assert_cmphex (glyphs[i].cluster, ==, cluster);
206 }
207
208 /* reverse, test, and reverse back */
209
210 hb_buffer_reverse (b);
211 for (i = 0; i < len; i++)
212 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]);
213
214 hb_buffer_reverse (b);
215 for (i = 0; i < len; i++)
216 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
217
218 /* reverse_clusters works same as reverse for now since each codepoint is
219 * in its own cluster */
220
221 hb_buffer_reverse_clusters (b);
222 for (i = 0; i < len; i++)
223 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]);
224
225 hb_buffer_reverse_clusters (b);
226 for (i = 0; i < len; i++)
227 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
228
229 /* now form a cluster and test again */
230 glyphs[2].cluster = glyphs[1].cluster;
231
232 /* reverse, test, and reverse back */
233
234 hb_buffer_reverse (b);
235 for (i = 0; i < len; i++)
236 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[len-i]);
237
238 hb_buffer_reverse (b);
239 for (i = 0; i < len; i++)
240 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
241
242 /* reverse_clusters twice still should return the original string,
243 * but when applied once, the 1-2 cluster should be retained. */
244
245 hb_buffer_reverse_clusters (b);
246 for (i = 0; i < len; i++) {
247 unsigned int j = len-1-i;
248 if (j == 1)
249 j = 2;
250 else if (j == 2)
251 j = 1;
252 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+j]);
253 }
254
255 hb_buffer_reverse_clusters (b);
256 for (i = 0; i < len; i++)
257 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
258
259
260 /* test setting length */
261
262 /* enlarge */
263 g_assert (hb_buffer_set_length (b, 10));
264 glyphs = hb_buffer_get_glyph_infos (b, NULL);
265 g_assert_cmpint (hb_buffer_get_length (b), ==, 10);
266 for (i = 0; i < 5; i++)
267 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
268 for (i = 5; i < 10; i++)
269 g_assert_cmphex (glyphs[i].codepoint, ==, 0);
270 /* shrink */
271 g_assert (hb_buffer_set_length (b, 3));
272 glyphs = hb_buffer_get_glyph_infos (b, NULL);
273 g_assert_cmpint (hb_buffer_get_length (b), ==, 3);
274 for (i = 0; i < 3; i++)
275 g_assert_cmphex (glyphs[i].codepoint, ==, utf32[1+i]);
276
277
278 g_assert (hb_buffer_allocation_successful (b));
279
280
281 /* test reset clears content */
282
283 hb_buffer_reset (b);
284 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
285 }
286
287 static void
test_buffer_positions(fixture_t * fixture,gconstpointer user_data)288 test_buffer_positions (fixture_t *fixture, gconstpointer user_data)
289 {
290 hb_buffer_t *b = fixture->buffer;
291 unsigned int i, len, len2;
292 hb_glyph_position_t *positions;
293
294 /* Without shaping, positions should all be zero */
295 len = hb_buffer_get_length (b);
296 hb_buffer_get_glyph_positions (b, NULL); /* test NULL */
297 positions = hb_buffer_get_glyph_positions (b, &len2);
298 g_assert_cmpint (len, ==, len2);
299 for (i = 0; i < len; i++) {
300 g_assert_cmpint (0, ==, positions[i].x_advance);
301 g_assert_cmpint (0, ==, positions[i].y_advance);
302 g_assert_cmpint (0, ==, positions[i].x_offset);
303 g_assert_cmpint (0, ==, positions[i].y_offset);
304 g_assert_cmpint (0, ==, positions[i].var.i32);
305 }
306
307 /* test reset clears content */
308 hb_buffer_reset (b);
309 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
310 }
311
312 static void
test_buffer_allocation(fixture_t * fixture,gconstpointer user_data)313 test_buffer_allocation (fixture_t *fixture, gconstpointer user_data)
314 {
315 hb_buffer_t *b = fixture->buffer;
316
317 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
318
319 g_assert (hb_buffer_pre_allocate (b, 100));
320 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
321 g_assert (hb_buffer_allocation_successful (b));
322
323 /* lets try a huge allocation, make sure it fails */
324 g_assert (!hb_buffer_pre_allocate (b, (unsigned int) -1));
325 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
326 g_assert (!hb_buffer_allocation_successful (b));
327
328 /* small one again */
329 g_assert (hb_buffer_pre_allocate (b, 50));
330 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
331 g_assert (!hb_buffer_allocation_successful (b));
332
333 hb_buffer_reset (b);
334 g_assert (hb_buffer_allocation_successful (b));
335
336 /* all allocation and size */
337 g_assert (!hb_buffer_pre_allocate (b, ((unsigned int) -1) / 20 + 1));
338 g_assert (!hb_buffer_allocation_successful (b));
339
340 hb_buffer_reset (b);
341 g_assert (hb_buffer_allocation_successful (b));
342
343 /* technically, this one can actually pass on 64bit machines, but
344 * I'm doubtful that any malloc allows 4GB allocations at a time.
345 * But let's only enable it on a 32-bit machine. */
346 if (sizeof (long) == 4) {
347 g_assert (!hb_buffer_pre_allocate (b, ((unsigned int) -1) / 20 - 1));
348 g_assert (!hb_buffer_allocation_successful (b));
349 }
350
351 hb_buffer_reset (b);
352 g_assert (hb_buffer_allocation_successful (b));
353 }
354
355
356 typedef struct {
357 const char utf8[8];
358 const uint32_t codepoints[8];
359 } utf8_conversion_test_t;
360
361 /* note: we skip the first and last byte when adding to buffer */
362 static const utf8_conversion_test_t utf8_conversion_tests[] = {
363 {"a\303\207", {-1}},
364 {"a\303\207b", {0xC7}},
365 {"ab\303cd", {'b', -1, 'c'}},
366 {"ab\303\302\301cd", {'b', -1, -1, -1, 'c'}}
367 };
368
369 static void
test_buffer_utf8_conversion(void)370 test_buffer_utf8_conversion (void)
371 {
372 hb_buffer_t *b;
373 hb_glyph_info_t *glyphs;
374 unsigned int bytes, chars, i, j, len;
375
376 b = hb_buffer_create ();
377 hb_buffer_set_replacement_codepoint (b, (hb_codepoint_t) -1);
378
379 for (i = 0; i < G_N_ELEMENTS (utf8_conversion_tests); i++)
380 {
381 const utf8_conversion_test_t *test = &utf8_conversion_tests[i];
382 char *escaped;
383
384 escaped = g_strescape (test->utf8, NULL);
385 g_test_message ("UTF-8 test #%d: %s", i, escaped);
386 g_free (escaped);
387
388 bytes = strlen (test->utf8);
389 for (chars = 0; test->codepoints[chars]; chars++)
390 ;
391
392 hb_buffer_clear_contents (b);
393 hb_buffer_add_utf8 (b, test->utf8, bytes, 1, bytes - 2);
394
395 glyphs = hb_buffer_get_glyph_infos (b, &len);
396 g_assert_cmpint (len, ==, chars);
397 for (j = 0; j < chars; j++)
398 g_assert_cmphex (glyphs[j].codepoint, ==, test->codepoints[j]);
399 }
400
401 hb_buffer_destroy (b);
402 }
403
404
405
406 /* Following test table is adapted from glib/glib/tests/utf8-validate.c
407 * with relicensing permission from Matthias Clasen. */
408
409 typedef struct {
410 const char *utf8;
411 int max_len;
412 unsigned int offset;
413 gboolean valid;
414 } utf8_validity_test_t;
415
416 static const utf8_validity_test_t utf8_validity_tests[] = {
417 /* some tests to check max_len handling */
418 /* length 1 */
419 { "abcde", -1, 5, TRUE },
420 { "abcde", 3, 3, TRUE },
421 { "abcde", 5, 5, TRUE },
422 /* length 2 */
423 { "\xc2\xa9\xc2\xa9\xc2\xa9", -1, 6, TRUE },
424 { "\xc2\xa9\xc2\xa9\xc2\xa9", 1, 0, FALSE },
425 { "\xc2\xa9\xc2\xa9\xc2\xa9", 2, 2, TRUE },
426 { "\xc2\xa9\xc2\xa9\xc2\xa9", 3, 2, FALSE },
427 { "\xc2\xa9\xc2\xa9\xc2\xa9", 4, 4, TRUE },
428 { "\xc2\xa9\xc2\xa9\xc2\xa9", 5, 4, FALSE },
429 { "\xc2\xa9\xc2\xa9\xc2\xa9", 6, 6, TRUE },
430 /* length 3 */
431 { "\xe2\x89\xa0\xe2\x89\xa0", -1, 6, TRUE },
432 { "\xe2\x89\xa0\xe2\x89\xa0", 1, 0, FALSE },
433 { "\xe2\x89\xa0\xe2\x89\xa0", 2, 0, FALSE },
434 { "\xe2\x89\xa0\xe2\x89\xa0", 3, 3, TRUE },
435 { "\xe2\x89\xa0\xe2\x89\xa0", 4, 3, FALSE },
436 { "\xe2\x89\xa0\xe2\x89\xa0", 5, 3, FALSE },
437 { "\xe2\x89\xa0\xe2\x89\xa0", 6, 6, TRUE },
438
439 /* examples from http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt */
440 /* greek 'kosme' */
441 { "\xce\xba\xe1\xbd\xb9\xcf\x83\xce\xbc\xce\xb5", -1, 11, TRUE },
442 /* first sequence of each length */
443 { "\x00", -1, 0, TRUE },
444 { "\xc2\x80", -1, 2, TRUE },
445 { "\xe0\xa0\x80", -1, 3, TRUE },
446 { "\xf0\x90\x80\x80", -1, 4, TRUE },
447 { "\xf8\x88\x80\x80\x80", -1, 0, FALSE },
448 { "\xfc\x84\x80\x80\x80\x80", -1, 0, FALSE },
449 /* last sequence of each length */
450 { "\x7f", -1, 1, TRUE },
451 { "\xdf\xbf", -1, 2, TRUE },
452 { "\xef\xbf\xbf", -1, 0, TRUE },
453 { "\xf4\x8f\xbf\xbf", -1, 0, TRUE },
454 { "\xf4\x90\xbf\xbf", -1, 0, FALSE },
455 { "\xf7\xbf\xbf\xbf", -1, 0, FALSE },
456 { "\xfb\xbf\xbf\xbf\xbf", -1, 0, FALSE },
457 { "\xfd\xbf\xbf\xbf\xbf\xbf", -1, 0, FALSE },
458 /* other boundary conditions */
459 { "\xed\x9f\xbf", -1, 3, TRUE },
460 { "\xed\xa0\x80", -1, 0, FALSE },
461 { "\xed\xbf\xbf", -1, 0, FALSE },
462 { "\xee\x80\x80", -1, 3, TRUE },
463 { "\xef\xbf\xbd", -1, 3, TRUE },
464 { "\xf4\x8f\xbf\xbf", -1, 0, TRUE },
465 /* malformed sequences */
466 /* continuation bytes */
467 { "\x80", -1, 0, FALSE },
468 { "\xbf", -1, 0, FALSE },
469 { "\x80\xbf", -1, 0, FALSE },
470 { "\x80\xbf\x80", -1, 0, FALSE },
471 { "\x80\xbf\x80\xbf", -1, 0, FALSE },
472 { "\x80\xbf\x80\xbf\x80", -1, 0, FALSE },
473 { "\x80\xbf\x80\xbf\x80\xbf", -1, 0, FALSE },
474 { "\x80\xbf\x80\xbf\x80\xbf\x80", -1, 0, FALSE },
475
476 /* all possible continuation byte */
477 { "\x80", -1, 0, FALSE },
478 { "\x81", -1, 0, FALSE },
479 { "\x82", -1, 0, FALSE },
480 { "\x83", -1, 0, FALSE },
481 { "\x84", -1, 0, FALSE },
482 { "\x85", -1, 0, FALSE },
483 { "\x86", -1, 0, FALSE },
484 { "\x87", -1, 0, FALSE },
485 { "\x88", -1, 0, FALSE },
486 { "\x89", -1, 0, FALSE },
487 { "\x8a", -1, 0, FALSE },
488 { "\x8b", -1, 0, FALSE },
489 { "\x8c", -1, 0, FALSE },
490 { "\x8d", -1, 0, FALSE },
491 { "\x8e", -1, 0, FALSE },
492 { "\x8f", -1, 0, FALSE },
493 { "\x90", -1, 0, FALSE },
494 { "\x91", -1, 0, FALSE },
495 { "\x92", -1, 0, FALSE },
496 { "\x93", -1, 0, FALSE },
497 { "\x94", -1, 0, FALSE },
498 { "\x95", -1, 0, FALSE },
499 { "\x96", -1, 0, FALSE },
500 { "\x97", -1, 0, FALSE },
501 { "\x98", -1, 0, FALSE },
502 { "\x99", -1, 0, FALSE },
503 { "\x9a", -1, 0, FALSE },
504 { "\x9b", -1, 0, FALSE },
505 { "\x9c", -1, 0, FALSE },
506 { "\x9d", -1, 0, FALSE },
507 { "\x9e", -1, 0, FALSE },
508 { "\x9f", -1, 0, FALSE },
509 { "\xa0", -1, 0, FALSE },
510 { "\xa1", -1, 0, FALSE },
511 { "\xa2", -1, 0, FALSE },
512 { "\xa3", -1, 0, FALSE },
513 { "\xa4", -1, 0, FALSE },
514 { "\xa5", -1, 0, FALSE },
515 { "\xa6", -1, 0, FALSE },
516 { "\xa7", -1, 0, FALSE },
517 { "\xa8", -1, 0, FALSE },
518 { "\xa9", -1, 0, FALSE },
519 { "\xaa", -1, 0, FALSE },
520 { "\xab", -1, 0, FALSE },
521 { "\xac", -1, 0, FALSE },
522 { "\xad", -1, 0, FALSE },
523 { "\xae", -1, 0, FALSE },
524 { "\xaf", -1, 0, FALSE },
525 { "\xb0", -1, 0, FALSE },
526 { "\xb1", -1, 0, FALSE },
527 { "\xb2", -1, 0, FALSE },
528 { "\xb3", -1, 0, FALSE },
529 { "\xb4", -1, 0, FALSE },
530 { "\xb5", -1, 0, FALSE },
531 { "\xb6", -1, 0, FALSE },
532 { "\xb7", -1, 0, FALSE },
533 { "\xb8", -1, 0, FALSE },
534 { "\xb9", -1, 0, FALSE },
535 { "\xba", -1, 0, FALSE },
536 { "\xbb", -1, 0, FALSE },
537 { "\xbc", -1, 0, FALSE },
538 { "\xbd", -1, 0, FALSE },
539 { "\xbe", -1, 0, FALSE },
540 { "\xbf", -1, 0, FALSE },
541 /* lone start characters */
542 { "\xc0\x20", -1, 0, FALSE },
543 { "\xc1\x20", -1, 0, FALSE },
544 { "\xc2\x20", -1, 0, FALSE },
545 { "\xc3\x20", -1, 0, FALSE },
546 { "\xc4\x20", -1, 0, FALSE },
547 { "\xc5\x20", -1, 0, FALSE },
548 { "\xc6\x20", -1, 0, FALSE },
549 { "\xc7\x20", -1, 0, FALSE },
550 { "\xc8\x20", -1, 0, FALSE },
551 { "\xc9\x20", -1, 0, FALSE },
552 { "\xca\x20", -1, 0, FALSE },
553 { "\xcb\x20", -1, 0, FALSE },
554 { "\xcc\x20", -1, 0, FALSE },
555 { "\xcd\x20", -1, 0, FALSE },
556 { "\xce\x20", -1, 0, FALSE },
557 { "\xcf\x20", -1, 0, FALSE },
558 { "\xd0\x20", -1, 0, FALSE },
559 { "\xd1\x20", -1, 0, FALSE },
560 { "\xd2\x20", -1, 0, FALSE },
561 { "\xd3\x20", -1, 0, FALSE },
562 { "\xd4\x20", -1, 0, FALSE },
563 { "\xd5\x20", -1, 0, FALSE },
564 { "\xd6\x20", -1, 0, FALSE },
565 { "\xd7\x20", -1, 0, FALSE },
566 { "\xd8\x20", -1, 0, FALSE },
567 { "\xd9\x20", -1, 0, FALSE },
568 { "\xda\x20", -1, 0, FALSE },
569 { "\xdb\x20", -1, 0, FALSE },
570 { "\xdc\x20", -1, 0, FALSE },
571 { "\xdd\x20", -1, 0, FALSE },
572 { "\xde\x20", -1, 0, FALSE },
573 { "\xdf\x20", -1, 0, FALSE },
574 { "\xe0\x20", -1, 0, FALSE },
575 { "\xe1\x20", -1, 0, FALSE },
576 { "\xe2\x20", -1, 0, FALSE },
577 { "\xe3\x20", -1, 0, FALSE },
578 { "\xe4\x20", -1, 0, FALSE },
579 { "\xe5\x20", -1, 0, FALSE },
580 { "\xe6\x20", -1, 0, FALSE },
581 { "\xe7\x20", -1, 0, FALSE },
582 { "\xe8\x20", -1, 0, FALSE },
583 { "\xe9\x20", -1, 0, FALSE },
584 { "\xea\x20", -1, 0, FALSE },
585 { "\xeb\x20", -1, 0, FALSE },
586 { "\xec\x20", -1, 0, FALSE },
587 { "\xed\x20", -1, 0, FALSE },
588 { "\xee\x20", -1, 0, FALSE },
589 { "\xef\x20", -1, 0, FALSE },
590 { "\xf0\x20", -1, 0, FALSE },
591 { "\xf1\x20", -1, 0, FALSE },
592 { "\xf2\x20", -1, 0, FALSE },
593 { "\xf3\x20", -1, 0, FALSE },
594 { "\xf4\x20", -1, 0, FALSE },
595 { "\xf5\x20", -1, 0, FALSE },
596 { "\xf6\x20", -1, 0, FALSE },
597 { "\xf7\x20", -1, 0, FALSE },
598 { "\xf8\x20", -1, 0, FALSE },
599 { "\xf9\x20", -1, 0, FALSE },
600 { "\xfa\x20", -1, 0, FALSE },
601 { "\xfb\x20", -1, 0, FALSE },
602 { "\xfc\x20", -1, 0, FALSE },
603 { "\xfd\x20", -1, 0, FALSE },
604 /* missing continuation bytes */
605 { "\x20\xc0", -1, 1, FALSE },
606 { "\x20\xe0\x80", -1, 1, FALSE },
607 { "\x20\xf0\x80\x80", -1, 1, FALSE },
608 { "\x20\xf8\x80\x80\x80", -1, 1, FALSE },
609 { "\x20\xfc\x80\x80\x80\x80", -1, 1, FALSE },
610 { "\x20\xdf", -1, 1, FALSE },
611 { "\x20\xef\xbf", -1, 1, FALSE },
612 { "\x20\xf7\xbf\xbf", -1, 1, FALSE },
613 { "\x20\xfb\xbf\xbf\xbf", -1, 1, FALSE },
614 { "\x20\xfd\xbf\xbf\xbf\xbf", -1, 1, FALSE },
615 /* impossible bytes */
616 { "\x20\xfe\x20", -1, 1, FALSE },
617 { "\x20\xff\x20", -1, 1, FALSE },
618 /* overlong sequences */
619 { "\x20\xc0\xaf\x20", -1, 1, FALSE },
620 { "\x20\xe0\x80\xaf\x20", -1, 1, FALSE },
621 { "\x20\xf0\x80\x80\xaf\x20", -1, 1, FALSE },
622 { "\x20\xf8\x80\x80\x80\xaf\x20", -1, 1, FALSE },
623 { "\x20\xfc\x80\x80\x80\x80\xaf\x20", -1, 1, FALSE },
624 { "\x20\xc1\xbf\x20", -1, 1, FALSE },
625 { "\x20\xe0\x9f\xbf\x20", -1, 1, FALSE },
626 { "\x20\xf0\x8f\xbf\xbf\x20", -1, 1, FALSE },
627 { "\x20\xf8\x87\xbf\xbf\xbf\x20", -1, 1, FALSE },
628 { "\x20\xfc\x83\xbf\xbf\xbf\xbf\x20", -1, 1, FALSE },
629 { "\x20\xc0\x80\x20", -1, 1, FALSE },
630 { "\x20\xe0\x80\x80\x20", -1, 1, FALSE },
631 { "\x20\xf0\x80\x80\x80\x20", -1, 1, FALSE },
632 { "\x20\xf8\x80\x80\x80\x80\x20", -1, 1, FALSE },
633 { "\x20\xfc\x80\x80\x80\x80\x80\x20", -1, 1, FALSE },
634 /* illegal code positions */
635 { "\x20\xed\xa0\x80\x20", -1, 1, FALSE },
636 { "\x20\xed\xad\xbf\x20", -1, 1, FALSE },
637 { "\x20\xed\xae\x80\x20", -1, 1, FALSE },
638 { "\x20\xed\xaf\xbf\x20", -1, 1, FALSE },
639 { "\x20\xed\xb0\x80\x20", -1, 1, FALSE },
640 { "\x20\xed\xbe\x80\x20", -1, 1, FALSE },
641 { "\x20\xed\xbf\xbf\x20", -1, 1, FALSE },
642 { "\x20\xed\xa0\x80\xed\xb0\x80\x20", -1, 1, FALSE },
643 { "\x20\xed\xa0\x80\xed\xbf\xbf\x20", -1, 1, FALSE },
644 { "\x20\xed\xad\xbf\xed\xb0\x80\x20", -1, 1, FALSE },
645 { "\x20\xed\xad\xbf\xed\xbf\xbf\x20", -1, 1, FALSE },
646 { "\x20\xed\xae\x80\xed\xb0\x80\x20", -1, 1, FALSE },
647 { "\x20\xed\xae\x80\xed\xbf\xbf\x20", -1, 1, FALSE },
648 { "\x20\xed\xaf\xbf\xed\xb0\x80\x20", -1, 1, FALSE },
649 { "\x20\xed\xaf\xbf\xed\xbf\xbf\x20", -1, 1, FALSE },
650 #if 0 /* We don't consider U+FFFE / U+FFFF and similar invalid. */
651 { "\x20\xef\xbf\xbe\x20", -1, 1, FALSE },
652 { "\x20\xef\xbf\xbf\x20", -1, 1, FALSE },
653 #endif
654 { "", -1, 0, TRUE }
655 };
656
657 static void
test_buffer_utf8_validity(void)658 test_buffer_utf8_validity (void)
659 {
660 hb_buffer_t *b;
661 unsigned int i;
662
663 b = hb_buffer_create ();
664 hb_buffer_set_replacement_codepoint (b, (hb_codepoint_t) -1);
665
666 for (i = 0; i < G_N_ELEMENTS (utf8_validity_tests); i++)
667 {
668 const utf8_validity_test_t *test = &utf8_validity_tests[i];
669 unsigned int text_bytes, segment_bytes, j, len;
670 hb_glyph_info_t *glyphs;
671 char *escaped;
672
673 escaped = g_strescape (test->utf8, NULL);
674 g_test_message ("UTF-8 test #%d: %s", i, escaped);
675 g_free (escaped);
676
677 text_bytes = strlen (test->utf8);
678 if (test->max_len == -1)
679 segment_bytes = text_bytes;
680 else
681 segment_bytes = test->max_len;
682
683 hb_buffer_clear_contents (b);
684 hb_buffer_add_utf8 (b, test->utf8, text_bytes, 0, segment_bytes);
685
686 glyphs = hb_buffer_get_glyph_infos (b, &len);
687 for (j = 0; j < len; j++)
688 if (glyphs[j].codepoint == (hb_codepoint_t) -1)
689 break;
690
691 g_assert (test->valid ? j == len : j < len);
692 if (!test->valid)
693 g_assert (glyphs[j].cluster == test->offset);
694 }
695
696 hb_buffer_destroy (b);
697 }
698
699
700 typedef struct {
701 const uint16_t utf16[8];
702 const uint32_t codepoints[8];
703 } utf16_conversion_test_t;
704
705 /* note: we skip the first and last item from utf16 when adding to buffer */
706 static const utf16_conversion_test_t utf16_conversion_tests[] = {
707 {{0x41, 0x004D, 0x0430, 0x4E8C, 0xD800, 0xDF02, 0x61} , {0x004D, 0x0430, 0x4E8C, 0x10302}},
708 {{0x41, 0xD800, 0xDF02, 0x61}, {0x10302}},
709 {{0x41, 0xD800, 0xDF02}, {-1}},
710 {{0x41, 0x61, 0xD800, 0xDF02}, {0x61, -1}},
711 {{0x41, 0xD800, 0x61, 0xDF02}, {-1, 0x61}},
712 {{0x41, 0xDF00, 0x61}, {-1}},
713 {{0x41, 0x61}, {0}}
714 };
715
716 static void
test_buffer_utf16_conversion(void)717 test_buffer_utf16_conversion (void)
718 {
719 hb_buffer_t *b;
720 unsigned int i;
721
722 b = hb_buffer_create ();
723 hb_buffer_set_replacement_codepoint (b, (hb_codepoint_t) -1);
724
725 for (i = 0; i < G_N_ELEMENTS (utf16_conversion_tests); i++)
726 {
727 const utf16_conversion_test_t *test = &utf16_conversion_tests[i];
728 unsigned int u_len, chars, j, len;
729 hb_glyph_info_t *glyphs;
730
731 g_test_message ("UTF-16 test #%d", i);
732
733 for (u_len = 0; test->utf16[u_len]; u_len++)
734 ;
735 for (chars = 0; test->codepoints[chars]; chars++)
736 ;
737
738 hb_buffer_clear_contents (b);
739 hb_buffer_add_utf16 (b, test->utf16, u_len, 1, u_len - 2);
740
741 glyphs = hb_buffer_get_glyph_infos (b, &len);
742 g_assert_cmpint (len, ==, chars);
743 for (j = 0; j < chars; j++)
744 g_assert_cmphex (glyphs[j].codepoint, ==, test->codepoints[j]);
745 }
746
747 hb_buffer_destroy (b);
748 }
749
750
751 typedef struct {
752 const uint32_t utf32[8];
753 const uint32_t codepoints[8];
754 } utf32_conversion_test_t;
755
756 /* note: we skip the first and last item from utf32 when adding to buffer */
757 static const utf32_conversion_test_t utf32_conversion_tests[] = {
758 {{0x41, 0x004D, 0x0430, 0x4E8C, 0xD800, 0xDF02, 0x61} , {0x004D, 0x0430, 0x4E8C, -3, -3}},
759 {{0x41, 0x004D, 0x0430, 0x4E8C, 0x10302, 0x61} , {0x004D, 0x0430, 0x4E8C, 0x10302}},
760 {{0x41, 0xD800, 0xDF02, 0x61}, {-3, -3}},
761 {{0x41, 0xD800, 0xDF02}, {-3}},
762 {{0x41, 0x61, 0xD800, 0xDF02}, {0x61, -3}},
763 {{0x41, 0xD800, 0x61, 0xDF02}, {-3, 0x61}},
764 {{0x41, 0xDF00, 0x61}, {-3}},
765 {{0x41, 0x10FFFF, 0x61}, {0x10FFFF}},
766 {{0x41, 0x110000, 0x61}, {-3}},
767 {{0x41, 0x61}, {0}}
768 };
769
770 static void
test_buffer_utf32_conversion(void)771 test_buffer_utf32_conversion (void)
772 {
773 hb_buffer_t *b;
774 unsigned int i;
775
776 b = hb_buffer_create ();
777 hb_buffer_set_replacement_codepoint (b, (hb_codepoint_t) -3);
778
779 for (i = 0; i < G_N_ELEMENTS (utf32_conversion_tests); i++)
780 {
781 const utf32_conversion_test_t *test = &utf32_conversion_tests[i];
782 unsigned int u_len, chars, j, len;
783 hb_glyph_info_t *glyphs;
784
785 g_test_message ("UTF-32 test #%d", i);
786
787 for (u_len = 0; test->utf32[u_len]; u_len++)
788 ;
789 for (chars = 0; test->codepoints[chars]; chars++)
790 ;
791
792 hb_buffer_clear_contents (b);
793 hb_buffer_add_utf32 (b, test->utf32, u_len, 1, u_len - 2);
794
795 glyphs = hb_buffer_get_glyph_infos (b, &len);
796 g_assert_cmpint (len, ==, chars);
797 for (j = 0; j < chars; j++)
798 g_assert_cmphex (glyphs[j].codepoint, ==, test->codepoints[j]);
799 }
800
801 hb_buffer_destroy (b);
802 }
803
804
805 static void
test_empty(hb_buffer_t * b)806 test_empty (hb_buffer_t *b)
807 {
808 g_assert_cmpint (hb_buffer_get_length (b), ==, 0);
809 g_assert (!hb_buffer_get_glyph_infos (b, NULL));
810 g_assert (!hb_buffer_get_glyph_positions (b, NULL));
811 }
812
813 static void
test_buffer_empty(void)814 test_buffer_empty (void)
815 {
816 hb_buffer_t *b = hb_buffer_get_empty ();
817
818 g_assert (hb_buffer_get_empty ());
819 g_assert (hb_buffer_get_empty () == b);
820
821 g_assert (!hb_buffer_allocation_successful (b));
822
823 test_empty (b);
824
825 hb_buffer_add_utf32 (b, utf32, G_N_ELEMENTS (utf32), 1, G_N_ELEMENTS (utf32) - 2);
826
827 test_empty (b);
828
829 hb_buffer_reverse (b);
830 hb_buffer_reverse_clusters (b);
831
832 g_assert (!hb_buffer_set_length (b, 10));
833
834 test_empty (b);
835
836 g_assert (hb_buffer_set_length (b, 0));
837
838 test_empty (b);
839
840 g_assert (!hb_buffer_allocation_successful (b));
841
842 hb_buffer_reset (b);
843
844 test_empty (b);
845
846 g_assert (!hb_buffer_allocation_successful (b));
847 }
848
849 int
main(int argc,char ** argv)850 main (int argc, char **argv)
851 {
852 unsigned int i;
853
854 hb_test_init (&argc, &argv);
855
856 for (i = 0; i < BUFFER_NUM_TYPES; i++)
857 {
858 const void *buffer_type = GINT_TO_POINTER (i);
859 const char *buffer_name = buffer_names[i];
860
861 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_properties);
862 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_contents);
863 hb_test_add_fixture_flavor (fixture, buffer_type, buffer_name, test_buffer_positions);
864 }
865
866 hb_test_add_fixture (fixture, GINT_TO_POINTER (BUFFER_EMPTY), test_buffer_allocation);
867
868 hb_test_add (test_buffer_utf8_conversion);
869 hb_test_add (test_buffer_utf8_validity);
870 hb_test_add (test_buffer_utf16_conversion);
871 hb_test_add (test_buffer_utf32_conversion);
872 hb_test_add (test_buffer_empty);
873
874 return hb_test_run();
875 }
876