]>
Commit | Line | Data |
---|---|---|
b75a7d8f A |
1 | /* |
2 | ****************************************************************************** | |
3 | * | |
46f4442e | 4 | * Copyright (C) 2001-2006, International Business Machines |
b75a7d8f A |
5 | * Corporation and others. All Rights Reserved. |
6 | * | |
7 | ****************************************************************************** | |
8 | * file name: trietest.c | |
9 | * encoding: US-ASCII | |
10 | * tab size: 8 (not used) | |
11 | * indentation:4 | |
12 | * | |
13 | * created on: 2001nov20 | |
14 | * created by: Markus W. Scherer | |
15 | */ | |
16 | ||
17 | #include <stdio.h> | |
18 | #include "unicode/utypes.h" | |
19 | #include "utrie.h" | |
20 | #include "cstring.h" | |
21 | #include "cmemory.h" | |
22 | ||
23 | #if 1 | |
24 | #include "cintltst.h" | |
25 | #else | |
26 | /* definitions from standalone utrie development */ | |
27 | #define log_err printf | |
28 | #define log_verbose printf | |
29 | ||
30 | #undef u_errorName | |
31 | #define u_errorName(errorCode) "some error code" | |
32 | #endif | |
33 | ||
34 | #define ARRAY_LENGTH(array) (sizeof(array)/sizeof(array[0])) | |
35 | ||
36 | /* Values for setting possibly overlapping, out-of-order ranges of values */ | |
37 | typedef struct SetRange { | |
38 | UChar32 start, limit; | |
39 | uint32_t value; | |
40 | UBool overwrite; | |
41 | } SetRange; | |
42 | ||
43 | /* | |
44 | * Values for testing: | |
45 | * value is set from the previous boundary's limit to before | |
46 | * this boundary's limit | |
47 | */ | |
48 | typedef struct CheckRange { | |
49 | UChar32 limit; | |
50 | uint32_t value; | |
51 | } CheckRange; | |
b75a7d8f A |
52 | |
53 | ||
54 | static uint32_t U_CALLCONV | |
55 | _testFoldedValue32(UNewTrie *trie, UChar32 start, int32_t offset) { | |
56 | uint32_t foldedValue, value; | |
57 | UChar32 limit; | |
58 | UBool inBlockZero; | |
59 | ||
60 | foldedValue=0; | |
61 | ||
62 | limit=start+0x400; | |
63 | while(start<limit) { | |
64 | value=utrie_get32(trie, start, &inBlockZero); | |
65 | if(inBlockZero) { | |
66 | start+=UTRIE_DATA_BLOCK_LENGTH; | |
67 | } else { | |
68 | foldedValue|=value; | |
69 | ++start; | |
70 | } | |
71 | } | |
72 | ||
73 | if(foldedValue!=0) { | |
74 | return ((uint32_t)offset<<16)|foldedValue; | |
75 | } else { | |
76 | return 0; | |
77 | } | |
78 | } | |
79 | ||
80 | static int32_t U_CALLCONV | |
81 | _testFoldingOffset32(uint32_t data) { | |
82 | return (int32_t)(data>>16); | |
83 | } | |
84 | ||
85 | static uint32_t U_CALLCONV | |
86 | _testFoldedValue16(UNewTrie *trie, UChar32 start, int32_t offset) { | |
87 | uint32_t foldedValue, value; | |
88 | UChar32 limit; | |
89 | UBool inBlockZero; | |
90 | ||
91 | foldedValue=0; | |
92 | ||
93 | limit=start+0x400; | |
94 | while(start<limit) { | |
95 | value=utrie_get32(trie, start, &inBlockZero); | |
96 | if(inBlockZero) { | |
97 | start+=UTRIE_DATA_BLOCK_LENGTH; | |
98 | } else { | |
99 | foldedValue|=value; | |
100 | ++start; | |
101 | } | |
102 | } | |
103 | ||
104 | if(foldedValue!=0) { | |
105 | return (uint32_t)(offset|0x8000); | |
106 | } else { | |
107 | return 0; | |
108 | } | |
109 | } | |
110 | ||
111 | static int32_t U_CALLCONV | |
112 | _testFoldingOffset16(uint32_t data) { | |
113 | if(data&0x8000) { | |
114 | return (int32_t)(data&0x7fff); | |
115 | } else { | |
116 | return 0; | |
117 | } | |
118 | } | |
119 | ||
120 | static uint32_t U_CALLCONV | |
121 | _testEnumValue(const void *context, uint32_t value) { | |
122 | return value^0x5555; | |
123 | } | |
124 | ||
125 | static UBool U_CALLCONV | |
126 | _testEnumRange(const void *context, UChar32 start, UChar32 limit, uint32_t value) { | |
127 | const CheckRange **pb=(const CheckRange **)context; | |
128 | const CheckRange *b=(*pb)++; | |
129 | ||
130 | value^=0x5555; | |
131 | if(start!=(b-1)->limit || limit!=b->limit || value!=b->value) { | |
132 | log_err("error: utrie_enum() delivers wrong range [U+%04lx..U+%04lx[.0x%lx instead of [U+%04lx..U+%04lx[.0x%lx\n", | |
133 | start, limit, value, | |
134 | (b-1)->limit, b->limit, b->value); | |
135 | } | |
136 | return TRUE; | |
137 | } | |
138 | ||
139 | static void | |
140 | testTrieIteration(const char *testName, | |
141 | const UTrie *trie, | |
142 | const CheckRange checkRanges[], int32_t countCheckRanges) { | |
143 | UChar s[100]; | |
144 | uint32_t values[30]; | |
145 | ||
146 | const UChar *p, *limit; | |
147 | ||
148 | uint32_t value; | |
149 | UChar32 c; | |
150 | int32_t i, length, countValues; | |
151 | UChar c2; | |
152 | ||
153 | /* write a string */ | |
154 | length=countValues=0; | |
155 | for(i=0; i<countCheckRanges; ++i) { | |
156 | c=checkRanges[i].limit; | |
157 | if(c!=0) { | |
158 | --c; | |
159 | UTF_APPEND_CHAR_UNSAFE(s, length, c); | |
160 | values[countValues++]=checkRanges[i].value; | |
161 | } | |
162 | } | |
163 | limit=s+length; | |
164 | ||
165 | /* try forward */ | |
166 | p=s; | |
167 | i=0; | |
168 | while(p<limit) { | |
169 | c=c2=0x33; | |
170 | if(trie->data32!=NULL) { | |
171 | UTRIE_NEXT32(trie, p, limit, c, c2, value); | |
172 | } else { | |
173 | UTRIE_NEXT16(trie, p, limit, c, c2, value); | |
174 | } | |
175 | if(value!=values[i]) { | |
176 | log_err("error: wrong value from UTRIE_NEXT(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
177 | testName, c, c2, value, values[i]); | |
178 | } | |
179 | if( | |
180 | c2==0 ? | |
181 | c!=*(p-1) : | |
182 | !UTF_IS_LEAD(c) || !UTF_IS_TRAIL(c2) || c!=*(p-2) || c2!=*(p-1) | |
183 | ) { | |
184 | log_err("error: wrong (c, c2) from UTRIE_NEXT(%s): (U+%04lx, U+%04lx)\n", | |
185 | testName, c, c2); | |
186 | continue; | |
187 | } | |
188 | if(c2!=0) { | |
189 | int32_t offset; | |
190 | ||
191 | if(trie->data32==NULL) { | |
192 | value=UTRIE_GET16_FROM_LEAD(trie, c); | |
193 | offset=trie->getFoldingOffset(value); | |
194 | if(offset>0) { | |
195 | value=UTRIE_GET16_FROM_OFFSET_TRAIL(trie, offset, c2); | |
196 | } else { | |
197 | value=trie->initialValue; | |
198 | } | |
199 | } else { | |
200 | value=UTRIE_GET32_FROM_LEAD(trie, c); | |
201 | offset=trie->getFoldingOffset(value); | |
202 | if(offset>0) { | |
203 | value=UTRIE_GET32_FROM_OFFSET_TRAIL(trie, offset, c2); | |
204 | } else { | |
205 | value=trie->initialValue; | |
206 | } | |
207 | } | |
208 | if(value!=values[i]) { | |
209 | log_err("error: wrong value from UTRIE_GETXX_FROM_OFFSET_TRAIL(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
210 | testName, c, c2, value, values[i]); | |
211 | } | |
212 | } | |
213 | if(c2!=0) { | |
214 | value=0x44; | |
215 | if(trie->data32==NULL) { | |
216 | UTRIE_GET16_FROM_PAIR(trie, c, c2, value); | |
217 | } else { | |
218 | UTRIE_GET32_FROM_PAIR(trie, c, c2, value); | |
219 | } | |
220 | if(value!=values[i]) { | |
221 | log_err("error: wrong value from UTRIE_GETXX_FROM_PAIR(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
222 | testName, c, c2, value, values[i]); | |
223 | } | |
224 | } | |
225 | ++i; | |
226 | } | |
227 | ||
228 | /* try backward */ | |
229 | p=limit; | |
230 | i=countValues; | |
231 | while(s<p) { | |
232 | --i; | |
233 | c=c2=0x33; | |
234 | if(trie->data32!=NULL) { | |
235 | UTRIE_PREVIOUS32(trie, s, p, c, c2, value); | |
236 | } else { | |
237 | UTRIE_PREVIOUS16(trie, s, p, c, c2, value); | |
238 | } | |
239 | if(value!=values[i]) { | |
240 | log_err("error: wrong value from UTRIE_PREVIOUS(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
241 | testName, c, c2, value, values[i]); | |
242 | } | |
243 | if( | |
244 | c2==0 ? | |
245 | c!=*p: | |
246 | !UTF_IS_LEAD(c) || !UTF_IS_TRAIL(c2) || c!=*p || c2!=*(p+1) | |
247 | ) { | |
248 | log_err("error: wrong (c, c2) from UTRIE_PREVIOUS(%s): (U+%04lx, U+%04lx)\n", | |
249 | testName, c, c2); | |
250 | } | |
251 | } | |
252 | } | |
253 | ||
254 | static void | |
255 | testTrieRangesWithMalloc(const char *testName, | |
256 | const SetRange setRanges[], int32_t countSetRanges, | |
257 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
258 | UBool dataIs32, UBool latin1Linear) { | |
259 | UTrieGetFoldingOffset *getFoldingOffset; | |
260 | const CheckRange *enumRanges; | |
261 | UNewTrie *newTrie; | |
262 | UTrie trie={ 0 }; | |
263 | uint32_t value, value2; | |
264 | UChar32 start, limit; | |
265 | int32_t i, length; | |
266 | UErrorCode errorCode; | |
267 | UBool overwrite, ok; | |
268 | uint8_t* storage =NULL; | |
46f4442e A |
269 | static const int32_t DEFAULT_STORAGE_SIZE = 32768; |
270 | storage = (uint8_t*) uprv_malloc(sizeof(uint8_t)*DEFAULT_STORAGE_SIZE); | |
b75a7d8f A |
271 | |
272 | log_verbose("\ntesting Trie '%s'\n", testName); | |
374ca955 A |
273 | newTrie=utrie_open(NULL, NULL, 2000, |
274 | checkRanges[0].value, checkRanges[0].value, | |
275 | latin1Linear); | |
b75a7d8f A |
276 | |
277 | /* set values from setRanges[] */ | |
278 | ok=TRUE; | |
279 | for(i=0; i<countSetRanges; ++i) { | |
280 | start=setRanges[i].start; | |
281 | limit=setRanges[i].limit; | |
282 | value=setRanges[i].value; | |
283 | overwrite=setRanges[i].overwrite; | |
284 | if((limit-start)==1 && overwrite) { | |
285 | ok&=utrie_set32(newTrie, start, value); | |
286 | } else { | |
287 | ok&=utrie_setRange32(newTrie, start, limit, value, overwrite); | |
288 | } | |
289 | } | |
290 | if(!ok) { | |
291 | log_err("error: setting values into a trie failed (%s)\n", testName); | |
292 | return; | |
293 | } | |
294 | ||
295 | /* verify that all these values are in the new Trie */ | |
296 | start=0; | |
297 | for(i=0; i<countCheckRanges; ++i) { | |
298 | limit=checkRanges[i].limit; | |
299 | value=checkRanges[i].value; | |
300 | ||
301 | while(start<limit) { | |
302 | if(value!=utrie_get32(newTrie, start, NULL)) { | |
303 | log_err("error: newTrie(%s)[U+%04lx]==0x%lx instead of 0x%lx\n", | |
304 | testName, start, utrie_get32(newTrie, start, NULL), value); | |
305 | } | |
306 | ++start; | |
307 | } | |
308 | } | |
309 | ||
310 | if(dataIs32) { | |
311 | getFoldingOffset=_testFoldingOffset32; | |
312 | } else { | |
313 | getFoldingOffset=_testFoldingOffset16; | |
314 | } | |
315 | ||
316 | errorCode=U_ZERO_ERROR; | |
46f4442e | 317 | length=utrie_serialize(newTrie, storage, DEFAULT_STORAGE_SIZE, |
b75a7d8f A |
318 | dataIs32 ? _testFoldedValue32 : _testFoldedValue16, |
319 | (UBool)!dataIs32, | |
320 | &errorCode); | |
321 | if(U_FAILURE(errorCode)) { | |
322 | log_err("error: utrie_serialize(%s) failed: %s\n", testName, u_errorName(errorCode)); | |
323 | utrie_close(newTrie); | |
324 | return; | |
325 | } | |
326 | ||
327 | /* test linear Latin-1 range from utrie_getData() */ | |
328 | if(latin1Linear) { | |
329 | uint32_t *data; | |
330 | int32_t dataLength; | |
331 | ||
332 | data=utrie_getData(newTrie, &dataLength); | |
333 | start=0; | |
334 | for(i=0; i<countCheckRanges && start<=0xff; ++i) { | |
335 | limit=checkRanges[i].limit; | |
336 | value=checkRanges[i].value; | |
337 | ||
338 | while(start<limit && start<=0xff) { | |
339 | if(value!=data[UTRIE_DATA_BLOCK_LENGTH+start]) { | |
340 | log_err("error: newTrie(%s).latin1Data[U+%04lx]==0x%lx instead of 0x%lx\n", | |
341 | testName, start, data[UTRIE_DATA_BLOCK_LENGTH+start], value); | |
342 | } | |
343 | ++start; | |
344 | } | |
345 | } | |
346 | } | |
347 | ||
348 | utrie_close(newTrie); | |
349 | ||
350 | errorCode=U_ZERO_ERROR; | |
351 | if(!utrie_unserialize(&trie, storage, length, &errorCode)) { | |
352 | log_err("error: utrie_unserialize() failed, %s\n", u_errorName(errorCode)); | |
353 | return; | |
354 | } | |
355 | trie.getFoldingOffset=getFoldingOffset; | |
356 | ||
357 | if(dataIs32!=(trie.data32!=NULL)) { | |
358 | log_err("error: trie serialization (%s) did not preserve 32-bitness\n", testName); | |
359 | } | |
360 | if(latin1Linear!=trie.isLatin1Linear) { | |
361 | log_err("error: trie serialization (%s) did not preserve Latin-1-linearity\n", testName); | |
362 | } | |
363 | ||
364 | /* verify that all these values are in the unserialized Trie */ | |
365 | start=0; | |
366 | for(i=0; i<countCheckRanges; ++i) { | |
367 | limit=checkRanges[i].limit; | |
368 | value=checkRanges[i].value; | |
369 | ||
370 | if(start==0xd800) { | |
371 | /* skip surrogates */ | |
372 | start=limit; | |
373 | continue; | |
374 | } | |
375 | ||
376 | while(start<limit) { | |
377 | if(start<=0xffff) { | |
378 | if(dataIs32) { | |
379 | value2=UTRIE_GET32_FROM_BMP(&trie, start); | |
380 | } else { | |
381 | value2=UTRIE_GET16_FROM_BMP(&trie, start); | |
382 | } | |
383 | if(value!=value2) { | |
384 | log_err("error: unserialized trie(%s).fromBMP(U+%04lx)==0x%lx instead of 0x%lx\n", | |
385 | testName, start, value2, value); | |
386 | } | |
387 | if(!UTF_IS_LEAD(start)) { | |
388 | if(dataIs32) { | |
389 | value2=UTRIE_GET32_FROM_LEAD(&trie, start); | |
390 | } else { | |
391 | value2=UTRIE_GET16_FROM_LEAD(&trie, start); | |
392 | } | |
393 | if(value!=value2) { | |
394 | log_err("error: unserialized trie(%s).fromLead(U+%04lx)==0x%lx instead of 0x%lx\n", | |
395 | testName, start, value2, value); | |
396 | } | |
397 | } | |
398 | } | |
399 | if(dataIs32) { | |
400 | UTRIE_GET32(&trie, start, value2); | |
401 | } else { | |
402 | UTRIE_GET16(&trie, start, value2); | |
403 | } | |
404 | if(value!=value2) { | |
405 | log_err("error: unserialized trie(%s).get(U+%04lx)==0x%lx instead of 0x%lx\n", | |
406 | testName, start, value2, value); | |
407 | } | |
408 | ++start; | |
409 | } | |
410 | } | |
411 | ||
412 | /* enumerate and verify all ranges */ | |
413 | enumRanges=checkRanges+1; | |
414 | utrie_enum(&trie, _testEnumValue, _testEnumRange, &enumRanges); | |
415 | ||
416 | /* test linear Latin-1 range */ | |
417 | if(trie.isLatin1Linear) { | |
418 | if(trie.data32!=NULL) { | |
419 | const uint32_t *latin1=UTRIE_GET32_LATIN1(&trie); | |
420 | ||
421 | for(start=0; start<0x100; ++start) { | |
422 | if(latin1[start]!=UTRIE_GET32_FROM_LEAD(&trie, start)) { | |
423 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get32(U+%04lx)\n", | |
424 | testName, start, latin1[start], UTRIE_GET32_FROM_LEAD(&trie, start), start); | |
425 | } | |
426 | } | |
427 | } else { | |
428 | const uint16_t *latin1=UTRIE_GET16_LATIN1(&trie); | |
429 | ||
430 | for(start=0; start<0x100; ++start) { | |
431 | if(latin1[start]!=UTRIE_GET16_FROM_LEAD(&trie, start)) { | |
432 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get16(U+%04lx)\n", | |
433 | testName, start, latin1[start], UTRIE_GET16_FROM_LEAD(&trie, start), start); | |
434 | } | |
435 | } | |
436 | } | |
437 | } | |
438 | ||
439 | testTrieIteration(testName, &trie, checkRanges, countCheckRanges); | |
440 | uprv_free(storage); | |
441 | } | |
442 | ||
443 | static void | |
444 | testTrieRanges(const char *testName, | |
445 | const SetRange setRanges[], int32_t countSetRanges, | |
446 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
447 | UBool dataIs32, UBool latin1Linear) { | |
46f4442e A |
448 | union{ |
449 | double bogus; /* needed for aligining the storage */ | |
450 | uint8_t storage[32768]; | |
451 | } storageHolder; | |
b75a7d8f | 452 | UTrieGetFoldingOffset *getFoldingOffset; |
73c04bcf | 453 | UNewTrieGetFoldedValue *getFoldedValue; |
b75a7d8f A |
454 | const CheckRange *enumRanges; |
455 | UNewTrie *newTrie; | |
456 | UTrie trie={ 0 }; | |
457 | uint32_t value, value2; | |
458 | UChar32 start, limit; | |
459 | int32_t i, length; | |
460 | UErrorCode errorCode; | |
461 | UBool overwrite, ok; | |
462 | ||
463 | log_verbose("\ntesting Trie '%s'\n", testName); | |
374ca955 A |
464 | newTrie=utrie_open(NULL, NULL, 2000, |
465 | checkRanges[0].value, checkRanges[0].value, | |
466 | latin1Linear); | |
b75a7d8f A |
467 | |
468 | /* set values from setRanges[] */ | |
469 | ok=TRUE; | |
470 | for(i=0; i<countSetRanges; ++i) { | |
471 | start=setRanges[i].start; | |
472 | limit=setRanges[i].limit; | |
473 | value=setRanges[i].value; | |
474 | overwrite=setRanges[i].overwrite; | |
475 | if((limit-start)==1 && overwrite) { | |
476 | ok&=utrie_set32(newTrie, start, value); | |
477 | } else { | |
478 | ok&=utrie_setRange32(newTrie, start, limit, value, overwrite); | |
479 | } | |
480 | } | |
481 | if(!ok) { | |
482 | log_err("error: setting values into a trie failed (%s)\n", testName); | |
483 | return; | |
484 | } | |
485 | ||
486 | /* verify that all these values are in the new Trie */ | |
487 | start=0; | |
488 | for(i=0; i<countCheckRanges; ++i) { | |
489 | limit=checkRanges[i].limit; | |
490 | value=checkRanges[i].value; | |
491 | ||
492 | while(start<limit) { | |
493 | if(value!=utrie_get32(newTrie, start, NULL)) { | |
494 | log_err("error: newTrie(%s)[U+%04lx]==0x%lx instead of 0x%lx\n", | |
495 | testName, start, utrie_get32(newTrie, start, NULL), value); | |
496 | } | |
497 | ++start; | |
498 | } | |
499 | } | |
500 | ||
501 | if(dataIs32) { | |
502 | getFoldingOffset=_testFoldingOffset32; | |
73c04bcf | 503 | getFoldedValue=_testFoldedValue32; |
b75a7d8f A |
504 | } else { |
505 | getFoldingOffset=_testFoldingOffset16; | |
73c04bcf A |
506 | getFoldedValue=_testFoldedValue16; |
507 | } | |
508 | ||
509 | /* | |
510 | * code coverage for utrie.c/defaultGetFoldedValue(), | |
511 | * pick some combination of parameters for selecting the UTrie defaults | |
512 | */ | |
513 | if(!dataIs32 && latin1Linear) { | |
514 | getFoldingOffset=NULL; | |
515 | getFoldedValue=NULL; | |
b75a7d8f A |
516 | } |
517 | ||
518 | errorCode=U_ZERO_ERROR; | |
519 | length=utrie_serialize(newTrie, storageHolder.storage, sizeof(storageHolder.storage), | |
73c04bcf | 520 | getFoldedValue, |
b75a7d8f A |
521 | (UBool)!dataIs32, |
522 | &errorCode); | |
523 | if(U_FAILURE(errorCode)) { | |
524 | log_err("error: utrie_serialize(%s) failed: %s\n", testName, u_errorName(errorCode)); | |
525 | utrie_close(newTrie); | |
526 | return; | |
527 | } | |
46f4442e A |
528 | if (length >= (int32_t)sizeof(storageHolder.storage)) { |
529 | log_err("error: utrie_serialize(%s) needs more memory\n", testName); | |
530 | utrie_close(newTrie); | |
531 | return; | |
532 | } | |
b75a7d8f A |
533 | |
534 | /* test linear Latin-1 range from utrie_getData() */ | |
535 | if(latin1Linear) { | |
536 | uint32_t *data; | |
537 | int32_t dataLength; | |
538 | ||
539 | data=utrie_getData(newTrie, &dataLength); | |
540 | start=0; | |
541 | for(i=0; i<countCheckRanges && start<=0xff; ++i) { | |
542 | limit=checkRanges[i].limit; | |
543 | value=checkRanges[i].value; | |
544 | ||
545 | while(start<limit && start<=0xff) { | |
546 | if(value!=data[UTRIE_DATA_BLOCK_LENGTH+start]) { | |
547 | log_err("error: newTrie(%s).latin1Data[U+%04lx]==0x%lx instead of 0x%lx\n", | |
548 | testName, start, data[UTRIE_DATA_BLOCK_LENGTH+start], value); | |
549 | } | |
550 | ++start; | |
551 | } | |
552 | } | |
553 | } | |
554 | ||
555 | utrie_close(newTrie); | |
556 | ||
557 | errorCode=U_ZERO_ERROR; | |
558 | if(!utrie_unserialize(&trie, storageHolder.storage, length, &errorCode)) { | |
559 | log_err("error: utrie_unserialize() failed, %s\n", u_errorName(errorCode)); | |
560 | return; | |
561 | } | |
73c04bcf A |
562 | if(getFoldingOffset!=NULL) { |
563 | trie.getFoldingOffset=getFoldingOffset; | |
564 | } | |
b75a7d8f A |
565 | |
566 | if(dataIs32!=(trie.data32!=NULL)) { | |
567 | log_err("error: trie serialization (%s) did not preserve 32-bitness\n", testName); | |
568 | } | |
569 | if(latin1Linear!=trie.isLatin1Linear) { | |
570 | log_err("error: trie serialization (%s) did not preserve Latin-1-linearity\n", testName); | |
571 | } | |
572 | ||
573 | /* verify that all these values are in the unserialized Trie */ | |
574 | start=0; | |
575 | for(i=0; i<countCheckRanges; ++i) { | |
576 | limit=checkRanges[i].limit; | |
577 | value=checkRanges[i].value; | |
578 | ||
579 | if(start==0xd800) { | |
580 | /* skip surrogates */ | |
581 | start=limit; | |
582 | continue; | |
583 | } | |
584 | ||
585 | while(start<limit) { | |
586 | if(start<=0xffff) { | |
587 | if(dataIs32) { | |
588 | value2=UTRIE_GET32_FROM_BMP(&trie, start); | |
589 | } else { | |
590 | value2=UTRIE_GET16_FROM_BMP(&trie, start); | |
591 | } | |
592 | if(value!=value2) { | |
593 | log_err("error: unserialized trie(%s).fromBMP(U+%04lx)==0x%lx instead of 0x%lx\n", | |
594 | testName, start, value2, value); | |
595 | } | |
596 | if(!UTF_IS_LEAD(start)) { | |
597 | if(dataIs32) { | |
598 | value2=UTRIE_GET32_FROM_LEAD(&trie, start); | |
599 | } else { | |
600 | value2=UTRIE_GET16_FROM_LEAD(&trie, start); | |
601 | } | |
602 | if(value!=value2) { | |
603 | log_err("error: unserialized trie(%s).fromLead(U+%04lx)==0x%lx instead of 0x%lx\n", | |
604 | testName, start, value2, value); | |
605 | } | |
606 | } | |
607 | } | |
608 | if(dataIs32) { | |
609 | UTRIE_GET32(&trie, start, value2); | |
610 | } else { | |
611 | UTRIE_GET16(&trie, start, value2); | |
612 | } | |
613 | if(value!=value2) { | |
614 | log_err("error: unserialized trie(%s).get(U+%04lx)==0x%lx instead of 0x%lx\n", | |
615 | testName, start, value2, value); | |
616 | } | |
617 | ++start; | |
618 | } | |
619 | } | |
620 | ||
621 | /* enumerate and verify all ranges */ | |
622 | enumRanges=checkRanges+1; | |
623 | utrie_enum(&trie, _testEnumValue, _testEnumRange, &enumRanges); | |
624 | ||
625 | /* test linear Latin-1 range */ | |
626 | if(trie.isLatin1Linear) { | |
627 | if(trie.data32!=NULL) { | |
628 | const uint32_t *latin1=UTRIE_GET32_LATIN1(&trie); | |
629 | ||
630 | for(start=0; start<0x100; ++start) { | |
631 | if(latin1[start]!=UTRIE_GET32_FROM_LEAD(&trie, start)) { | |
632 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get32(U+%04lx)\n", | |
633 | testName, start, latin1[start], UTRIE_GET32_FROM_LEAD(&trie, start), start); | |
634 | } | |
635 | } | |
636 | } else { | |
637 | const uint16_t *latin1=UTRIE_GET16_LATIN1(&trie); | |
638 | ||
639 | for(start=0; start<0x100; ++start) { | |
640 | if(latin1[start]!=UTRIE_GET16_FROM_LEAD(&trie, start)) { | |
641 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get16(U+%04lx)\n", | |
642 | testName, start, latin1[start], UTRIE_GET16_FROM_LEAD(&trie, start), start); | |
643 | } | |
644 | } | |
645 | } | |
646 | } | |
647 | ||
648 | testTrieIteration(testName, &trie, checkRanges, countCheckRanges); | |
649 | } | |
650 | ||
651 | static void | |
652 | testTrieRanges2(const char *testName, | |
653 | const SetRange setRanges[], int32_t countSetRanges, | |
654 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
655 | UBool dataIs32) { | |
656 | char name[40]; | |
657 | ||
658 | testTrieRanges(testName, | |
659 | setRanges, countSetRanges, | |
660 | checkRanges, countCheckRanges, | |
661 | dataIs32, FALSE); | |
662 | testTrieRangesWithMalloc(testName, | |
663 | setRanges, countSetRanges, | |
664 | checkRanges, countCheckRanges, | |
665 | dataIs32, FALSE); | |
666 | ||
667 | uprv_strcpy(name, testName); | |
668 | uprv_strcat(name, "-latin1Linear"); | |
669 | testTrieRanges(name, | |
670 | setRanges, countSetRanges, | |
671 | checkRanges, countCheckRanges, | |
672 | dataIs32, TRUE); | |
673 | testTrieRangesWithMalloc(name, | |
674 | setRanges, countSetRanges, | |
675 | checkRanges, countCheckRanges, | |
676 | dataIs32, TRUE); | |
677 | } | |
678 | ||
679 | static void | |
680 | testTrieRanges4(const char *testName, | |
681 | const SetRange setRanges[], int32_t countSetRanges, | |
682 | const CheckRange checkRanges[], int32_t countCheckRanges) { | |
683 | char name[40]; | |
684 | ||
685 | uprv_strcpy(name, testName); | |
686 | uprv_strcat(name, ".32"); | |
687 | testTrieRanges2(name, | |
688 | setRanges, countSetRanges, | |
689 | checkRanges, countCheckRanges, | |
690 | TRUE); | |
691 | ||
692 | uprv_strcpy(name, testName); | |
693 | uprv_strcat(name, ".16"); | |
694 | testTrieRanges2(name, | |
695 | setRanges, countSetRanges, | |
696 | checkRanges, countCheckRanges, | |
697 | FALSE); | |
698 | } | |
699 | ||
700 | /* test data ----------------------------------------------------------------*/ | |
701 | ||
702 | /* set consecutive ranges, even with value 0 */ | |
703 | static const SetRange | |
704 | setRanges1[]={ | |
705 | {0, 0x20, 0, FALSE}, | |
706 | {0x20, 0xa7, 0x1234, FALSE}, | |
707 | {0xa7, 0x3400, 0, FALSE}, | |
708 | {0x3400, 0x9fa6, 0x6162, FALSE}, | |
374ca955 A |
709 | {0x9fa6, 0xda9e, 0x3132, FALSE}, |
710 | {0xdada, 0xeeee, 0x87ff, FALSE}, /* try to disrupt _testFoldingOffset16() */ | |
b75a7d8f A |
711 | {0xeeee, 0x11111, 1, FALSE}, |
712 | {0x11111, 0x44444, 0x6162, FALSE}, | |
374ca955 | 713 | {0x44444, 0x60003, 0, FALSE}, |
b75a7d8f A |
714 | {0xf0003, 0xf0004, 0xf, FALSE}, |
715 | {0xf0004, 0xf0006, 0x10, FALSE}, | |
716 | {0xf0006, 0xf0007, 0x11, FALSE}, | |
717 | {0xf0007, 0xf0020, 0x12, FALSE}, | |
718 | {0xf0020, 0x110000, 0, FALSE} | |
719 | }; | |
720 | ||
721 | static const CheckRange | |
722 | checkRanges1[]={ | |
723 | {0, 0}, /* dummy start range to make _testEnumRange() simpler */ | |
724 | {0x20, 0}, | |
725 | {0xa7, 0x1234}, | |
726 | {0x3400, 0}, | |
727 | {0x9fa6, 0x6162}, | |
374ca955 A |
728 | {0xda9e, 0x3132}, |
729 | {0xdada, 0}, | |
730 | {0xeeee, 0x87ff}, | |
b75a7d8f A |
731 | {0x11111,1}, |
732 | {0x44444,0x6162}, | |
733 | {0xf0003,0}, | |
734 | {0xf0004,0xf}, | |
735 | {0xf0006,0x10}, | |
736 | {0xf0007,0x11}, | |
737 | {0xf0020,0x12}, | |
738 | {0x110000, 0} | |
739 | }; | |
740 | ||
741 | /* set some interesting overlapping ranges */ | |
742 | static const SetRange | |
743 | setRanges2[]={ | |
744 | {0x21, 0x7f, 0x5555, TRUE}, | |
745 | {0x2f800,0x2fedc, 0x7a, TRUE}, | |
746 | {0x72, 0xdd, 3, TRUE}, | |
747 | {0xdd, 0xde, 4, FALSE}, | |
748 | {0x2f987,0x2fa98, 5, TRUE}, | |
749 | {0x2f777,0x2f833, 0, TRUE}, | |
750 | {0x2f900,0x2ffee, 1, FALSE}, | |
751 | {0x2ffee,0x2ffef, 2, TRUE} | |
752 | }; | |
753 | ||
754 | static const CheckRange | |
755 | checkRanges2[]={ | |
756 | {0, 0}, /* dummy start range to make _testEnumRange() simpler */ | |
757 | {0x21, 0}, | |
758 | {0x72, 0x5555}, | |
759 | {0xdd, 3}, | |
760 | {0xde, 4}, | |
761 | {0x2f833,0}, | |
762 | {0x2f987,0x7a}, | |
763 | {0x2fa98,5}, | |
764 | {0x2fedc,0x7a}, | |
765 | {0x2ffee,1}, | |
766 | {0x2ffef,2}, | |
767 | {0x110000, 0} | |
768 | }; | |
769 | ||
770 | /* use a non-zero initial value */ | |
771 | static const SetRange | |
772 | setRanges3[]={ | |
773 | {0x31, 0xa4, 1, FALSE}, | |
774 | {0x3400, 0x6789, 2, FALSE}, | |
775 | {0x30000,0x34567,9, TRUE}, | |
776 | {0x45678,0x56789,3, TRUE} | |
777 | }; | |
778 | ||
779 | static const CheckRange | |
780 | checkRanges3[]={ | |
781 | {0, 9}, /* dummy start range, also carries the initial value */ | |
782 | {0x31, 9}, | |
783 | {0xa4, 1}, | |
784 | {0x3400, 9}, | |
785 | {0x6789, 2}, | |
786 | {0x45678,9}, | |
787 | {0x56789,3}, | |
788 | {0x110000,9} | |
789 | }; | |
790 | ||
791 | static void | |
792 | TrieTest(void) { | |
793 | testTrieRanges4("set1", | |
794 | setRanges1, ARRAY_LENGTH(setRanges1), | |
795 | checkRanges1, ARRAY_LENGTH(checkRanges1)); | |
796 | testTrieRanges4("set2-overlap", | |
797 | setRanges2, ARRAY_LENGTH(setRanges2), | |
798 | checkRanges2, ARRAY_LENGTH(checkRanges2)); | |
799 | testTrieRanges4("set3-initial-9", | |
800 | setRanges3, ARRAY_LENGTH(setRanges3), | |
801 | checkRanges3, ARRAY_LENGTH(checkRanges3)); | |
802 | } | |
803 | ||
73c04bcf A |
804 | /* test utrie_unserializeDummy() -------------------------------------------- */ |
805 | ||
806 | static int32_t U_CALLCONV | |
807 | dummyGetFoldingOffset(uint32_t data) { | |
808 | return -1; /* never get non-initialValue data for supplementary code points */ | |
809 | } | |
810 | ||
811 | static void | |
812 | dummyTest(UBool make16BitTrie) { | |
46f4442e | 813 | int32_t mem[UTRIE_DUMMY_SIZE/4]; |
73c04bcf A |
814 | |
815 | UTrie trie; | |
816 | UErrorCode errorCode; | |
817 | UChar32 c; | |
818 | ||
819 | uint32_t value, initialValue, leadUnitValue; | |
820 | ||
821 | if(make16BitTrie) { | |
822 | initialValue=0x313; | |
823 | leadUnitValue=0xaffe; | |
824 | } else { | |
825 | initialValue=0x01234567; | |
826 | leadUnitValue=0x89abcdef; | |
827 | } | |
828 | ||
829 | errorCode=U_ZERO_ERROR; | |
830 | utrie_unserializeDummy(&trie, mem, sizeof(mem), initialValue, leadUnitValue, make16BitTrie, &errorCode); | |
831 | if(U_FAILURE(errorCode)) { | |
832 | log_err("utrie_unserializeDummy(make16BitTrie=%d) failed - %s\n", make16BitTrie, u_errorName(errorCode)); | |
833 | return; | |
834 | } | |
835 | trie.getFoldingOffset=dummyGetFoldingOffset; | |
836 | ||
837 | /* test that all code points have initialValue */ | |
838 | for(c=0; c<=0x10ffff; ++c) { | |
839 | if(make16BitTrie) { | |
840 | UTRIE_GET16(&trie, c, value); | |
841 | } else { | |
842 | UTRIE_GET32(&trie, c, value); | |
843 | } | |
844 | if(value!=initialValue) { | |
845 | log_err("UTRIE_GET%s(dummy, U+%04lx)=0x%lx instead of 0x%lx\n", | |
846 | make16BitTrie ? "16" : "32", (long)c, (long)value, (long)initialValue); | |
847 | } | |
848 | } | |
849 | ||
850 | /* test that the lead surrogate code units have leadUnitValue */ | |
851 | for(c=0xd800; c<=0xdbff; ++c) { | |
852 | if(make16BitTrie) { | |
853 | value=UTRIE_GET16_FROM_LEAD(&trie, c); | |
854 | } else { | |
855 | value=UTRIE_GET32_FROM_LEAD(&trie, c); | |
856 | } | |
857 | if(value!=leadUnitValue) { | |
858 | log_err("UTRIE_GET%s_FROM_LEAD(dummy, U+%04lx)=0x%lx instead of 0x%lx\n", | |
859 | make16BitTrie ? "16" : "32", (long)c, (long)value, (long)leadUnitValue); | |
860 | } | |
861 | } | |
862 | } | |
863 | ||
864 | static void | |
865 | DummyTrieTest(void) { | |
866 | dummyTest(TRUE); | |
867 | dummyTest(FALSE); | |
868 | } | |
869 | ||
b75a7d8f A |
870 | void |
871 | addTrieTest(TestNode** root); | |
872 | ||
873 | void | |
874 | addTrieTest(TestNode** root) { | |
374ca955 | 875 | addTest(root, &TrieTest, "tsutil/trietest/TrieTest"); |
73c04bcf | 876 | addTest(root, &DummyTrieTest, "tsutil/trietest/DummyTrieTest"); |
b75a7d8f | 877 | } |