1 /* This file is part of the YAZ toolkit.
2 * Copyright (C) 1995-2010 Index Data
3 * See the file LICENSE for details.
6 /* DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8 */
15 #include <yaz/timing.h>
21 #include <yaz/icu_I18N.h>
26 #define MAX_KEY_SIZE 256
29 uint8_t sort_key[MAX_KEY_SIZE]; /* standard C string '\0' terminated */
30 char disp_term[MAX_KEY_SIZE]; /* standard C utf-8 string */
34 int icu_termmap_cmp(const void *vp1, const void *vp2)
36 struct icu_termmap *itmp1 = *(struct icu_termmap **) vp1;
37 struct icu_termmap *itmp2 = *(struct icu_termmap **) vp2;
41 cmp = strcmp((const char *)itmp1->sort_key,
42 (const char *)itmp2->sort_key);
47 int test_icu_casemap(const char * locale, char action,
48 const char * src8cstr, const char * chk8cstr)
51 UErrorCode status = U_ZERO_ERROR;
53 struct icu_buf_utf8 * src8 = icu_buf_utf8_create(0);
54 struct icu_buf_utf8 * dest8 = icu_buf_utf8_create(0);
55 struct icu_buf_utf16 * src16 = icu_buf_utf16_create(0);
56 struct icu_buf_utf16 * dest16 = icu_buf_utf16_create(0);
59 int src8cstr_len = strlen(src8cstr);
60 int chk8cstr_len = strlen(chk8cstr);
62 /* converting to UTF16 */
63 icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
65 /* perform case mapping */
66 icu_utf16_casemap(dest16, src16, locale, action, &status);
68 /* converting to UTF8 */
69 icu_utf16_to_utf8(dest8, dest16, &status);
71 /* determine success */
73 && (dest8->utf8_len == strlen(chk8cstr))
74 && !strcmp(chk8cstr, (const char *) dest8->utf8))
82 printf("original string: '%s' (%d)\n", src8cstr, src8cstr_len);
83 printf("icu_casemap '%s:%c' '%s' (%d)\n",
84 locale, action, dest8->utf8, dest8->utf8_len);
85 printf("expected string: '%s' (%d)\n", chk8cstr, chk8cstr_len);
88 /* clean the buffers */
89 icu_buf_utf8_destroy(src8);
90 icu_buf_utf8_destroy(dest8);
91 icu_buf_utf16_destroy(src16);
92 icu_buf_utf16_destroy(dest16);
97 void test_icu_I18N_casemap(int argc, char **argv)
101 /* successful tests */
102 YAZ_CHECK(test_icu_casemap("en", 'l',
103 "A ReD fOx hunTS sQUirriLs",
104 "a red fox hunts squirrils"));
106 YAZ_CHECK(test_icu_casemap("en", 'u',
107 "A ReD fOx hunTS sQUirriLs",
108 "A RED FOX HUNTS SQUIRRILS"));
110 YAZ_CHECK(test_icu_casemap("en", 'f',
111 "A ReD fOx hunTS sQUirriLs",
112 "a red fox hunts squirrils"));
114 YAZ_CHECK(test_icu_casemap("en", 't',
115 "A ReD fOx hunTS sQUirriLs",
116 "A Red Fox Hunts Squirrils"));
120 /* success expected */
121 YAZ_CHECK(test_icu_casemap("da", 'l',
122 "åh ÆbLE, øs fLØde i Åen efter bLåBærGRødeN",
123 "åh æble, øs fløde i åen efter blåbærgrøden"));
125 YAZ_CHECK(test_icu_casemap("da", 'u',
126 "åh ÆbLE, øs fLØde i Åen efter bLåBærGRødeN",
127 "ÅH ÆBLE, ØS FLØDE I ÅEN EFTER BLÅBÆRGRØDEN"));
129 YAZ_CHECK(test_icu_casemap("da", 'f',
130 "åh ÆbLE, øs fLØde i Åen efter bLåBærGRødeN",
131 "åh æble, øs fløde i åen efter blåbærgrøden"));
133 YAZ_CHECK(test_icu_casemap("da", 't',
134 "åh ÆbLE, øs fLØde i Åen efter bLåBærGRødeN",
135 "Åh Æble, Øs Fløde I Åen Efter Blåbærgrøden"));
139 /* success expected */
140 YAZ_CHECK(test_icu_casemap("de", 'l',
141 "zWÖlf ärgerliche Würste rollen ÜBer die StRAße",
142 "zwölf ärgerliche würste rollen über die straße"));
144 YAZ_CHECK(test_icu_casemap("de", 'u',
145 "zWÖlf ärgerliche Würste rollen ÜBer die StRAße",
146 "ZWÖLF ÄRGERLICHE WÜRSTE ROLLEN ÜBER DIE STRASSE"));
148 YAZ_CHECK(test_icu_casemap("de", 'f',
149 "zWÖlf ärgerliche Würste rollen ÜBer die StRAße",
150 "zwölf ärgerliche würste rollen über die strasse"));
152 YAZ_CHECK(test_icu_casemap("de", 't',
153 "zWÖlf ärgerliche Würste rollen ÜBer die StRAße",
154 "Zwölf Ärgerliche Würste Rollen Über Die Straße"));
158 int test_icu_sortmap(const char * locale, int src_list_len,
159 const char ** src_list, const char ** chk_list)
163 UErrorCode status = U_ZERO_ERROR;
165 struct icu_buf_utf8 * buf8 = icu_buf_utf8_create(0);
166 struct icu_buf_utf16 * buf16 = icu_buf_utf16_create(0);
170 struct icu_termmap * list[src_list_len];
172 UCollator *coll = ucol_open(locale, &status);
173 icu_check_status(status);
175 if(U_FAILURE(status))
178 /* assigning display terms and sort keys using buf 8 and buf16 */
179 for (i = 0; i < src_list_len; i++)
182 list[i] = (struct icu_termmap *) malloc(sizeof(struct icu_termmap));
184 /* copy display term */
185 strcpy(list[i]->disp_term, src_list[i]);
187 /* transforming to UTF16 */
188 icu_utf16_from_utf8_cstr(buf16, list[i]->disp_term, &status);
189 icu_check_status(status);
191 /* computing sortkeys */
192 icu_sortkey8_from_utf16(coll, buf8, buf16, &status);
193 icu_check_status(status);
195 /* assigning sortkeys */
196 memcpy(list[i]->sort_key, buf8->utf8, buf8->utf8_len);
200 qsort(list, src_list_len, sizeof(struct icu_termmap *), icu_termmap_cmp);
202 /* checking correct sorting */
203 for (i = 0; i < src_list_len; i++)
205 if (0 != strcmp(list[i]->disp_term, chk_list[i])){
213 printf("Input str: '%s' : ", locale);
214 for (i = 0; i < src_list_len; i++) {
215 printf(" '%s'", list[i]->disp_term);
218 printf("ICU sort: '%s' : ", locale);
219 for (i = 0; i < src_list_len; i++) {
220 printf(" '%s'", list[i]->disp_term);
223 printf("Expected: '%s' : ", locale);
224 for (i = 0; i < src_list_len; i++) {
225 printf(" '%s'", chk_list[i]);
230 for (i = 0; i < src_list_len; i++)
235 icu_buf_utf8_destroy(buf8);
236 icu_buf_utf16_destroy(buf16);
241 void test_icu_I18N_sortmap(int argc, char **argv)
243 /* successful tests */
245 const char * en_1_src[6] = {"z", "K", "a", "A", "Z", "k"};
246 const char * en_1_cck[6] = {"a", "A", "k", "K", "z", "Z"};
247 YAZ_CHECK(test_icu_sortmap("en", en_1_len, en_1_src, en_1_cck));
248 YAZ_CHECK(test_icu_sortmap("en_AU", en_1_len, en_1_src, en_1_cck));
249 YAZ_CHECK(test_icu_sortmap("en_CA", en_1_len, en_1_src, en_1_cck));
250 YAZ_CHECK(test_icu_sortmap("en_GB", en_1_len, en_1_src, en_1_cck));
251 YAZ_CHECK(test_icu_sortmap("en_US", en_1_len, en_1_src, en_1_cck));
253 /* successful tests */
256 const char * da_1_src[6] = {"z", "å", "o", "æ", "a", "ø"};
257 const char * da_1_cck[6] = {"a", "o", "z", "æ", "ø", "å"};
258 YAZ_CHECK(test_icu_sortmap("da", da_1_len, da_1_src, da_1_cck));
259 YAZ_CHECK(test_icu_sortmap("da_DK", da_1_len, da_1_src, da_1_cck));
261 /* successful tests */
264 const char * de_1_src[9] = {"u", "ä", "o", "t", "s", "ß", "ü", "ö", "a"};
265 const char * de_1_cck[9] = {"a","ä", "o", "ö", "s", "ß", "t", "u", "ü"};
266 YAZ_CHECK(test_icu_sortmap("de", de_1_len, de_1_src, de_1_cck));
267 YAZ_CHECK(test_icu_sortmap("de_AT", de_1_len, de_1_src, de_1_cck));
268 YAZ_CHECK(test_icu_sortmap("de_DE", de_1_len, de_1_src, de_1_cck));
273 int test_icu_normalizer(const char * rules8cstr,
274 const char * src8cstr,
275 const char * chk8cstr)
279 UErrorCode status = U_ZERO_ERROR;
281 struct icu_buf_utf16 * src16 = icu_buf_utf16_create(0);
282 struct icu_buf_utf16 * dest16 = icu_buf_utf16_create(0);
283 struct icu_buf_utf8 * dest8 = icu_buf_utf8_create(0);
284 struct icu_transform * transform
285 = icu_transform_create(rules8cstr, 'f', 0, &status);
286 icu_check_status(status);
288 icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
289 icu_check_status(status);
291 icu_transform_trans(transform, dest16, src16, &status);
292 icu_check_status(status);
294 icu_utf16_to_utf8(dest8, dest16, &status);
295 icu_check_status(status);
298 if(!strcmp((const char *) dest8->utf8,
299 (const char *) chk8cstr))
304 printf("Normalization\n");
305 printf("Rules: '%s'\n", rules8cstr);
306 printf("Input: '%s'\n", src8cstr);
307 printf("Normalized: '%s'\n", dest8->utf8);
308 printf("Expected: '%s'\n", chk8cstr);
311 icu_transform_destroy(transform);
312 icu_buf_utf16_destroy(src16);
313 icu_buf_utf16_destroy(dest16);
314 icu_buf_utf8_destroy(dest8);
319 void test_icu_I18N_normalizer(int argc, char **argv)
321 YAZ_CHECK(test_icu_normalizer("[:Punctuation:] Any-Remove",
325 YAZ_CHECK(test_icu_normalizer("[:Control:] Any-Remove",
329 YAZ_CHECK(test_icu_normalizer("[:Decimal_Number:] Any-Remove",
333 YAZ_CHECK(test_icu_normalizer("Lower; [:^Letter:] Remove",
337 YAZ_CHECK(test_icu_normalizer("[:^Number:] Remove",
338 "Monday 15th of April",
341 YAZ_CHECK(test_icu_normalizer("Lower;"
342 "[[:WhiteSpace:][:Punctuation:]] Remove",
346 YAZ_CHECK(test_icu_normalizer("NFD; [:Nonspacing Mark:] Remove; NFC",
347 "à côté de l'alcôve ovoïde",
348 "a cote de l'alcove ovoide"));
351 int test_icu_tokenizer(const char * locale, char action,
352 const char * src8cstr, int count)
356 UErrorCode status = U_ZERO_ERROR;
357 struct icu_buf_utf16 * src16 = icu_buf_utf16_create(0);
358 struct icu_buf_utf16 * tkn16 = icu_buf_utf16_create(0);
359 struct icu_buf_utf8 * tkn8 = icu_buf_utf8_create(0);
360 struct icu_tokenizer * tokenizer = 0;
362 /* transforming to UTF16 */
363 icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
364 icu_check_status(status);
366 /* set up tokenizer */
367 tokenizer = icu_tokenizer_create(locale, action, &status);
368 icu_check_status(status);
369 YAZ_CHECK(tokenizer);
371 /* attach text buffer to tokenizer */
372 icu_tokenizer_attach(tokenizer, src16, &status);
373 icu_check_status(status);
374 YAZ_CHECK(tokenizer->bi);
376 /* perform work on tokens */
377 while(icu_tokenizer_next_token(tokenizer, tkn16, &status)){
378 icu_check_status(status);
380 /* converting to UTF8 */
381 icu_utf16_to_utf8(tkn8, tkn16, &status);
384 if (count != icu_tokenizer_token_count(tokenizer)){
386 printf("\nTokenizer '%s:%c' Error: \n", locale, action);
387 printf("Input: '%s'\n", src8cstr);
388 printf("Tokens: %d", icu_tokenizer_token_count(tokenizer));
389 printf(", expected: %d\n", count);
392 icu_tokenizer_destroy(tokenizer);
393 icu_buf_utf16_destroy(src16);
394 icu_buf_utf16_destroy(tkn16);
395 icu_buf_utf8_destroy(tkn8);
400 void test_icu_I18N_tokenizer(int argc, char **argv)
403 = "O Romeo, Romeo! wherefore art thou Romeo?";
405 YAZ_CHECK(test_icu_tokenizer("en", 's', en_str, 2));
406 YAZ_CHECK(test_icu_tokenizer("en", 'l', en_str, 7));
407 YAZ_CHECK(test_icu_tokenizer("en", 'w', en_str, 16));
408 YAZ_CHECK(test_icu_tokenizer("en", 'c', en_str, 41));
412 = "Blåbærtærte. Denne kage stammer fra Finland. "
413 "Den er med blåbær, men alle sommerens forskellige bær kan bruges.";
415 YAZ_CHECK(test_icu_tokenizer("da", 's', da_str, 3));
416 YAZ_CHECK(test_icu_tokenizer("dar", 'l', da_str, 17));
417 YAZ_CHECK(test_icu_tokenizer("da", 'w', da_str, 37));
418 YAZ_CHECK(test_icu_tokenizer("da", 'c', da_str, 110));
422 void test_icu_I18N_chain(int argc, char **argv)
425 = "O Romeo, Romeo! wherefore art thou\t Romeo?";
427 UErrorCode status = U_ZERO_ERROR;
428 struct icu_chain * chain = 0;
430 const char * xml_str = "<icu locale=\"en\">"
431 "<transform rule=\"[:Control:] Any-Remove\"/>"
432 "<tokenize rule=\"l\"/>"
433 "<transform rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
435 "<casemap rule=\"l\"/>"
439 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
440 xmlNode *xml_node = xmlDocGetRootElement(doc);
443 chain = icu_chain_xml_config(xml_node, 0, &status);
448 YAZ_CHECK(icu_chain_assign_cstr(chain, en_str, &status));
450 while (icu_chain_next_token(chain, &status))
453 /* printf("%d '%s' '%s'\n",
454 icu_chain_token_number(chain),
455 icu_chain_token_norm(chain),
456 icu_chain_token_display(chain)); */
459 YAZ_CHECK_EQ(icu_chain_token_number(chain), 7);
462 YAZ_CHECK(icu_chain_assign_cstr(chain, "what is this?", &status));
464 while (icu_chain_next_token(chain, &status))
467 /* printf("%d '%s' '%s'\n",
468 icu_chain_token_number(chain),
469 icu_chain_token_norm(chain),
470 icu_chain_token_display(chain)); */
474 YAZ_CHECK_EQ(icu_chain_token_number(chain), 3);
476 icu_chain_destroy(chain);
480 void test_bug_1140(void)
482 UErrorCode status = U_ZERO_ERROR;
483 struct icu_chain * chain = 0;
485 const char * xml_str = "<icu locale=\"en\">"
487 /* if the first rule is normalize instead. Then it works */
489 "<transform rule=\"[:Control:] Any-Remove\"/>"
491 "<tokenize rule=\"l\"/>"
492 "<transform rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
494 "<casemap rule=\"l\"/>"
498 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
499 xmlNode *xml_node = xmlDocGetRootElement(doc);
502 chain = icu_chain_xml_config(xml_node, 0, &status);
507 YAZ_CHECK(icu_chain_assign_cstr(
508 chain, "O Romeo, Romeo! wherefore art thou\t Romeo?",
511 while (icu_chain_next_token(chain, &status))
514 /* printf("%d '%s' '%s'\n",
515 icu_chain_token_number(chain),
516 icu_chain_token_norm(chain),
517 icu_chain_token_display(chain)); */
521 YAZ_CHECK_EQ(icu_chain_token_number(chain), 7);
523 YAZ_CHECK(icu_chain_assign_cstr(chain, "what is this?", &status));
525 while (icu_chain_next_token(chain, &status))
528 /* printf("%d '%s' '%s'\n",
529 icu_chain_token_number(chain),
530 icu_chain_token_norm(chain),
531 icu_chain_token_display(chain)); */
534 /* we expect 'what' 'is' 'this', i.e. 3 tokens */
535 YAZ_CHECK_EQ(icu_chain_token_number(chain), 3);
537 icu_chain_destroy(chain);
541 void test_chain_empty_token(void)
543 UErrorCode status = U_ZERO_ERROR;
544 struct icu_chain * chain = 0;
546 const char * xml_str = "<icu locale=\"en\">"
547 "<tokenize rule=\"w\"/>"
548 "<transform rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
551 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
552 xmlNode *xml_node = xmlDocGetRootElement(doc);
555 chain = icu_chain_xml_config(xml_node, 0, &status);
560 YAZ_CHECK(icu_chain_assign_cstr(
561 chain, "a string with 15 tokenss and 8 displays",
564 while (icu_chain_next_token(chain, &status))
567 /* printf("%d '%s' '%s'\n",
568 icu_chain_token_number(chain),
569 icu_chain_token_norm(chain),
570 icu_chain_token_display(chain)); */
573 YAZ_CHECK_EQ(icu_chain_token_number(chain), 15);
575 icu_chain_destroy(chain);
578 void test_chain_empty_chain(void)
580 UErrorCode status = U_ZERO_ERROR;
581 struct icu_chain * chain = 0;
583 const char * xml_str = "<icu locale=\"en\">"
586 const char * src8 = "some 5487 weired !¤%&(/& sTuFf";
589 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
590 xmlNode *xml_node = xmlDocGetRootElement(doc);
593 chain = icu_chain_xml_config(xml_node, 0, &status);
598 YAZ_CHECK(icu_chain_assign_cstr(
602 while (icu_chain_next_token(chain, &status))
605 /* printf("%d '%s' '%s'\n",
606 icu_chain_token_number(chain),
607 icu_chain_token_norm(chain),
608 icu_chain_token_display(chain)); */
611 YAZ_CHECK_EQ(icu_chain_token_number(chain), 1);
613 dest8 = (char *) icu_chain_token_norm(chain);
614 YAZ_CHECK_EQ(strcmp(src8, dest8), 0);
616 icu_chain_destroy(chain);
619 void test_icu_iter1(void)
621 UErrorCode status = U_ZERO_ERROR;
622 struct icu_chain * chain = 0;
624 struct icu_iter *iter;
625 struct icu_buf_utf8 *token;
627 const char * xml_str = "<icu locale=\"en\">"
628 "<tokenize rule=\"w\"/>"
629 "<transform rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
632 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
636 xml_node = xmlDocGetRootElement(doc);
641 chain = icu_chain_xml_config(xml_node, 0, &status);
646 iter = icu_iter_create(chain, "a string with 15 tokens and 8 displays");
651 token = icu_buf_utf8_create(0);
652 while (icu_iter_next(iter, token))
654 printf("[%.*s]", (int) token->utf8_len, token->utf8);
656 icu_buf_utf8_destroy(token);
659 icu_iter_destroy(iter);
660 icu_chain_destroy(chain);
664 void test_icu_iter2(void)
666 UErrorCode status = U_ZERO_ERROR;
667 struct icu_chain * chain = 0;
669 struct icu_iter *iter;
670 struct icu_buf_utf8 *token;
672 const char * xml_str = "<icu locale=\"en\">"
673 "<transform rule=\"[:Control:] Any-Remove\"/>"
674 "<tokenize rule=\"l\"/>"
675 "<tokenize rule=\"w\"/>"
676 "<transform rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
678 "<casemap rule=\"l\"/>"
681 xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
685 xml_node = xmlDocGetRootElement(doc);
690 chain = icu_chain_xml_config(xml_node, 0, &status);
695 iter = icu_iter_create(chain, "Adobe Acrobat Reader, 1991-1999.");
699 token = icu_buf_utf8_create(0);
700 while (icu_iter_next(iter, token))
702 printf("[%.*s]", (int) token->utf8_len, token->utf8);
704 icu_buf_utf8_destroy(token);
706 icu_iter_destroy(iter);
707 icu_chain_destroy(chain);
710 #endif /* YAZ_HAVE_ICU */
712 int main(int argc, char **argv)
714 YAZ_CHECK_INIT(argc, argv);
719 test_icu_I18N_casemap(argc, argv);
720 test_icu_I18N_sortmap(argc, argv);
721 test_icu_I18N_normalizer(argc, argv);
722 test_icu_I18N_tokenizer(argc, argv);
723 test_icu_I18N_chain(argc, argv);
724 test_chain_empty_token();
725 test_chain_empty_chain();
731 #else /* YAZ_HAVE_ICU */
733 printf("ICU unit tests omitted.\n"
734 "Please install libicu36-dev and icu-doc or similar\n");
737 #endif /* YAZ_HAVE_ICU */
745 * c-file-style: "Stroustrup"
746 * indent-tabs-mode: nil
748 * vim: shiftwidth=4 tabstop=8 expandtab