]>
Commit | Line | Data |
---|---|---|
f3c0d7a5 A |
1 | // © 2016 and later: Unicode, Inc. and others. |
2 | // License & terms of use: http://www.unicode.org/copyright.html | |
b75a7d8f A |
3 | /* |
4 | ****************************************************************************** | |
5 | * | |
2ca993e8 | 6 | * Copyright (C) 2001-2016, International Business Machines |
b75a7d8f A |
7 | * Corporation and others. All Rights Reserved. |
8 | * | |
9 | ****************************************************************************** | |
10 | * file name: trietest.c | |
f3c0d7a5 | 11 | * encoding: UTF-8 |
b75a7d8f A |
12 | * tab size: 8 (not used) |
13 | * indentation:4 | |
14 | * | |
15 | * created on: 2001nov20 | |
16 | * created by: Markus W. Scherer | |
17 | */ | |
18 | ||
19 | #include <stdio.h> | |
20 | #include "unicode/utypes.h" | |
4388f060 | 21 | #include "unicode/utf16.h" |
b75a7d8f A |
22 | #include "utrie.h" |
23 | #include "cstring.h" | |
24 | #include "cmemory.h" | |
25 | ||
26 | #if 1 | |
27 | #include "cintltst.h" | |
28 | #else | |
29 | /* definitions from standalone utrie development */ | |
30 | #define log_err printf | |
31 | #define log_verbose printf | |
32 | ||
33 | #undef u_errorName | |
34 | #define u_errorName(errorCode) "some error code" | |
35 | #endif | |
36 | ||
b75a7d8f A |
37 | /* Values for setting possibly overlapping, out-of-order ranges of values */ |
38 | typedef struct SetRange { | |
39 | UChar32 start, limit; | |
40 | uint32_t value; | |
41 | UBool overwrite; | |
42 | } SetRange; | |
43 | ||
44 | /* | |
45 | * Values for testing: | |
46 | * value is set from the previous boundary's limit to before | |
47 | * this boundary's limit | |
48 | */ | |
49 | typedef struct CheckRange { | |
50 | UChar32 limit; | |
51 | uint32_t value; | |
52 | } CheckRange; | |
b75a7d8f A |
53 | |
54 | ||
55 | static uint32_t U_CALLCONV | |
56 | _testFoldedValue32(UNewTrie *trie, UChar32 start, int32_t offset) { | |
57 | uint32_t foldedValue, value; | |
58 | UChar32 limit; | |
59 | UBool inBlockZero; | |
60 | ||
61 | foldedValue=0; | |
62 | ||
63 | limit=start+0x400; | |
64 | while(start<limit) { | |
65 | value=utrie_get32(trie, start, &inBlockZero); | |
66 | if(inBlockZero) { | |
67 | start+=UTRIE_DATA_BLOCK_LENGTH; | |
68 | } else { | |
69 | foldedValue|=value; | |
70 | ++start; | |
71 | } | |
72 | } | |
73 | ||
74 | if(foldedValue!=0) { | |
75 | return ((uint32_t)offset<<16)|foldedValue; | |
76 | } else { | |
77 | return 0; | |
78 | } | |
79 | } | |
80 | ||
81 | static int32_t U_CALLCONV | |
82 | _testFoldingOffset32(uint32_t data) { | |
83 | return (int32_t)(data>>16); | |
84 | } | |
85 | ||
86 | static uint32_t U_CALLCONV | |
87 | _testFoldedValue16(UNewTrie *trie, UChar32 start, int32_t offset) { | |
88 | uint32_t foldedValue, value; | |
89 | UChar32 limit; | |
90 | UBool inBlockZero; | |
91 | ||
92 | foldedValue=0; | |
93 | ||
94 | limit=start+0x400; | |
95 | while(start<limit) { | |
96 | value=utrie_get32(trie, start, &inBlockZero); | |
97 | if(inBlockZero) { | |
98 | start+=UTRIE_DATA_BLOCK_LENGTH; | |
99 | } else { | |
100 | foldedValue|=value; | |
101 | ++start; | |
102 | } | |
103 | } | |
104 | ||
105 | if(foldedValue!=0) { | |
106 | return (uint32_t)(offset|0x8000); | |
107 | } else { | |
108 | return 0; | |
109 | } | |
110 | } | |
111 | ||
112 | static int32_t U_CALLCONV | |
113 | _testFoldingOffset16(uint32_t data) { | |
114 | if(data&0x8000) { | |
115 | return (int32_t)(data&0x7fff); | |
116 | } else { | |
117 | return 0; | |
118 | } | |
119 | } | |
120 | ||
121 | static uint32_t U_CALLCONV | |
122 | _testEnumValue(const void *context, uint32_t value) { | |
340931cb | 123 | (void)context; // suppress compiler warnings about unused variable |
b75a7d8f A |
124 | return value^0x5555; |
125 | } | |
126 | ||
127 | static UBool U_CALLCONV | |
128 | _testEnumRange(const void *context, UChar32 start, UChar32 limit, uint32_t value) { | |
129 | const CheckRange **pb=(const CheckRange **)context; | |
130 | const CheckRange *b=(*pb)++; | |
131 | ||
132 | value^=0x5555; | |
133 | if(start!=(b-1)->limit || limit!=b->limit || value!=b->value) { | |
134 | log_err("error: utrie_enum() delivers wrong range [U+%04lx..U+%04lx[.0x%lx instead of [U+%04lx..U+%04lx[.0x%lx\n", | |
135 | start, limit, value, | |
136 | (b-1)->limit, b->limit, b->value); | |
137 | } | |
138 | return TRUE; | |
139 | } | |
140 | ||
141 | static void | |
142 | testTrieIteration(const char *testName, | |
143 | const UTrie *trie, | |
144 | const CheckRange checkRanges[], int32_t countCheckRanges) { | |
145 | UChar s[100]; | |
146 | uint32_t values[30]; | |
147 | ||
148 | const UChar *p, *limit; | |
149 | ||
150 | uint32_t value; | |
151 | UChar32 c; | |
152 | int32_t i, length, countValues; | |
153 | UChar c2; | |
154 | ||
155 | /* write a string */ | |
156 | length=countValues=0; | |
157 | for(i=0; i<countCheckRanges; ++i) { | |
158 | c=checkRanges[i].limit; | |
159 | if(c!=0) { | |
160 | --c; | |
4388f060 | 161 | U16_APPEND_UNSAFE(s, length, c); |
b75a7d8f A |
162 | values[countValues++]=checkRanges[i].value; |
163 | } | |
164 | } | |
165 | limit=s+length; | |
166 | ||
167 | /* try forward */ | |
168 | p=s; | |
169 | i=0; | |
170 | while(p<limit) { | |
171 | c=c2=0x33; | |
172 | if(trie->data32!=NULL) { | |
173 | UTRIE_NEXT32(trie, p, limit, c, c2, value); | |
174 | } else { | |
175 | UTRIE_NEXT16(trie, p, limit, c, c2, value); | |
176 | } | |
177 | if(value!=values[i]) { | |
178 | log_err("error: wrong value from UTRIE_NEXT(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
179 | testName, c, c2, value, values[i]); | |
180 | } | |
181 | if( | |
182 | c2==0 ? | |
183 | c!=*(p-1) : | |
4388f060 | 184 | !U16_IS_LEAD(c) || !U16_IS_TRAIL(c2) || c!=*(p-2) || c2!=*(p-1) |
b75a7d8f A |
185 | ) { |
186 | log_err("error: wrong (c, c2) from UTRIE_NEXT(%s): (U+%04lx, U+%04lx)\n", | |
187 | testName, c, c2); | |
188 | continue; | |
189 | } | |
190 | if(c2!=0) { | |
191 | int32_t offset; | |
192 | ||
193 | if(trie->data32==NULL) { | |
194 | value=UTRIE_GET16_FROM_LEAD(trie, c); | |
195 | offset=trie->getFoldingOffset(value); | |
196 | if(offset>0) { | |
197 | value=UTRIE_GET16_FROM_OFFSET_TRAIL(trie, offset, c2); | |
198 | } else { | |
199 | value=trie->initialValue; | |
200 | } | |
201 | } else { | |
202 | value=UTRIE_GET32_FROM_LEAD(trie, c); | |
203 | offset=trie->getFoldingOffset(value); | |
204 | if(offset>0) { | |
205 | value=UTRIE_GET32_FROM_OFFSET_TRAIL(trie, offset, c2); | |
206 | } else { | |
207 | value=trie->initialValue; | |
208 | } | |
209 | } | |
210 | if(value!=values[i]) { | |
211 | log_err("error: wrong value from UTRIE_GETXX_FROM_OFFSET_TRAIL(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
212 | testName, c, c2, value, values[i]); | |
213 | } | |
214 | } | |
215 | if(c2!=0) { | |
216 | value=0x44; | |
217 | if(trie->data32==NULL) { | |
218 | UTRIE_GET16_FROM_PAIR(trie, c, c2, value); | |
219 | } else { | |
220 | UTRIE_GET32_FROM_PAIR(trie, c, c2, value); | |
221 | } | |
222 | if(value!=values[i]) { | |
223 | log_err("error: wrong value from UTRIE_GETXX_FROM_PAIR(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
224 | testName, c, c2, value, values[i]); | |
225 | } | |
226 | } | |
227 | ++i; | |
228 | } | |
229 | ||
230 | /* try backward */ | |
231 | p=limit; | |
232 | i=countValues; | |
233 | while(s<p) { | |
234 | --i; | |
235 | c=c2=0x33; | |
236 | if(trie->data32!=NULL) { | |
237 | UTRIE_PREVIOUS32(trie, s, p, c, c2, value); | |
238 | } else { | |
239 | UTRIE_PREVIOUS16(trie, s, p, c, c2, value); | |
240 | } | |
241 | if(value!=values[i]) { | |
242 | log_err("error: wrong value from UTRIE_PREVIOUS(%s)(U+%04lx, U+%04lx): 0x%lx instead of 0x%lx\n", | |
243 | testName, c, c2, value, values[i]); | |
244 | } | |
245 | if( | |
246 | c2==0 ? | |
247 | c!=*p: | |
4388f060 | 248 | !U16_IS_LEAD(c) || !U16_IS_TRAIL(c2) || c!=*p || c2!=*(p+1) |
b75a7d8f A |
249 | ) { |
250 | log_err("error: wrong (c, c2) from UTRIE_PREVIOUS(%s): (U+%04lx, U+%04lx)\n", | |
251 | testName, c, c2); | |
252 | } | |
253 | } | |
254 | } | |
255 | ||
256 | static void | |
257 | testTrieRangesWithMalloc(const char *testName, | |
258 | const SetRange setRanges[], int32_t countSetRanges, | |
259 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
260 | UBool dataIs32, UBool latin1Linear) { | |
261 | UTrieGetFoldingOffset *getFoldingOffset; | |
262 | const CheckRange *enumRanges; | |
263 | UNewTrie *newTrie; | |
264 | UTrie trie={ 0 }; | |
265 | uint32_t value, value2; | |
266 | UChar32 start, limit; | |
267 | int32_t i, length; | |
268 | UErrorCode errorCode; | |
269 | UBool overwrite, ok; | |
270 | uint8_t* storage =NULL; | |
46f4442e A |
271 | static const int32_t DEFAULT_STORAGE_SIZE = 32768; |
272 | storage = (uint8_t*) uprv_malloc(sizeof(uint8_t)*DEFAULT_STORAGE_SIZE); | |
b75a7d8f A |
273 | |
274 | log_verbose("\ntesting Trie '%s'\n", testName); | |
374ca955 A |
275 | newTrie=utrie_open(NULL, NULL, 2000, |
276 | checkRanges[0].value, checkRanges[0].value, | |
277 | latin1Linear); | |
b75a7d8f A |
278 | |
279 | /* set values from setRanges[] */ | |
280 | ok=TRUE; | |
281 | for(i=0; i<countSetRanges; ++i) { | |
282 | start=setRanges[i].start; | |
283 | limit=setRanges[i].limit; | |
284 | value=setRanges[i].value; | |
285 | overwrite=setRanges[i].overwrite; | |
286 | if((limit-start)==1 && overwrite) { | |
287 | ok&=utrie_set32(newTrie, start, value); | |
288 | } else { | |
289 | ok&=utrie_setRange32(newTrie, start, limit, value, overwrite); | |
290 | } | |
291 | } | |
292 | if(!ok) { | |
293 | log_err("error: setting values into a trie failed (%s)\n", testName); | |
294 | return; | |
295 | } | |
296 | ||
297 | /* verify that all these values are in the new Trie */ | |
298 | start=0; | |
299 | for(i=0; i<countCheckRanges; ++i) { | |
300 | limit=checkRanges[i].limit; | |
301 | value=checkRanges[i].value; | |
302 | ||
303 | while(start<limit) { | |
304 | if(value!=utrie_get32(newTrie, start, NULL)) { | |
305 | log_err("error: newTrie(%s)[U+%04lx]==0x%lx instead of 0x%lx\n", | |
306 | testName, start, utrie_get32(newTrie, start, NULL), value); | |
307 | } | |
308 | ++start; | |
309 | } | |
310 | } | |
311 | ||
312 | if(dataIs32) { | |
313 | getFoldingOffset=_testFoldingOffset32; | |
314 | } else { | |
315 | getFoldingOffset=_testFoldingOffset16; | |
316 | } | |
317 | ||
318 | errorCode=U_ZERO_ERROR; | |
46f4442e | 319 | length=utrie_serialize(newTrie, storage, DEFAULT_STORAGE_SIZE, |
b75a7d8f A |
320 | dataIs32 ? _testFoldedValue32 : _testFoldedValue16, |
321 | (UBool)!dataIs32, | |
322 | &errorCode); | |
323 | if(U_FAILURE(errorCode)) { | |
324 | log_err("error: utrie_serialize(%s) failed: %s\n", testName, u_errorName(errorCode)); | |
325 | utrie_close(newTrie); | |
326 | return; | |
327 | } | |
328 | ||
329 | /* test linear Latin-1 range from utrie_getData() */ | |
330 | if(latin1Linear) { | |
331 | uint32_t *data; | |
332 | int32_t dataLength; | |
333 | ||
334 | data=utrie_getData(newTrie, &dataLength); | |
335 | start=0; | |
336 | for(i=0; i<countCheckRanges && start<=0xff; ++i) { | |
337 | limit=checkRanges[i].limit; | |
338 | value=checkRanges[i].value; | |
339 | ||
340 | while(start<limit && start<=0xff) { | |
341 | if(value!=data[UTRIE_DATA_BLOCK_LENGTH+start]) { | |
342 | log_err("error: newTrie(%s).latin1Data[U+%04lx]==0x%lx instead of 0x%lx\n", | |
343 | testName, start, data[UTRIE_DATA_BLOCK_LENGTH+start], value); | |
344 | } | |
345 | ++start; | |
346 | } | |
347 | } | |
348 | } | |
349 | ||
350 | utrie_close(newTrie); | |
351 | ||
352 | errorCode=U_ZERO_ERROR; | |
353 | if(!utrie_unserialize(&trie, storage, length, &errorCode)) { | |
354 | log_err("error: utrie_unserialize() failed, %s\n", u_errorName(errorCode)); | |
355 | return; | |
356 | } | |
357 | trie.getFoldingOffset=getFoldingOffset; | |
358 | ||
359 | if(dataIs32!=(trie.data32!=NULL)) { | |
360 | log_err("error: trie serialization (%s) did not preserve 32-bitness\n", testName); | |
361 | } | |
362 | if(latin1Linear!=trie.isLatin1Linear) { | |
363 | log_err("error: trie serialization (%s) did not preserve Latin-1-linearity\n", testName); | |
364 | } | |
365 | ||
366 | /* verify that all these values are in the unserialized Trie */ | |
367 | start=0; | |
368 | for(i=0; i<countCheckRanges; ++i) { | |
369 | limit=checkRanges[i].limit; | |
370 | value=checkRanges[i].value; | |
371 | ||
372 | if(start==0xd800) { | |
373 | /* skip surrogates */ | |
374 | start=limit; | |
375 | continue; | |
376 | } | |
377 | ||
378 | while(start<limit) { | |
379 | if(start<=0xffff) { | |
380 | if(dataIs32) { | |
381 | value2=UTRIE_GET32_FROM_BMP(&trie, start); | |
382 | } else { | |
383 | value2=UTRIE_GET16_FROM_BMP(&trie, start); | |
384 | } | |
385 | if(value!=value2) { | |
386 | log_err("error: unserialized trie(%s).fromBMP(U+%04lx)==0x%lx instead of 0x%lx\n", | |
387 | testName, start, value2, value); | |
388 | } | |
4388f060 | 389 | if(!U16_IS_LEAD(start)) { |
b75a7d8f A |
390 | if(dataIs32) { |
391 | value2=UTRIE_GET32_FROM_LEAD(&trie, start); | |
392 | } else { | |
393 | value2=UTRIE_GET16_FROM_LEAD(&trie, start); | |
394 | } | |
395 | if(value!=value2) { | |
396 | log_err("error: unserialized trie(%s).fromLead(U+%04lx)==0x%lx instead of 0x%lx\n", | |
397 | testName, start, value2, value); | |
398 | } | |
399 | } | |
400 | } | |
401 | if(dataIs32) { | |
402 | UTRIE_GET32(&trie, start, value2); | |
403 | } else { | |
404 | UTRIE_GET16(&trie, start, value2); | |
405 | } | |
406 | if(value!=value2) { | |
407 | log_err("error: unserialized trie(%s).get(U+%04lx)==0x%lx instead of 0x%lx\n", | |
408 | testName, start, value2, value); | |
409 | } | |
410 | ++start; | |
411 | } | |
412 | } | |
413 | ||
414 | /* enumerate and verify all ranges */ | |
415 | enumRanges=checkRanges+1; | |
416 | utrie_enum(&trie, _testEnumValue, _testEnumRange, &enumRanges); | |
417 | ||
418 | /* test linear Latin-1 range */ | |
419 | if(trie.isLatin1Linear) { | |
420 | if(trie.data32!=NULL) { | |
421 | const uint32_t *latin1=UTRIE_GET32_LATIN1(&trie); | |
422 | ||
423 | for(start=0; start<0x100; ++start) { | |
424 | if(latin1[start]!=UTRIE_GET32_FROM_LEAD(&trie, start)) { | |
425 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get32(U+%04lx)\n", | |
426 | testName, start, latin1[start], UTRIE_GET32_FROM_LEAD(&trie, start), start); | |
427 | } | |
428 | } | |
429 | } else { | |
430 | const uint16_t *latin1=UTRIE_GET16_LATIN1(&trie); | |
431 | ||
432 | for(start=0; start<0x100; ++start) { | |
433 | if(latin1[start]!=UTRIE_GET16_FROM_LEAD(&trie, start)) { | |
434 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get16(U+%04lx)\n", | |
435 | testName, start, latin1[start], UTRIE_GET16_FROM_LEAD(&trie, start), start); | |
436 | } | |
437 | } | |
438 | } | |
439 | } | |
440 | ||
441 | testTrieIteration(testName, &trie, checkRanges, countCheckRanges); | |
442 | uprv_free(storage); | |
443 | } | |
444 | ||
445 | static void | |
446 | testTrieRanges(const char *testName, | |
447 | const SetRange setRanges[], int32_t countSetRanges, | |
448 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
449 | UBool dataIs32, UBool latin1Linear) { | |
46f4442e A |
450 | union{ |
451 | double bogus; /* needed for aligining the storage */ | |
452 | uint8_t storage[32768]; | |
453 | } storageHolder; | |
b75a7d8f | 454 | UTrieGetFoldingOffset *getFoldingOffset; |
73c04bcf | 455 | UNewTrieGetFoldedValue *getFoldedValue; |
b75a7d8f A |
456 | const CheckRange *enumRanges; |
457 | UNewTrie *newTrie; | |
458 | UTrie trie={ 0 }; | |
459 | uint32_t value, value2; | |
460 | UChar32 start, limit; | |
461 | int32_t i, length; | |
462 | UErrorCode errorCode; | |
463 | UBool overwrite, ok; | |
464 | ||
465 | log_verbose("\ntesting Trie '%s'\n", testName); | |
374ca955 A |
466 | newTrie=utrie_open(NULL, NULL, 2000, |
467 | checkRanges[0].value, checkRanges[0].value, | |
468 | latin1Linear); | |
b75a7d8f A |
469 | |
470 | /* set values from setRanges[] */ | |
471 | ok=TRUE; | |
472 | for(i=0; i<countSetRanges; ++i) { | |
473 | start=setRanges[i].start; | |
474 | limit=setRanges[i].limit; | |
475 | value=setRanges[i].value; | |
476 | overwrite=setRanges[i].overwrite; | |
477 | if((limit-start)==1 && overwrite) { | |
478 | ok&=utrie_set32(newTrie, start, value); | |
479 | } else { | |
480 | ok&=utrie_setRange32(newTrie, start, limit, value, overwrite); | |
481 | } | |
482 | } | |
483 | if(!ok) { | |
484 | log_err("error: setting values into a trie failed (%s)\n", testName); | |
485 | return; | |
486 | } | |
487 | ||
488 | /* verify that all these values are in the new Trie */ | |
489 | start=0; | |
490 | for(i=0; i<countCheckRanges; ++i) { | |
491 | limit=checkRanges[i].limit; | |
492 | value=checkRanges[i].value; | |
493 | ||
494 | while(start<limit) { | |
495 | if(value!=utrie_get32(newTrie, start, NULL)) { | |
496 | log_err("error: newTrie(%s)[U+%04lx]==0x%lx instead of 0x%lx\n", | |
497 | testName, start, utrie_get32(newTrie, start, NULL), value); | |
498 | } | |
499 | ++start; | |
500 | } | |
501 | } | |
502 | ||
503 | if(dataIs32) { | |
504 | getFoldingOffset=_testFoldingOffset32; | |
73c04bcf | 505 | getFoldedValue=_testFoldedValue32; |
b75a7d8f A |
506 | } else { |
507 | getFoldingOffset=_testFoldingOffset16; | |
73c04bcf A |
508 | getFoldedValue=_testFoldedValue16; |
509 | } | |
510 | ||
511 | /* | |
512 | * code coverage for utrie.c/defaultGetFoldedValue(), | |
513 | * pick some combination of parameters for selecting the UTrie defaults | |
514 | */ | |
515 | if(!dataIs32 && latin1Linear) { | |
516 | getFoldingOffset=NULL; | |
517 | getFoldedValue=NULL; | |
b75a7d8f A |
518 | } |
519 | ||
520 | errorCode=U_ZERO_ERROR; | |
521 | length=utrie_serialize(newTrie, storageHolder.storage, sizeof(storageHolder.storage), | |
73c04bcf | 522 | getFoldedValue, |
b75a7d8f A |
523 | (UBool)!dataIs32, |
524 | &errorCode); | |
525 | if(U_FAILURE(errorCode)) { | |
526 | log_err("error: utrie_serialize(%s) failed: %s\n", testName, u_errorName(errorCode)); | |
527 | utrie_close(newTrie); | |
528 | return; | |
529 | } | |
46f4442e A |
530 | if (length >= (int32_t)sizeof(storageHolder.storage)) { |
531 | log_err("error: utrie_serialize(%s) needs more memory\n", testName); | |
532 | utrie_close(newTrie); | |
533 | return; | |
534 | } | |
b75a7d8f A |
535 | |
536 | /* test linear Latin-1 range from utrie_getData() */ | |
537 | if(latin1Linear) { | |
538 | uint32_t *data; | |
539 | int32_t dataLength; | |
540 | ||
541 | data=utrie_getData(newTrie, &dataLength); | |
542 | start=0; | |
543 | for(i=0; i<countCheckRanges && start<=0xff; ++i) { | |
544 | limit=checkRanges[i].limit; | |
545 | value=checkRanges[i].value; | |
546 | ||
547 | while(start<limit && start<=0xff) { | |
548 | if(value!=data[UTRIE_DATA_BLOCK_LENGTH+start]) { | |
549 | log_err("error: newTrie(%s).latin1Data[U+%04lx]==0x%lx instead of 0x%lx\n", | |
550 | testName, start, data[UTRIE_DATA_BLOCK_LENGTH+start], value); | |
551 | } | |
552 | ++start; | |
553 | } | |
554 | } | |
555 | } | |
556 | ||
557 | utrie_close(newTrie); | |
558 | ||
559 | errorCode=U_ZERO_ERROR; | |
560 | if(!utrie_unserialize(&trie, storageHolder.storage, length, &errorCode)) { | |
561 | log_err("error: utrie_unserialize() failed, %s\n", u_errorName(errorCode)); | |
562 | return; | |
563 | } | |
73c04bcf A |
564 | if(getFoldingOffset!=NULL) { |
565 | trie.getFoldingOffset=getFoldingOffset; | |
566 | } | |
b75a7d8f A |
567 | |
568 | if(dataIs32!=(trie.data32!=NULL)) { | |
569 | log_err("error: trie serialization (%s) did not preserve 32-bitness\n", testName); | |
570 | } | |
571 | if(latin1Linear!=trie.isLatin1Linear) { | |
572 | log_err("error: trie serialization (%s) did not preserve Latin-1-linearity\n", testName); | |
573 | } | |
574 | ||
575 | /* verify that all these values are in the unserialized Trie */ | |
576 | start=0; | |
577 | for(i=0; i<countCheckRanges; ++i) { | |
578 | limit=checkRanges[i].limit; | |
579 | value=checkRanges[i].value; | |
580 | ||
581 | if(start==0xd800) { | |
582 | /* skip surrogates */ | |
583 | start=limit; | |
584 | continue; | |
585 | } | |
586 | ||
587 | while(start<limit) { | |
588 | if(start<=0xffff) { | |
589 | if(dataIs32) { | |
590 | value2=UTRIE_GET32_FROM_BMP(&trie, start); | |
591 | } else { | |
592 | value2=UTRIE_GET16_FROM_BMP(&trie, start); | |
593 | } | |
594 | if(value!=value2) { | |
595 | log_err("error: unserialized trie(%s).fromBMP(U+%04lx)==0x%lx instead of 0x%lx\n", | |
596 | testName, start, value2, value); | |
597 | } | |
4388f060 | 598 | if(!U16_IS_LEAD(start)) { |
b75a7d8f A |
599 | if(dataIs32) { |
600 | value2=UTRIE_GET32_FROM_LEAD(&trie, start); | |
601 | } else { | |
602 | value2=UTRIE_GET16_FROM_LEAD(&trie, start); | |
603 | } | |
604 | if(value!=value2) { | |
605 | log_err("error: unserialized trie(%s).fromLead(U+%04lx)==0x%lx instead of 0x%lx\n", | |
606 | testName, start, value2, value); | |
607 | } | |
608 | } | |
609 | } | |
610 | if(dataIs32) { | |
611 | UTRIE_GET32(&trie, start, value2); | |
612 | } else { | |
613 | UTRIE_GET16(&trie, start, value2); | |
614 | } | |
615 | if(value!=value2) { | |
616 | log_err("error: unserialized trie(%s).get(U+%04lx)==0x%lx instead of 0x%lx\n", | |
617 | testName, start, value2, value); | |
618 | } | |
619 | ++start; | |
620 | } | |
621 | } | |
622 | ||
623 | /* enumerate and verify all ranges */ | |
624 | enumRanges=checkRanges+1; | |
625 | utrie_enum(&trie, _testEnumValue, _testEnumRange, &enumRanges); | |
626 | ||
627 | /* test linear Latin-1 range */ | |
628 | if(trie.isLatin1Linear) { | |
629 | if(trie.data32!=NULL) { | |
630 | const uint32_t *latin1=UTRIE_GET32_LATIN1(&trie); | |
631 | ||
632 | for(start=0; start<0x100; ++start) { | |
633 | if(latin1[start]!=UTRIE_GET32_FROM_LEAD(&trie, start)) { | |
634 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get32(U+%04lx)\n", | |
635 | testName, start, latin1[start], UTRIE_GET32_FROM_LEAD(&trie, start), start); | |
636 | } | |
637 | } | |
638 | } else { | |
639 | const uint16_t *latin1=UTRIE_GET16_LATIN1(&trie); | |
640 | ||
641 | for(start=0; start<0x100; ++start) { | |
642 | if(latin1[start]!=UTRIE_GET16_FROM_LEAD(&trie, start)) { | |
643 | log_err("error: (%s) trie.latin1[U+%04lx]=0x%lx!=0x%lx=trie.get16(U+%04lx)\n", | |
644 | testName, start, latin1[start], UTRIE_GET16_FROM_LEAD(&trie, start), start); | |
645 | } | |
646 | } | |
647 | } | |
648 | } | |
649 | ||
650 | testTrieIteration(testName, &trie, checkRanges, countCheckRanges); | |
651 | } | |
652 | ||
653 | static void | |
654 | testTrieRanges2(const char *testName, | |
655 | const SetRange setRanges[], int32_t countSetRanges, | |
656 | const CheckRange checkRanges[], int32_t countCheckRanges, | |
657 | UBool dataIs32) { | |
658 | char name[40]; | |
659 | ||
660 | testTrieRanges(testName, | |
661 | setRanges, countSetRanges, | |
662 | checkRanges, countCheckRanges, | |
663 | dataIs32, FALSE); | |
664 | testTrieRangesWithMalloc(testName, | |
665 | setRanges, countSetRanges, | |
666 | checkRanges, countCheckRanges, | |
667 | dataIs32, FALSE); | |
668 | ||
669 | uprv_strcpy(name, testName); | |
670 | uprv_strcat(name, "-latin1Linear"); | |
671 | testTrieRanges(name, | |
672 | setRanges, countSetRanges, | |
673 | checkRanges, countCheckRanges, | |
674 | dataIs32, TRUE); | |
675 | testTrieRangesWithMalloc(name, | |
676 | setRanges, countSetRanges, | |
677 | checkRanges, countCheckRanges, | |
678 | dataIs32, TRUE); | |
679 | } | |
680 | ||
681 | static void | |
682 | testTrieRanges4(const char *testName, | |
683 | const SetRange setRanges[], int32_t countSetRanges, | |
684 | const CheckRange checkRanges[], int32_t countCheckRanges) { | |
685 | char name[40]; | |
686 | ||
687 | uprv_strcpy(name, testName); | |
688 | uprv_strcat(name, ".32"); | |
689 | testTrieRanges2(name, | |
690 | setRanges, countSetRanges, | |
691 | checkRanges, countCheckRanges, | |
692 | TRUE); | |
693 | ||
694 | uprv_strcpy(name, testName); | |
695 | uprv_strcat(name, ".16"); | |
696 | testTrieRanges2(name, | |
697 | setRanges, countSetRanges, | |
698 | checkRanges, countCheckRanges, | |
699 | FALSE); | |
700 | } | |
701 | ||
702 | /* test data ----------------------------------------------------------------*/ | |
703 | ||
704 | /* set consecutive ranges, even with value 0 */ | |
705 | static const SetRange | |
706 | setRanges1[]={ | |
707 | {0, 0x20, 0, FALSE}, | |
708 | {0x20, 0xa7, 0x1234, FALSE}, | |
709 | {0xa7, 0x3400, 0, FALSE}, | |
710 | {0x3400, 0x9fa6, 0x6162, FALSE}, | |
374ca955 A |
711 | {0x9fa6, 0xda9e, 0x3132, FALSE}, |
712 | {0xdada, 0xeeee, 0x87ff, FALSE}, /* try to disrupt _testFoldingOffset16() */ | |
b75a7d8f A |
713 | {0xeeee, 0x11111, 1, FALSE}, |
714 | {0x11111, 0x44444, 0x6162, FALSE}, | |
374ca955 | 715 | {0x44444, 0x60003, 0, FALSE}, |
b75a7d8f A |
716 | {0xf0003, 0xf0004, 0xf, FALSE}, |
717 | {0xf0004, 0xf0006, 0x10, FALSE}, | |
718 | {0xf0006, 0xf0007, 0x11, FALSE}, | |
719 | {0xf0007, 0xf0020, 0x12, FALSE}, | |
720 | {0xf0020, 0x110000, 0, FALSE} | |
721 | }; | |
722 | ||
723 | static const CheckRange | |
724 | checkRanges1[]={ | |
725 | {0, 0}, /* dummy start range to make _testEnumRange() simpler */ | |
726 | {0x20, 0}, | |
727 | {0xa7, 0x1234}, | |
728 | {0x3400, 0}, | |
729 | {0x9fa6, 0x6162}, | |
374ca955 A |
730 | {0xda9e, 0x3132}, |
731 | {0xdada, 0}, | |
732 | {0xeeee, 0x87ff}, | |
b75a7d8f A |
733 | {0x11111,1}, |
734 | {0x44444,0x6162}, | |
735 | {0xf0003,0}, | |
736 | {0xf0004,0xf}, | |
737 | {0xf0006,0x10}, | |
738 | {0xf0007,0x11}, | |
739 | {0xf0020,0x12}, | |
740 | {0x110000, 0} | |
741 | }; | |
742 | ||
743 | /* set some interesting overlapping ranges */ | |
744 | static const SetRange | |
745 | setRanges2[]={ | |
746 | {0x21, 0x7f, 0x5555, TRUE}, | |
747 | {0x2f800,0x2fedc, 0x7a, TRUE}, | |
748 | {0x72, 0xdd, 3, TRUE}, | |
749 | {0xdd, 0xde, 4, FALSE}, | |
729e4ab9 A |
750 | {0x201, 0x220, 6, TRUE}, /* 3 consecutive blocks with the same pattern but discontiguous value ranges */ |
751 | {0x221, 0x240, 6, TRUE}, | |
752 | {0x241, 0x260, 6, TRUE}, | |
b75a7d8f A |
753 | {0x2f987,0x2fa98, 5, TRUE}, |
754 | {0x2f777,0x2f833, 0, TRUE}, | |
755 | {0x2f900,0x2ffee, 1, FALSE}, | |
756 | {0x2ffee,0x2ffef, 2, TRUE} | |
757 | }; | |
758 | ||
759 | static const CheckRange | |
760 | checkRanges2[]={ | |
761 | {0, 0}, /* dummy start range to make _testEnumRange() simpler */ | |
762 | {0x21, 0}, | |
763 | {0x72, 0x5555}, | |
764 | {0xdd, 3}, | |
765 | {0xde, 4}, | |
729e4ab9 A |
766 | {0x201, 0}, |
767 | {0x220, 6}, | |
768 | {0x221, 0}, | |
769 | {0x240, 6}, | |
770 | {0x241, 0}, | |
771 | {0x260, 6}, | |
b75a7d8f A |
772 | {0x2f833,0}, |
773 | {0x2f987,0x7a}, | |
774 | {0x2fa98,5}, | |
775 | {0x2fedc,0x7a}, | |
776 | {0x2ffee,1}, | |
777 | {0x2ffef,2}, | |
778 | {0x110000, 0} | |
779 | }; | |
780 | ||
781 | /* use a non-zero initial value */ | |
782 | static const SetRange | |
783 | setRanges3[]={ | |
784 | {0x31, 0xa4, 1, FALSE}, | |
785 | {0x3400, 0x6789, 2, FALSE}, | |
786 | {0x30000,0x34567,9, TRUE}, | |
787 | {0x45678,0x56789,3, TRUE} | |
788 | }; | |
789 | ||
790 | static const CheckRange | |
791 | checkRanges3[]={ | |
792 | {0, 9}, /* dummy start range, also carries the initial value */ | |
793 | {0x31, 9}, | |
794 | {0xa4, 1}, | |
795 | {0x3400, 9}, | |
796 | {0x6789, 2}, | |
797 | {0x45678,9}, | |
798 | {0x56789,3}, | |
799 | {0x110000,9} | |
800 | }; | |
801 | ||
802 | static void | |
803 | TrieTest(void) { | |
804 | testTrieRanges4("set1", | |
2ca993e8 A |
805 | setRanges1, UPRV_LENGTHOF(setRanges1), |
806 | checkRanges1, UPRV_LENGTHOF(checkRanges1)); | |
b75a7d8f | 807 | testTrieRanges4("set2-overlap", |
2ca993e8 A |
808 | setRanges2, UPRV_LENGTHOF(setRanges2), |
809 | checkRanges2, UPRV_LENGTHOF(checkRanges2)); | |
b75a7d8f | 810 | testTrieRanges4("set3-initial-9", |
2ca993e8 A |
811 | setRanges3, UPRV_LENGTHOF(setRanges3), |
812 | checkRanges3, UPRV_LENGTHOF(checkRanges3)); | |
b75a7d8f A |
813 | } |
814 | ||
73c04bcf A |
815 | /* test utrie_unserializeDummy() -------------------------------------------- */ |
816 | ||
817 | static int32_t U_CALLCONV | |
818 | dummyGetFoldingOffset(uint32_t data) { | |
340931cb | 819 | (void)data; // suppress compiler warnings about unused variable |
73c04bcf A |
820 | return -1; /* never get non-initialValue data for supplementary code points */ |
821 | } | |
822 | ||
823 | static void | |
824 | dummyTest(UBool make16BitTrie) { | |
46f4442e | 825 | int32_t mem[UTRIE_DUMMY_SIZE/4]; |
73c04bcf A |
826 | |
827 | UTrie trie; | |
828 | UErrorCode errorCode; | |
829 | UChar32 c; | |
830 | ||
831 | uint32_t value, initialValue, leadUnitValue; | |
832 | ||
833 | if(make16BitTrie) { | |
834 | initialValue=0x313; | |
835 | leadUnitValue=0xaffe; | |
836 | } else { | |
837 | initialValue=0x01234567; | |
838 | leadUnitValue=0x89abcdef; | |
839 | } | |
840 | ||
841 | errorCode=U_ZERO_ERROR; | |
842 | utrie_unserializeDummy(&trie, mem, sizeof(mem), initialValue, leadUnitValue, make16BitTrie, &errorCode); | |
843 | if(U_FAILURE(errorCode)) { | |
844 | log_err("utrie_unserializeDummy(make16BitTrie=%d) failed - %s\n", make16BitTrie, u_errorName(errorCode)); | |
845 | return; | |
846 | } | |
847 | trie.getFoldingOffset=dummyGetFoldingOffset; | |
848 | ||
849 | /* test that all code points have initialValue */ | |
850 | for(c=0; c<=0x10ffff; ++c) { | |
851 | if(make16BitTrie) { | |
852 | UTRIE_GET16(&trie, c, value); | |
853 | } else { | |
854 | UTRIE_GET32(&trie, c, value); | |
855 | } | |
856 | if(value!=initialValue) { | |
857 | log_err("UTRIE_GET%s(dummy, U+%04lx)=0x%lx instead of 0x%lx\n", | |
858 | make16BitTrie ? "16" : "32", (long)c, (long)value, (long)initialValue); | |
859 | } | |
860 | } | |
861 | ||
862 | /* test that the lead surrogate code units have leadUnitValue */ | |
863 | for(c=0xd800; c<=0xdbff; ++c) { | |
864 | if(make16BitTrie) { | |
865 | value=UTRIE_GET16_FROM_LEAD(&trie, c); | |
866 | } else { | |
867 | value=UTRIE_GET32_FROM_LEAD(&trie, c); | |
868 | } | |
869 | if(value!=leadUnitValue) { | |
870 | log_err("UTRIE_GET%s_FROM_LEAD(dummy, U+%04lx)=0x%lx instead of 0x%lx\n", | |
871 | make16BitTrie ? "16" : "32", (long)c, (long)value, (long)leadUnitValue); | |
872 | } | |
873 | } | |
874 | } | |
875 | ||
876 | static void | |
877 | DummyTrieTest(void) { | |
878 | dummyTest(TRUE); | |
879 | dummyTest(FALSE); | |
880 | } | |
881 | ||
b75a7d8f A |
882 | void |
883 | addTrieTest(TestNode** root); | |
884 | ||
885 | void | |
886 | addTrieTest(TestNode** root) { | |
374ca955 | 887 | addTest(root, &TrieTest, "tsutil/trietest/TrieTest"); |
73c04bcf | 888 | addTest(root, &DummyTrieTest, "tsutil/trietest/DummyTrieTest"); |
b75a7d8f | 889 | } |