-/* $Id: icu_I18N.c,v 1.13 2007-05-15 15:11:42 marc Exp $
+/* $Id: icu_I18N.c,v 1.14 2007-05-16 12:39:49 marc Exp $
Copyright (c) 2006-2007, Index Data.
This file is part of Pazpar2.
struct icu_chain_step * icu_chain_step_create(struct icu_chain * chain,
enum icu_chain_step_type type,
const uint8_t * rule,
- struct icu_buf_utf16 * src16,
+ struct icu_buf_utf16 * buf16,
UErrorCode *status)
{
struct icu_chain_step * step = 0;
step = (struct icu_chain_step *) malloc(sizeof(struct icu_chain_step));
+ step->type = type;
+ step->more_tokens = 0;
+
+ if (buf16)
+ step->buf16 = buf16;
+ else
+ step->buf16 = 0;
+
// create auxilary objects
switch(step->type) {
case ICU_chain_step_type_display:
break;
}
- if (src16)
- step->src16 = src16;
-
-
return step;
};
if (!step)
return;
-
- if (step->previous)
- icu_chain_step_destroy(step->previous);
-
- if (step->src16)
- icu_buf_utf16_destroy(step->src16);
- // destroy last living icu_chain_step
+ icu_chain_step_destroy(step->previous);
switch(step->type) {
case ICU_chain_step_type_display:
case ICU_chain_step_type_sort:
break;
case ICU_chain_step_type_charmap:
+ icu_buf_utf16_destroy(step->buf16);
break;
case ICU_chain_step_type_normalize:
icu_normalizer_destroy(step->u.normalizer);
+ icu_buf_utf16_destroy(step->buf16);
break;
case ICU_chain_step_type_tokenize:
icu_tokenizer_destroy(step->u.tokenizer);
+ icu_buf_utf16_destroy(step->buf16);
break;
default:
break;
{
struct icu_chain_step * step = 0;
struct icu_buf_utf16 * src16 = 0;
+ struct icu_buf_utf16 * buf16 = 0;
if (!chain || !type || !rule)
return 0;
- //if(chain->steps && chain->steps->src16)
+ // assign utf16 src buffers as needed
+ if (chain->steps && chain->steps->buf16)
+ src16 = chain->steps->buf16;
+ else if (chain->src16)
+ src16 = chain->src16;
+ else
+ return 0;
+
- // assign utf16 src buffers as needed
- switch(step->type) {
+ // assign utf16 destination buffers as needed, or
+ // re-use previous uft18 buffer if this step does not touch it
+ switch(type) {
case ICU_chain_step_type_display:
+ buf16 = src16;
break;
case ICU_chain_step_type_norm:
+ buf16 = src16;
break;
case ICU_chain_step_type_sort:
+ buf16 = src16;
break;
case ICU_chain_step_type_charmap:
+ buf16 = icu_buf_utf16_create(0);
break;
case ICU_chain_step_type_normalize:
+ buf16 = icu_buf_utf16_create(0);
break;
case ICU_chain_step_type_tokenize:
+ buf16 = icu_buf_utf16_create(0);
break;
default:
break;
}
// create actual chain step with this buffer
- // leave zero for implicit buffer creation
- step = icu_chain_step_create(chain, type, rule, src16, status);
+ step = icu_chain_step_create(chain, type, rule, buf16, status);
step->previous = chain->steps;
chain->steps = step;
};
+int icu_chain_step_next_token(struct icu_chain * chain,
+ struct icu_chain_step * step,
+ UErrorCode *status)
+{
+ struct icu_buf_utf16 * src16 = 0;
+
+ printf("icu_chain_step_next_token %d\n", (int) step);
+
+ if (!chain || !chain->src16 || !step || !step->more_tokens)
+ return 0;
+
+ // assign utf16 src buffers as neeed, advance in previous steps
+ // tokens, and setting stop condition
+ if (step->previous){
+ src16 = step->previous->buf16;
+ step->more_tokens
+ = icu_chain_step_next_token(chain, step->previous, status);
+ }
+ else { // first step can only work once on chain->src16 input buffer
+ src16 = chain->src16;
+ step->more_tokens = 1;
+ }
+
+ // stop if nothing to process
+ // i.e new token source was not properly assigned
+ if (!step->more_tokens || !src16 || !src16->utf16_len) //
+ return 0;
+
+ printf("icu_chain_step_next_token %d working\n", (int) step);
+
+
+ // perform the work, eventually put this steps output in
+ // step->buf16 or the chains UTF8 output buffers
+ switch(step->type) {
+ case ICU_chain_step_type_display:
+ icu_utf16_to_utf8(chain->display8, src16, status);
+ break;
+ case ICU_chain_step_type_norm:
+ icu_utf16_to_utf8(chain->norm8, src16, status);
+ break;
+ case ICU_chain_step_type_sort:
+ icu_utf16_to_utf8(chain->sort8, src16, status);
+ break;
+ case ICU_chain_step_type_charmap:
+ break;
+ case ICU_chain_step_type_normalize:
+ icu_normalizer_normalize(step->u.normalizer,
+ step->buf16, src16, status);
+ break;
+ case ICU_chain_step_type_tokenize:
+ // step->more_tokens
+ // = icu_tokenizer_next_token(step->u.tokenizer,
+ // step->buf16, status);
+ break;
+ default:
+ return 0;
+ break;
+ }
+
+
+ // stop further token processing if last step
+ if (!step->previous)
+ step->more_tokens = 0;
+
+
+ if (U_FAILURE(*status))
+ return 0;
+
+ return 1;
+};
+
+
+
+int icu_chain_assign_cstr(struct icu_chain * chain,
+ const char * src8cstr,
+ UErrorCode *status)
+{
+ struct icu_chain_step * stp = chain->steps;
+
+ if (!chain || !src8cstr)
+ return 0;
+
+ // clear token count
+ chain->token_count = 0;
+
+ // clear all steps stop states
+
+ while (stp){
+ stp->more_tokens = 1;
+ stp = stp->previous;
+ }
+
+ // finally convert UTF8 to UTF16 string
+ icu_utf16_from_utf8_cstr(chain->src16, src8cstr, status);
+
+ if (U_FAILURE(*status))
+ return 0;
+
+ return 1;
+};
+
+
+
+int icu_chain_next_token(struct icu_chain * chain,
+ UErrorCode *status)
+{
+ int success = 0;
+
+ if (!chain || !chain->steps)
+ return 0;
+
+ success = icu_chain_step_next_token(chain, chain->steps, status);
+
+ if (success){
+ chain->token_count++;
+ return chain->token_count;
+ }
+
+ return 0;
+};
+
+int icu_chain_get_token_count(struct icu_chain * chain)
+{
+ if (!chain)
+ return 0;
+
+ return chain->token_count;
+};
+
+
+
+const char * icu_chain_get_display(struct icu_chain * chain)
+{
+ if (chain->display8)
+ return (const char *) chain->display8->utf8;
+
+ return 0;
+};
+
+const char * icu_chain_get_norm(struct icu_chain * chain)
+{
+ if (chain->norm8)
+ return (const char *) chain->norm8->utf8;
+
+ return 0;
+};
+
+const char * icu_chain_get_sort(struct icu_chain * chain)
+{
+ if (chain->sort8)
+ return (const char *) chain->sort8->utf8;
+
+ return 0;
+};
+
+
+
+
#endif // HAVE_ICU
-/* $Id: icu_I18N.h,v 1.13 2007-05-15 15:11:42 marc Exp $
+/* $Id: icu_I18N.h,v 1.14 2007-05-16 12:39:49 marc Exp $
Copyright (c) 2006-2007, Index Data.
This file is part of Pazpar2.
struct icu_tokenizer * tokenizer;
} u;
// temprary post-action utf16 buffer
- struct icu_buf_utf16 * src16;
+ struct icu_buf_utf16 * buf16;
struct icu_chain_step * previous;
- int end_of_tokens;
+ int more_tokens;
};
struct icu_chain_step * icu_chain_step_create(struct icu_chain * chain,
enum icu_chain_step_type type,
const uint8_t * rule,
- struct icu_buf_utf16 * src16,
+ struct icu_buf_utf16 * buf16,
UErrorCode *status);
struct icu_chain * icu_chain_create(const uint8_t * identifier,
const uint8_t * locale);
-
void icu_chain_destroy(struct icu_chain * chain);
struct icu_chain_step * icu_chain_insert_step(struct icu_chain * chain,
UErrorCode *status);
+int icu_chain_step_next_token(struct icu_chain * chain,
+ struct icu_chain_step * step,
+ UErrorCode *status);
+
+int icu_chain_assign_cstr(struct icu_chain * chain,
+ const char * src8cstr,
+ UErrorCode *status);
+
+int icu_chain_next_token(struct icu_chain * chain,
+ UErrorCode *status);
+
+int icu_chain_get_token_count(struct icu_chain * chain);
+
+const char * icu_chain_get_display(struct icu_chain * chain);
+
+const char * icu_chain_get_norm(struct icu_chain * chain);
+
+const char * icu_chain_get_sort(struct icu_chain * chain);
+
+
+
#endif // HAVE_ICU
-/* $Id: test_icu_I18N.c,v 1.18 2007-05-11 22:59:36 marc Exp $
+/* $Id: test_icu_I18N.c,v 1.19 2007-05-16 12:39:49 marc Exp $
Copyright (c) 2006-2007, Index Data.
This file is part of Pazpar2.
// DO NOT EDIT THIS FILE IF YOUR EDITOR DOES NOT SUPPORT UTF-8
+
+
+
+int test_icu_transliterator(const char * rules8cstr,
+ const char * src8cstr,
+ const char * chk8cstr)
+{
+ int success = 0;
+
+ UErrorCode status = U_ZERO_ERROR;
+ UParseError parse_error[256];
+
+
+ struct icu_buf_utf16 * rules16 = icu_buf_utf16_create(0);
+ 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);
+
+ icu_utf16_from_utf8_cstr(rules16, rules8cstr, &status);
+ icu_check_status(status);
+
+ icu_utf16_from_utf8_cstr(src16, src8cstr, &status);
+ icu_check_status(status);
+
+
+ struct icu_normalizer * normalizer
+ = icu_normalizer_create((const char *) rules16, 'f', &status);
+ icu_check_status(status);
+
+ icu_normalizer_normalize(normalizer, dest16, src16, &status);
+
+
+ icu_utf16_to_utf8(dest8, src16, &status);
+ icu_check_status(status);
+
+
+ if(!strcmp((const char *) dest8->utf8,
+ (const char *) chk8cstr))
+ success = 1;
+ else {
+ success = 0;
+ printf("Normaliozation;");
+ 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(rules16);
+ icu_buf_utf16_destroy(src16);
+ icu_buf_utf16_destroy(dest16);
+ icu_buf_utf8_destroy(dest8);
+
+ return success;
+}
+
+
+#if 0
+
int test_icu_transliterator(const char * rules8cstr,
const char * src8cstr,
const char * chk8cstr)
rules8cstr);
}
- utrans_transUChars (trans, src16->utf16, &(src16->utf16_len),
+ utrans_transUChars(trans, src16->utf16, &(src16->utf16_len),
src16->utf16_cap,
0, &(src16->utf16_len), &status);
return success;
}
-#if 0
printf("\n\nUnicode Set Patterns:\n"
" Pattern Description\n"
" Ranges [a-z] The lower case letters a through z\n"
}
+void test_icu_I18N_chain(int argc, char **argv)
+{
+ const char * en_str
+ = "O Romeo, Romeo! wherefore art thou Romeo?";
+
+ UErrorCode status = U_ZERO_ERROR;
+ struct icu_chain_step * step = 0;
+ struct icu_chain * chain
+ = icu_chain_create((uint8_t *) "en:sentence", (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 *) "w", */
+/* &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_norm,
+ (const uint8_t *)"",
+ &status);
+/* step = icu_chain_insert_step(chain, ICU_chain_step_type_sort, */
+/* (const uint8_t *)"", */
+/* &status); */
+
+
+
+
+ YAZ_CHECK(icu_chain_assign_cstr(chain, en_str, &status));
+
+ while (icu_chain_next_token(chain, &status)){
+ printf("token %d norm: '%s' display: '%s'\n",
+ icu_chain_get_token_count(chain),
+ icu_chain_get_norm(chain),
+ icu_chain_get_display(chain));
+ }
+
+ icu_chain_destroy(chain);
+}
//test_icu_I18N_casemap_failures(argc, argv);
test_icu_I18N_casemap(argc, argv);
test_icu_I18N_sortmap(argc, argv);
- test_icu_I18N_transliterator(argc, argv);
+ //test_icu_I18N_transliterator(argc, argv);
test_icu_I18N_tokenizer(argc, argv);
+ //test_icu_I18N_chain(argc, argv);
#else // HAVE_ICU