X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=src%2Fcharsets.c;h=dfc1015993737e56001ff525c356fdfe3e4cb5de;hb=5ac0f94cc7fc45e3ae81e8575facb88aa1740a98;hp=9332aa0c0a42996ac2c24cb6618567defcea7c2b;hpb=2f29d901e7ae8091b92f228a6aa5212d5fe2a01a;p=pazpar2-moved-to-github.git diff --git a/src/charsets.c b/src/charsets.c index 9332aa0..dfc1015 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) 2006-2010 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 @@ -28,21 +28,28 @@ 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" #if YAZ_HAVE_ICU #include + +#if YAZ_VERSIONL >= 0x40002 +/* YAZ 4.0.2 or later as icu_iter */ +#define ICU_ITER 1 #endif +#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; @@ -51,11 +58,11 @@ struct pp2_charset_s { }; 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 */ @@ -65,6 +72,9 @@ 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 ICU_ITER + yaz_icu_iter_t iter; +#endif }; @@ -142,12 +152,30 @@ void pp2_charset_destroy(pp2_charset_t pct) } pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct, - const char *buf) + const char *buf, + int skip_article) { pp2_relevance_token_t prt = xmalloc(sizeof(*prt)); assert(pct); + if (skip_article) + { + 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; + } + prt->norm_str = wrbuf_alloc(); prt->sort_str = wrbuf_alloc(); prt->cp = buf; @@ -155,11 +183,20 @@ pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct, prt->pct = pct; #if YAZ_HAVE_ICU +#if ICU_ITER + prt->iter = 0; +#endif if (pct->icu_chn) { +#if ICU_ITER + prt->iter = icu_iter_create(pct->icu_chn); + icu_iter_first(prt->iter, buf); +#else int ok = 0; pct->icu_sts = U_ZERO_ERROR; + ok = icu_chain_assign_cstr(pct->icu_chn, buf, &pct->icu_sts); +#endif //printf("\nfield ok: %d '%s'\n", ok, buf); prt->pct = pct; } @@ -171,6 +208,10 @@ pp2_relevance_token_t pp2_relevance_tokenize(pp2_charset_t pct, void pp2_relevance_token_destroy(pp2_relevance_token_t prt) { assert(prt); +#if ICU_ITER + if (prt->iter) + icu_iter_destroy(prt->iter); +#endif if(prt->norm_str) wrbuf_destroy(prt->norm_str); if(prt->sort_str) @@ -184,9 +225,9 @@ 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) @@ -220,8 +261,7 @@ 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; @@ -229,7 +269,7 @@ 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); @@ -242,6 +282,12 @@ 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) { +#if ICU_ITER + if (icu_iter_next(prt->iter)) + { + return icu_iter_get_norm(prt->iter); + } +#else if (icu_chain_next_token(prt->pct->icu_chn, &prt->pct->icu_sts)) { if (U_FAILURE(prt->pct->icu_sts)) @@ -250,13 +296,17 @@ static const char *pp2_relevance_token_icu(pp2_relevance_token_t prt) } return icu_chain_token_norm(prt->pct->icu_chn); } +#endif 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) { +#if ICU_ITER + return icu_iter_get_sortkey(prt->iter); +#else return icu_chain_token_sortkey(prt->pct->icu_chn); +#endif } #endif // YAZ_HAVE_ICU