X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;ds=sidebyside;f=src%2Fcharsets.c;h=7bbe102e858788409e5f6d1b35ec70beb6d5145c;hb=01f19b8c3664fa1c78d1fb9efb8e358d50f85c39;hp=39ede26ab299d228b2df65e2f90ae242d8f4c9bf;hpb=9267d02c3601122f74bdf3a9d21f30080a4a1a9a;p=pazpar2-moved-to-github.git diff --git a/src/charsets.c b/src/charsets.c index 39ede26..7bbe102 100644 --- a/src/charsets.c +++ b/src/charsets.c @@ -1,5 +1,5 @@ /* This file is part of Pazpar2. - Copyright (C) 2006-2009 Index Data + Copyright (C) 2006-2011 Index Data Pazpar2 is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free @@ -28,8 +28,10 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA #include #include #include +#include #include #include +#include #include "charsets.h" #include "normalize7bit.h" @@ -38,23 +40,24 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA #include #endif - /* charset handle */ struct pp2_charset_s { const char *(*token_next_handler)(pp2_relevance_token_t prt); - const char *(*get_sort_handler)(pp2_relevance_token_t prt, int skip); + const char *(*get_sort_handler)(pp2_relevance_token_t prt); + int ref_count; #if YAZ_HAVE_ICU struct icu_chain * icu_chn; UErrorCode icu_sts; #endif }; +static const char *pp2_relevance_token_null(pp2_relevance_token_t prt); static const char *pp2_relevance_token_a_to_z(pp2_relevance_token_t prt); -static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt, int skip_article); +static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt); #if YAZ_HAVE_ICU static const char *pp2_relevance_token_icu(pp2_relevance_token_t prt); -static const char *pp2_get_sort_icu(pp2_relevance_token_t prt, int skip_article); +static const char *pp2_get_sort_icu(pp2_relevance_token_t prt); #endif /* tokenzier handle */ @@ -64,6 +67,9 @@ struct pp2_relevance_token_s { pp2_charset_t pct; /* our main charset handle (type+config) */ WRBUF norm_str; /* normized string we return (temporarily) */ WRBUF sort_str; /* sort string we return (temporarily) */ +#if YAZ_HAVE_ICU + yaz_icu_iter_t iter; +#endif }; @@ -99,13 +105,25 @@ pp2_charset_t pp2_charset_create_xml(xmlNode *xml_node) #endif // YAZ_HAVE_ICU } +void pp2_charset_incref(pp2_charset_t pct) +{ + (pct->ref_count)++; +} -pp2_charset_t pp2_charset_create(struct icu_chain * icu_chn) +pp2_charset_t pp2_charset_create_a_to_z(void) +{ + pp2_charset_t pct = pp2_charset_create(0); + pct->token_next_handler = pp2_relevance_token_a_to_z; + return pct; +} + +pp2_charset_t pp2_charset_create(struct icu_chain *icu_chn) { pp2_charset_t pct = xmalloc(sizeof(*pct)); - pct->token_next_handler = pp2_relevance_token_a_to_z; + pct->token_next_handler = pp2_relevance_token_null; pct->get_sort_handler = pp2_get_sort_ascii; + pct->ref_count = 1; #if YAZ_HAVE_ICU pct->icu_chn = 0; if (icu_chn) @@ -121,11 +139,21 @@ pp2_charset_t pp2_charset_create(struct icu_chain * icu_chn) void pp2_charset_destroy(pp2_charset_t pct) { - xfree(pct); + if (pct) + { + assert(pct->ref_count >= 1); + --(pct->ref_count); + if (pct->ref_count == 0) + { +#if YAZ_HAVE_ICU + icu_chain_destroy(pct->icu_chn); +#endif + xfree(pct); + } + } } -pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct, - const char *buf) +pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct) { pp2_relevance_token_t prt = xmalloc(sizeof(*prt)); @@ -133,27 +161,59 @@ pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct, prt->norm_str = wrbuf_alloc(); prt->sort_str = wrbuf_alloc(); - prt->cp = buf; + prt->cp = 0; prt->last_cp = 0; prt->pct = pct; #if YAZ_HAVE_ICU + prt->iter = 0; if (pct->icu_chn) + prt->iter = icu_iter_create(pct->icu_chn); +#endif + return prt; +} + +void pp2_relevance_first(pp2_relevance_token_t prt, + const char *buf, + int skip_article) +{ + if (skip_article) { - int ok = 0; - pct->icu_sts = U_ZERO_ERROR; - ok = icu_chain_assign_cstr(pct->icu_chn, buf, &pct->icu_sts); - //printf("\nfield ok: %d '%s'\n", ok, buf); - prt->pct = pct; + const char *p = buf; + char firstword[64]; + char *pout = firstword; + char articles[] = "the den der die des an a "; // must end in space + + while (*p && !isalnum(*(unsigned char *)p)) + p++; + for (; *p && *p != ' ' && pout - firstword < (sizeof(firstword)-2); p++) + *pout++ = tolower(*(unsigned char *)p); + *pout++ = ' '; + *pout++ = '\0'; + if (strstr(articles, firstword)) + buf = p; + } + + wrbuf_rewind(prt->norm_str); + wrbuf_rewind(prt->sort_str); + prt->cp = buf; + prt->last_cp = 0; + +#if YAZ_HAVE_ICU + if (prt->iter) + { + icu_iter_first(prt->iter, buf); } #endif // YAZ_HAVE_ICU - return prt; } - void pp2_relevance_token_destroy(pp2_relevance_token_t prt) { assert(prt); +#if YAZ_HAVE_ICU + if (prt->iter) + icu_iter_destroy(prt->iter); +#endif if(prt->norm_str) wrbuf_destroy(prt->norm_str); if(prt->sort_str) @@ -167,9 +227,9 @@ const char *pp2_relevance_token_next(pp2_relevance_token_t prt) return (prt->pct->token_next_handler)(prt); } -const char *pp2_get_sort(pp2_relevance_token_t prt, int skip) +const char *pp2_get_sort(pp2_relevance_token_t prt) { - return prt->pct->get_sort_handler(prt, skip); + return prt->pct->get_sort_handler(prt); } #define raw_char(c) (((c) >= 'a' && (c) <= 'z') ? (c) : -1) @@ -203,8 +263,7 @@ static const char *pp2_relevance_token_a_to_z(pp2_relevance_token_t prt) return wrbuf_cstr(prt->norm_str); } -static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt, - int skip_article) +static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt) { if (prt->last_cp == 0) return 0; @@ -212,7 +271,7 @@ static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt, { char *tmp = xstrdup(prt->last_cp); char *result = 0; - result = normalize7bit_mergekey(tmp, skip_article); + result = normalize7bit_mergekey(tmp); wrbuf_rewind(prt->sort_str); wrbuf_puts(prt->sort_str, result); @@ -221,25 +280,30 @@ static const char *pp2_get_sort_ascii(pp2_relevance_token_t prt, } } +static const char *pp2_relevance_token_null(pp2_relevance_token_t prt) +{ + const char *cp = prt->cp; + + prt->last_cp = *cp ? cp : 0; + while (*cp) + cp++; + prt->cp = cp; + return prt->last_cp; +} #if YAZ_HAVE_ICU static const char *pp2_relevance_token_icu(pp2_relevance_token_t prt) { - if (icu_chain_next_token(prt->pct->icu_chn, &prt->pct->icu_sts)) + if (icu_iter_next(prt->iter)) { - if (U_FAILURE(prt->pct->icu_sts)) - { - return 0; - } - return icu_chain_token_norm(prt->pct->icu_chn); + return icu_iter_get_norm(prt->iter); } return 0; } -static const char *pp2_get_sort_icu(pp2_relevance_token_t prt, - int skip_article) +static const char *pp2_get_sort_icu(pp2_relevance_token_t prt) { - return icu_chain_token_sortkey(prt->pct->icu_chn); + return icu_iter_get_sortkey(prt->iter); } #endif // YAZ_HAVE_ICU