/* This file is part of the YAZ toolkit.
- * Copyright (C) 1995-2008 Index Data
+ * Copyright (C) 1995-2009 Index Data
* See the file LICENSE for details.
*/
#include <unicode/ustring.h> /* some more string fcns*/
#include <unicode/uchar.h> /* char names */
-
-
#include <unicode/ucol.h>
-
-int icu_check_status (UErrorCode status)
+int icu_check_status(UErrorCode status)
{
if (U_FAILURE(status))
{
return 0;
}
return 1;
-
}
-
-
struct icu_buf_utf16 * icu_buf_utf16_create(size_t capacity)
{
struct icu_buf_utf16 * buf16
buf16->utf16_len = 0;
buf16->utf16_cap = 0;
- if (capacity > 0){
+ if (capacity > 0)
+ {
buf16->utf16 = (UChar *) xmalloc(sizeof(UChar) * capacity);
buf16->utf16[0] = (UChar) 0;
buf16->utf16_cap = capacity;
struct icu_buf_utf16 * icu_buf_utf16_clear(struct icu_buf_utf16 * buf16)
{
- if (buf16){
+ if (buf16)
+ {
if (buf16->utf16)
buf16->utf16[0] = (UChar) 0;
buf16->utf16_len = 0;
if (!buf16)
return 0;
- if (capacity > 0){
+ if (capacity > 0)
+ {
if (0 == buf16->utf16)
buf16->utf16 = (UChar *) xmalloc(sizeof(UChar) * capacity);
else
icu_buf_utf16_clear(buf16);
buf16->utf16_cap = capacity;
}
- else {
+ else
+ {
xfree(buf16->utf16);
buf16->utf16 = 0;
buf16->utf16_len = 0;
struct icu_buf_utf16 * icu_buf_utf16_copy(struct icu_buf_utf16 * dest16,
struct icu_buf_utf16 * src16)
{
- if(!dest16 || !src16
- || dest16 == src16)
+ if (!dest16 || !src16 || dest16 == src16)
return 0;
if (dest16->utf16_cap < src16->utf16_len)
return dest16;
}
-
void icu_buf_utf16_destroy(struct icu_buf_utf16 * buf16)
{
if (buf16)
xfree(buf16);
}
-
-
struct icu_buf_utf8 * icu_buf_utf8_create(size_t capacity)
{
struct icu_buf_utf8 * buf8
buf8->utf8_len = 0;
buf8->utf8_cap = 0;
- if (capacity > 0){
+ if (capacity > 0)
+ {
buf8->utf8 = (uint8_t *) xmalloc(sizeof(uint8_t) * capacity);
buf8->utf8[0] = (uint8_t) 0;
buf8->utf8_cap = capacity;
return buf8;
}
-
struct icu_buf_utf8 * icu_buf_utf8_clear(struct icu_buf_utf8 * buf8)
{
- if (buf8){
+ if (buf8)
+ {
if (buf8->utf8)
buf8->utf8[0] = (uint8_t) 0;
buf8->utf8_len = 0;
return buf8;
}
-
struct icu_buf_utf8 * icu_buf_utf8_resize(struct icu_buf_utf8 * buf8,
size_t capacity)
{
return buf8;
}
-
const char *icu_buf_utf8_to_cstr(struct icu_buf_utf8 *src8)
{
if (!src8 || src8->utf8_len == 0)
return (const char *) src8->utf8;
}
-
void icu_buf_utf8_destroy(struct icu_buf_utf8 * buf8)
{
if (buf8)
xfree(buf8);
}
-
-
-UErrorCode icu_utf16_from_utf8(struct icu_buf_utf16 * dest16,
- struct icu_buf_utf8 * src8,
- UErrorCode * status)
-{
- int32_t utf16_len = 0;
-
- u_strFromUTF8(dest16->utf16, dest16->utf16_cap,
- &utf16_len,
- (const char *) src8->utf8, src8->utf8_len, status);
-
- /* check for buffer overflow, resize and retry */
- if (*status == U_BUFFER_OVERFLOW_ERROR)
- {
- icu_buf_utf16_resize(dest16, utf16_len * 2);
- *status = U_ZERO_ERROR;
- u_strFromUTF8(dest16->utf16, dest16->utf16_cap,
- &utf16_len,
- (const char *) src8->utf8, src8->utf8_len, status);
- }
-
- if (U_SUCCESS(*status)
- && utf16_len <= dest16->utf16_cap)
- dest16->utf16_len = utf16_len;
- else
- icu_buf_utf16_clear(dest16);
-
- return *status;
-}
-
-
-
UErrorCode icu_utf16_from_utf8_cstr(struct icu_buf_utf16 * dest16,
const char * src8cstr,
UErrorCode * status)
return *status;
}
-
-
-
UErrorCode icu_utf16_to_utf8(struct icu_buf_utf8 * dest8,
struct icu_buf_utf16 * src16,
UErrorCode * status)
u_strToUTF8((char *) dest8->utf8, dest8->utf8_cap,
&utf8_len,
src16->utf16, src16->utf16_len, status);
-
}
if (U_SUCCESS(*status)
= (struct icu_casemap *) xmalloc(sizeof(struct icu_casemap));
casemap->action = action;
- switch(casemap->action) {
+ switch(casemap->action)
+ {
case 'l':
case 'L':
case 'u':
icu_casemap_destroy(casemap);
return 0;
}
-
return casemap;
}
xfree(casemap);
}
-
int icu_casemap_casemap(struct icu_casemap * casemap,
struct icu_buf_utf16 * dest16,
struct icu_buf_utf16 * src16,
casemap->action, status);
}
-
int icu_utf16_casemap(struct icu_buf_utf16 * dest16,
struct icu_buf_utf16 * src16,
const char *locale, char action,
{
int32_t dest16_len = 0;
-
- if (!src16->utf16_len){ /* guarding for empty source string */
+ if (!src16->utf16_len)
+ { /* guarding for empty source string */
if (dest16->utf16)
dest16->utf16[0] = (UChar) 0;
dest16->utf16_len = 0;
return U_ZERO_ERROR;
}
-
- switch(action) {
+ switch(action)
+ {
case 'l':
case 'L':
dest16_len = u_strToLower(dest16->utf16, dest16->utf16_cap,
/* check for buffer overflow, resize and retry */
if (*status == U_BUFFER_OVERFLOW_ERROR
&& dest16 != src16 /* do not resize if in-place conversion */
- ){
+ )
+ {
icu_buf_utf16_resize(dest16, dest16_len * 2);
*status = U_ZERO_ERROR;
-
switch(action) {
case 'l':
case 'L':
if (U_SUCCESS(*status)
&& dest16_len <= dest16->utf16_cap)
dest16->utf16_len = dest16_len;
- else {
+ else
+ {
if (dest16->utf16)
dest16->utf16[0] = (UChar) 0;
dest16->utf16_len = 0;
return *status;
}
-
-
void icu_sortkey8_from_utf16(UCollator *coll,
struct icu_buf_utf8 * dest8,
struct icu_buf_utf16 * src16,
UErrorCode * status)
{
-
int32_t sortkey_len = 0;
sortkey_len = ucol_getSortKey(coll, src16->utf16, src16->utf16_len,
dest8->utf8, dest8->utf8_cap);
/* check for buffer overflow, resize and retry */
- if (sortkey_len > dest8->utf8_cap) {
+ if (sortkey_len > dest8->utf8_cap)
+ {
icu_buf_utf8_resize(dest8, sortkey_len * 2);
sortkey_len = ucol_getSortKey(coll, src16->utf16, src16->utf16_len,
dest8->utf8, dest8->utf8_cap);
tokenizer->token_start = 0;
tokenizer->token_end = 0;
-
- switch(tokenizer->action) {
+ switch(tokenizer->action)
+ {
case 'l':
case 'L':
tokenizer->bi = ubrk_open(UBRK_LINE, locale, 0, 0, status);
if (!tokenizer || !tokenizer->bi || !src16)
return 0;
-
tokenizer->buf16 = src16;
tokenizer->token_count = 0;
tokenizer->token_id = 0;
tokenizer->token_end = 0;
ubrk_setText(tokenizer->bi, src16->utf16, src16->utf16_len, status);
-
-
+
if (U_FAILURE(*status))
return 0;
int32_t tkn_start = 0;
int32_t tkn_end = 0;
int32_t tkn_len = 0;
-
if (!tokenizer || !tokenizer->bi
|| !tokenizer->buf16 || !tokenizer->buf16->utf16_len)
return 0;
-
/*
never change tokenizer->buf16 and keep always invariant
0 <= tokenizer->token_start
tokenizer->token_id = 0;
}
tokenizer->token_start = tkn_start;
- tokenizer->token_end = tkn_end;
-
+ tokenizer->token_end = tkn_end;
/* copying into token buffer if it exists */
if (tkn16){
return tkn_len;
}
-
int32_t icu_tokenizer_token_id(struct icu_tokenizer * tokenizer)
{
return tokenizer->token_id;
return tokenizer->token_count;
}
+struct icu_transform * icu_transform_create(const char *id, char action,
+ const char *rules,
+ UErrorCode *status)
+{
+ struct icu_buf_utf16 *id16 = icu_buf_utf16_create(0);
+ struct icu_buf_utf16 *rules16 = icu_buf_utf16_create(0);
+ struct icu_transform * transform
+ = (struct icu_transform *) xmalloc(sizeof(struct icu_transform));
-struct icu_normalizer * icu_normalizer_create(const char *rules, char action,
- UErrorCode *status)
-{
+ transform->action = action;
+ transform->trans = 0;
- struct icu_normalizer * normalizer
- = (struct icu_normalizer *) xmalloc(sizeof(struct icu_normalizer));
+ if (id)
+ icu_utf16_from_utf8_cstr(id16, id, status);
+ if (rules)
+ icu_utf16_from_utf8_cstr(rules16, rules, status);
- normalizer->action = action;
- normalizer->trans = 0;
- normalizer->rules16 = icu_buf_utf16_create(0);
- icu_utf16_from_utf8_cstr(normalizer->rules16, rules, status);
-
- switch(normalizer->action) {
+ switch(transform->action)
+ {
case 'f':
case 'F':
- normalizer->trans
- = utrans_openU(normalizer->rules16->utf16,
- normalizer->rules16->utf16_len,
+ transform->trans
+ = utrans_openU(id16->utf16,
+ id16->utf16_len,
UTRANS_FORWARD,
- 0, 0,
- &normalizer->parse_error, status);
+ rules16->utf16,
+ rules16->utf16_len,
+ &transform->parse_error, status);
break;
case 'r':
case 'R':
- normalizer->trans
- = utrans_openU(normalizer->rules16->utf16,
- normalizer->rules16->utf16_len,
+ transform->trans
+ = utrans_openU(id16->utf16,
+ id16->utf16_len,
UTRANS_REVERSE ,
- 0, 0,
- &normalizer->parse_error, status);
+ rules16->utf16,
+ rules16->utf16_len,
+ &transform->parse_error, status);
break;
default:
*status = U_UNSUPPORTED_ERROR;
- return 0;
break;
}
+ icu_buf_utf16_destroy(rules16);
+ icu_buf_utf16_destroy(id16);
if (U_SUCCESS(*status))
- return normalizer;
+ return transform;
/* freeing if failed */
- icu_normalizer_destroy(normalizer);
+ icu_transform_destroy(transform);
return 0;
}
-
-void icu_normalizer_destroy(struct icu_normalizer * normalizer){
- if (normalizer) {
- if (normalizer->rules16)
- icu_buf_utf16_destroy(normalizer->rules16);
- if (normalizer->trans)
- utrans_close(normalizer->trans);
- xfree(normalizer);
+void icu_transform_destroy(struct icu_transform * transform)
+{
+ if (transform)
+ {
+ if (transform->trans)
+ utrans_close(transform->trans);
+ xfree(transform);
}
}
-
-
-int icu_normalizer_normalize(struct icu_normalizer * normalizer,
- struct icu_buf_utf16 * dest16,
- struct icu_buf_utf16 * src16,
- UErrorCode *status)
+int icu_transform_trans(struct icu_transform * transform,
+ struct icu_buf_utf16 * dest16,
+ struct icu_buf_utf16 * src16,
+ UErrorCode *status)
{
- if (!normalizer || !normalizer->trans
- || !src16
- || !dest16)
+ if (!transform || !transform->trans
+ || !src16 || !dest16)
return 0;
- if (!src16->utf16_len){ /* guarding for empty source string */
+ if (!src16->utf16_len)
+ { /* guarding for empty source string */
icu_buf_utf16_clear(dest16);
return 0;
}
if (!icu_buf_utf16_copy(dest16, src16))
return 0;
-
- utrans_transUChars (normalizer->trans,
+ utrans_transUChars (transform->trans,
dest16->utf16, &(dest16->utf16_len),
dest16->utf16_cap,
0, &(src16->utf16_len), status);
return dest16->utf16_len;
}
-
-
-
struct icu_chain_step * icu_chain_step_create(struct icu_chain * chain,
enum icu_chain_step_type type,
const uint8_t * rule,
step->buf16 = buf16;
/* create auxilary objects */
- switch(step->type) {
+ switch(step->type)
+ {
case ICU_chain_step_type_display:
break;
case ICU_chain_step_type_casemap:
step->u.casemap = icu_casemap_create(rule[0], status);
break;
- case ICU_chain_step_type_normalize:
- step->u.normalizer = icu_normalizer_create((char *) rule, 'f', status);
+ case ICU_chain_step_type_transform:
+ /* rule omitted. Only ID used */
+ step->u.transform = icu_transform_create((const char *) rule, 'f',
+ 0, status);
break;
case ICU_chain_step_type_tokenize:
step->u.tokenizer = icu_tokenizer_create((char *) chain->locale,
(char) rule[0], status);
break;
+ case ICU_chain_step_type_transliterate:
+ /* we pass a dummy ID to utrans_openU.. */
+ step->u.transform = icu_transform_create("custom", 'f',
+ (const char *) rule, status);
+ break;
default:
break;
}
-
return step;
}
-void icu_chain_step_destroy(struct icu_chain_step * step){
-
+void icu_chain_step_destroy(struct icu_chain_step * step)
+{
if (!step)
return;
icu_chain_step_destroy(step->previous);
- switch(step->type) {
+ switch(step->type)
+ {
case ICU_chain_step_type_display:
break;
case ICU_chain_step_type_casemap:
icu_casemap_destroy(step->u.casemap);
icu_buf_utf16_destroy(step->buf16);
break;
- case ICU_chain_step_type_normalize:
- icu_normalizer_destroy(step->u.normalizer);
+ case ICU_chain_step_type_transform:
+ case ICU_chain_step_type_transliterate:
+ icu_transform_destroy(step->u.transform);
icu_buf_utf16_destroy(step->buf16);
break;
case ICU_chain_step_type_tokenize:
}
-
struct icu_chain * icu_chain_create(const char *locale, int sort,
UErrorCode * status)
{
}
}
-
-
struct icu_chain * icu_chain_xml_config(const xmlNode *xml_node,
int sort,
UErrorCode * status)
step = icu_chain_insert_step(chain, ICU_chain_step_type_casemap,
(const uint8_t *) xml_rule, status);
else if (!strcmp((const char *) node->name, "transform"))
- step = icu_chain_insert_step(chain, ICU_chain_step_type_normalize,
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_transform,
+ (const uint8_t *) xml_rule, status);
+ else if (!strcmp((const char *) node->name, "transliterate"))
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_transliterate,
(const uint8_t *) xml_rule, status);
else if (!strcmp((const char *) node->name, "tokenize"))
step = icu_chain_insert_step(chain, ICU_chain_step_type_tokenize,
else if (!strcmp((const char *) node->name, "display"))
step = icu_chain_insert_step(chain, ICU_chain_step_type_display,
(const uint8_t *) "", status);
+ else if (!strcmp((const char *) node->name, "normalize"))
+ {
+ yaz_log(YLOG_WARN, "Element %s is deprecated. "
+ "Use transform instead", node->name);
+ step = icu_chain_insert_step(chain, ICU_chain_step_type_transform,
+ (const uint8_t *) xml_rule, status);
+ }
+ else if (!strcmp((const char *) node->name, "index")
+ || !strcmp((const char *) node->name, "sortkey"))
+ {
+ yaz_log(YLOG_WARN, "Element %s is no longer needed. "
+ "Remove it from the configuration", node->name);
+ }
+ else
+ {
+ yaz_log(YLOG_WARN, "Unknown element %s", node->name);
+ icu_chain_destroy(chain);
+ return 0;
+ }
xmlFree(xml_rule);
- if (!step || U_FAILURE(*status))
+ if (step && U_FAILURE(*status))
{
icu_chain_destroy(chain);
return 0;
}
-
-
}
return chain;
}
-
-
struct icu_chain_step * icu_chain_insert_step(struct icu_chain * chain,
enum icu_chain_step_type type,
const uint8_t * rule,
else
return 0;
-
/* create utf16 destination buffers as needed, or */
switch(type)
{
case ICU_chain_step_type_casemap:
buf16 = icu_buf_utf16_create(0);
break;
- case ICU_chain_step_type_normalize:
+ case ICU_chain_step_type_transform:
+ case ICU_chain_step_type_transliterate:
buf16 = icu_buf_utf16_create(0);
break;
case ICU_chain_step_type_tokenize:
buf16 = icu_buf_utf16_create(0);
break;
+ break;
default:
break;
}
step->buf16, src16, status,
chain->locale);
break;
- case ICU_chain_step_type_normalize:
- icu_normalizer_normalize(step->u.normalizer,
- step->buf16, src16, status);
+ case ICU_chain_step_type_transform:
+ case ICU_chain_step_type_transliterate:
+ icu_transform_trans(step->u.transform,
+ step->buf16, src16, status);
break;
case ICU_chain_step_type_tokenize:
/* attach to new src16 token only first time during splitting */
}
-int icu_chain_assign_cstr(struct icu_chain * chain,
- const char * src8cstr,
+int icu_chain_assign_cstr(struct icu_chain * chain, const char * src8cstr,
UErrorCode *status)
{
struct icu_chain_step * stp = 0;
return 1;
}
-
-
-int icu_chain_next_token(struct icu_chain * chain,
- UErrorCode *status)
+int icu_chain_next_token(struct icu_chain * chain, UErrorCode *status)
{
int got_token = 0;
icu_sortkey8_from_utf16(chain->coll,
chain->sort8, chain->steps->buf16,
status);
-
return chain->token_count;
}
}
return chain->token_count;
}
-
const char * icu_chain_token_display(struct icu_chain * chain)
{
if (chain->display8)
/*
* Local variables:
* c-basic-offset: 4
+ * c-file-style: "Stroustrup"
* indent-tabs-mode: nil
* End:
* vim: shiftwidth=4 tabstop=8 expandtab
*/
+
projects / yaz-moved-to-github.git / blobdiff