-/* $Id: test_icu_I18N.c,v 1.11 2007-05-09 14:01:21 marc Exp $
+/* $Id: test_icu_I18N.c,v 1.25 2007-05-24 10:35:21 adam Exp $
Copyright (c) 2006-2007, Index Data.
This file is part of Pazpar2.
printf("\n");
}
-
+
+
+ for( i = 0; i < src_list_len; i++)
+ free(list[i]);
+
+
ucol_close(coll);
icu_buf_utf8_destroy(buf8);
icu_buf_utf16_destroy(buf16);
-
-
return success;
}
// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
-void test_icu_I18N_normmap(int argc, char **argv)
+
+
+
+int test_icu_normalizer(const char * rules8cstr,
+ const char * src8cstr,
+ const char * chk8cstr)
{
+ int success = 0;
+
+ UErrorCode status = U_ZERO_ERROR;
+
+ struct icu_buf_utf16 * src16 = icu_buf_utf16_create(0);
+ struct icu_buf_utf16 * dest16 = icu_buf_utf16_create(0);
+ struct icu_buf_utf8 * dest8 = icu_buf_utf8_create(0);
+ struct icu_normalizer * normalizer
+ = icu_normalizer_create(rules8cstr, 'f', &status);
+ icu_check_status(status);
+
+ icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
+ icu_check_status(status);
+ icu_normalizer_normalize(normalizer, dest16, src16, &status);
+ icu_check_status(status);
-}
+ icu_utf16_to_utf8(dest8, dest16, &status);
+ icu_check_status(status);
+
+
+ if(!strcmp((const char *) dest8->utf8,
+ (const char *) chk8cstr))
+ success = 1;
+ else {
+ success = 0;
+ printf("Normalization\n");
+ printf("Rules: '%s'\n", rules8cstr);
+ printf("Input: '%s'\n", src8cstr);
+ printf("Normalized: '%s'\n", dest8->utf8);
+ printf("Expected: '%s'\n", chk8cstr);
+ }
+
+
+ icu_normalizer_destroy(normalizer);
+ icu_buf_utf16_destroy(src16);
+ icu_buf_utf16_destroy(dest16);
+ icu_buf_utf8_destroy(dest8);
+
+ return success;
+};
// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
-void test_icu_I18N_tokenizer(int argc, char **argv)
+void test_icu_I18N_normalizer(int argc, char **argv)
{
- const char * src8cstr
- = "Though I am not naturally honest, I am so sometimes by chance.";
+ YAZ_CHECK(test_icu_normalizer("[:Punctuation:] Any-Remove",
+ "Don't shoot!",
+ "Dont shoot"));
+
+ YAZ_CHECK(test_icu_normalizer("[:Control:] Any-Remove",
+ "Don't\n shoot!",
+ "Don't shoot!"));
+
+ YAZ_CHECK(test_icu_normalizer("[:Decimal_Number:] Any-Remove",
+ "This is 4 you!",
+ "This is you!"));
+
+ YAZ_CHECK(test_icu_normalizer("Lower; [:^Letter:] Remove",
+ "Don't shoot!",
+ "dontshoot"));
+
+ YAZ_CHECK(test_icu_normalizer("[:^Number:] Remove",
+ "Monday 15th of April",
+ "15"));
+
+ YAZ_CHECK(test_icu_normalizer("Lower;"
+ "[[:WhiteSpace:][:Punctuation:]] Remove",
+ " word4you? ",
+ "word4you"));
+
+
+ YAZ_CHECK(test_icu_normalizer("NFD; [:Nonspacing Mark:] Remove; NFC",
+ "à côté de l'alcôve ovoïde",
+ "a cote de l'alcove ovoide"));
+
+}
+
+
+// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
+
+int test_icu_tokenizer(const char * locale, char action,
+ const char * src8cstr, int count)
+{
+ int success = 1;
UErrorCode status = U_ZERO_ERROR;
struct icu_buf_utf16 * src16 = icu_buf_utf16_create(0);
struct icu_buf_utf16 * tkn16 = icu_buf_utf16_create(0);
struct icu_buf_utf8 * tkn8 = icu_buf_utf8_create(0);
- printf("Input: '%s'\n", src8cstr);
+ //printf("Input: '%s'\n", src8cstr);
// transforming to UTF16
icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
// set up tokenizer
struct icu_tokenizer * tokenizer
- = icu_tokenizer_create("en", 's', &status);
+ = icu_tokenizer_create(locale, action, &status);
icu_check_status(status);
YAZ_CHECK(tokenizer);
YAZ_CHECK(tokenizer->bi);
// perform work on tokens
- printf("Tokens: ");
+ //printf("Tokens: ");
while(icu_tokenizer_next_token(tokenizer, tkn16, &status)){
icu_check_status(status);
// converting to UTF8
icu_utf16_to_utf8(tkn8, tkn16, &status);
- printf("'%s' ", tkn8->utf8);
-
//printf("token %d %d %d %d '%s'\n",
- // icu_tokenizer_token_id(tokenizer),
+ //
// icu_tokenizer_token_start(tokenizer),
// icu_tokenizer_token_end(tokenizer),
// icu_tokenizer_token_length(tokenizer),
// tkn8->utf8);
}
- printf(" (%d)(%d)\n", icu_tokenizer_token_id(tokenizer),
- icu_tokenizer_token_count(tokenizer));
+
+ if (count != icu_tokenizer_token_count(tokenizer)){
+ success = 0;
+ printf("\nTokenizer '%s:%c' Error: \n", locale, action);
+ printf("Input: '%s'\n", src8cstr);
+ printf("Tokens: %d", icu_tokenizer_token_count(tokenizer));
+ printf(", expected: %d\n", count);
+ }
icu_tokenizer_destroy(tokenizer);
icu_buf_utf16_destroy(src16);
icu_buf_utf16_destroy(tkn16);
icu_buf_utf8_destroy(tkn8);
+
+ return success;
}
+// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
+void test_icu_I18N_tokenizer(int argc, char **argv)
+{
+ const char * en_str
+ = "O Romeo, Romeo! wherefore art thou Romeo?";
+
+ YAZ_CHECK(test_icu_tokenizer("en", 's', en_str, 2));
+ YAZ_CHECK(test_icu_tokenizer("en", 'l', en_str, 7));
+ YAZ_CHECK(test_icu_tokenizer("en", 'w', en_str, 16));
+ YAZ_CHECK(test_icu_tokenizer("en", 'c', en_str, 41));
+
+
+
+ const char * da_str
+ = "Blåbærtærte. Denne kage stammer fra Finland. "
+ "Den er med blåbær, men alle sommerens forskellige bær kan bruges.";
+
+ YAZ_CHECK(test_icu_tokenizer("da", 's', da_str, 3));
+ YAZ_CHECK(test_icu_tokenizer("dar", 'l', da_str, 17));
+ YAZ_CHECK(test_icu_tokenizer("da", 'w', da_str, 37));
+ YAZ_CHECK(test_icu_tokenizer("da", 'c', da_str, 110));
+
+}
+
+
+void test_icu_I18N_chain(int argc, char **argv)
+{
+ const char * en_str
+ = "O Romeo, Romeo! wherefore art thou\t Romeo?";
+
+ printf("ICU chain:\ninput: '%s'\n", en_str);
+
+ UErrorCode status = U_ZERO_ERROR;
+ //struct icu_chain_step * step = 0;
+ struct icu_chain * chain = 0;
+
+
+ const char * xml_str = "<icu_chain id=\"en:word\" locale=\"en\">"
+ "<normalize rule=\"[:Control:] Any-Remove\"/>"
+ "<tokenize rule=\"l\"/>"
+ "<normalize rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
+ "<display/>"
+ "<casemap rule=\"l\"/>"
+ "<normal/>"
+ "<sort/>"
+ "</icu_chain>";
+
+
+ xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
+ xmlNode *xml_node = xmlDocGetRootElement(doc);
+ YAZ_CHECK(xml_node);
+
+
+ chain = icu_chain_xml_config(xml_node, &status);
+
+#if 0
+ chain = icu_chain_create((uint8_t *) "en:word", (uint8_t *) "en");
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_normalize,
+ (const uint8_t *) "[:Control:] Any-Remove",
+ &status);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_tokenize,
+ (const uint8_t *) "s",
+ &status);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_tokenize,
+ (const uint8_t *) "l",
+ &status);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_normalize,
+ (const uint8_t *)
+ "[[:WhiteSpace:][:Punctuation:]] Any-Remove",
+ &status);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_display,
+ (const uint8_t *)"",
+ &status);
+/* step = icu_chain_insert_step(chain, ICU_chain_step_type_normalize, */
+/* (const uint8_t *) "Lower", */
+/* &status); */
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_casemap,
+ (const uint8_t *) "l",
+ &status);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_norm,
+ (const uint8_t *)"",
+ &status);
+/* step = icu_chain_insert_step(chain, ICU_chain_step_type_sort, */
+/* (const uint8_t *)"", */
+/* &status); */
+
+#endif
+
+ xmlFreeDoc(doc);
+ YAZ_CHECK(chain);
+
+ YAZ_CHECK(icu_chain_assign_cstr(chain, en_str, &status));
+
+ while (icu_chain_next_token(chain, &status)){
+ printf("%d '%s' '%s'\n",
+ icu_chain_get_token_count(chain),
+ icu_chain_get_norm(chain),
+ icu_chain_get_display(chain));
+ }
+
+ YAZ_CHECK_EQ(icu_chain_get_token_count(chain), 7);
+
+
+ YAZ_CHECK(icu_chain_assign_cstr(chain, "what is this?", &status));
+
+ while (icu_chain_next_token(chain, &status)){
+ printf("%d '%s' '%s'\n",
+ icu_chain_get_token_count(chain),
+ icu_chain_get_norm(chain),
+ icu_chain_get_display(chain));
+ }
+
+
+ YAZ_CHECK_EQ(icu_chain_get_token_count(chain), 3);
+
+ icu_chain_destroy(chain);
+}
+
+
+void test_bug_1140(void)
+{
+ const char * en_str
+ = "O Romeo, Romeo! wherefore art thou\t Romeo?";
+
+ printf("ICU chain:\ninput: '%s'\n", en_str);
+
+ UErrorCode status = U_ZERO_ERROR;
+ //struct icu_chain_step * step = 0;
+ struct icu_chain * chain = 0;
+
+ const char * xml_str = "<icu_chain id=\"en:word\" locale=\"en\">"
+
+ /* if the first rule is normalize instead. Then it works */
+#if 0
+ "<normalize rule=\"[:Control:] Any-Remove\"/>"
+#endif
+ "<tokenize rule=\"l\"/>"
+ "<normalize rule=\"[[:WhiteSpace:][:Punctuation:]] Remove\"/>"
+ "<display/>"
+ "<casemap rule=\"l\"/>"
+ "<normal/>"
+ "<sort/>"
+ "</icu_chain>";
+
+
+ xmlDoc *doc = xmlParseMemory(xml_str, strlen(xml_str));
+ xmlNode *xml_node = xmlDocGetRootElement(doc);
+ YAZ_CHECK(xml_node);
+
+ chain = icu_chain_xml_config(xml_node, &status);
+
+ xmlFreeDoc(doc);
+ YAZ_CHECK(chain);
+
+ YAZ_CHECK(icu_chain_assign_cstr(
+ chain, "O Romeo, Romeo! wherefore art thou\t Romeo?",
+ &status));
+
+ while (icu_chain_next_token(chain, &status))
+ ;
+
+ YAZ_CHECK_EQ(icu_chain_get_token_count(chain), 7);
+
+ YAZ_CHECK(icu_chain_assign_cstr(chain, "what is this?", &status));
+
+ while (icu_chain_next_token(chain, &status)){
+ printf("%d '%s' '%s'\n",
+ icu_chain_get_token_count(chain),
+ icu_chain_get_norm(chain),
+ icu_chain_get_display(chain));
+ }
+
+ /* we expect 'what' 'is' 'this', i.e. 3 tokens */
+ YAZ_CHECK_EQ(icu_chain_get_token_count(chain), 3);
+
+ icu_chain_destroy(chain);
+}
+
#endif // HAVE_ICU
// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
//test_icu_I18N_casemap_failures(argc, argv);
test_icu_I18N_casemap(argc, argv);
test_icu_I18N_sortmap(argc, argv);
- test_icu_I18N_normmap(argc, argv);
+ test_icu_I18N_normalizer(argc, argv);
test_icu_I18N_tokenizer(argc, argv);
+ test_icu_I18N_chain(argc, argv);
+#if 0
+ /* currently fails */
+ test_bug_1140();
+#endif
#else // HAVE_ICU
-// CRAP to follow
-#if 0
-
-// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
-
-void test_icu_I18N_casemap_failures(int argc, char **argv)
-{
-
- size_t buf_cap = 128;
- char buf[buf_cap];
- size_t dest8_len = 0;
- NMEM nmem = nmem_create();
- char * dest8 = 0;
-
- const char * src8 = "A ReD fOx hunTS sQUirriLs";
- //size_t src8_len = strlen(src8);
-
- //printf("original string: '%s' (%d)\n", src8, (int) src8_len);
-
- // some calling error needs investigation
- dest8 = icu_casemap(nmem, buf, buf_cap, &dest8_len,
- src8, "en", 't');
- YAZ_CHECK(0 == dest8_len);
- //printf("icu_casemap 'en:t' '%s' (%d)\n", dest8, (int) dest8_len);
-
-
- // attention: does not fail even if no locale 'xy_zz' defined
- // it seems to default to english locale
- dest8 = icu_casemap(nmem, buf, buf_cap, &dest8_len,
- src8, "zz_abc", 'l');
- YAZ_CHECK(dest8_len);
- //printf("icu_casemap 'zz:l' '%s' (%d)\n", dest8, (int) dest8_len);
-
-
- // shall fail - no buf buffer defined
- dest8 = icu_casemap(nmem, 0, buf_cap, &dest8_len,
- src8, "en", 'l');
- YAZ_CHECK(0 == dest8_len);
- //printf("icu_casemap 'en:l' '%s' (%d)\n", dest8, (int) dest8_len);
-
- // shall fail - no buf_cap defined
- dest8 = icu_casemap(nmem, buf, 0, &dest8_len,
- src8, "en", 'l');
- YAZ_CHECK(0 == dest8_len);
- //printf("icu_casemap 'en:l' '%s' (%d)\n", dest8, (int) dest8_len);
-
- // shall fail - no action 'x' defined
- dest8 = icu_casemap(nmem, buf, buf_cap, &dest8_len,
- src8, "en", 'x');
- YAZ_CHECK(0 == dest8_len);
- //printf("icu_casemap 'en:x' '%s' (%d)\n", dest8, (int) dest8_len);
-
- nmem_destroy(nmem);
-}
-
-
-
-#endif
-
-
-
/*
* Local variables:
* c-basic-offset: 4