/* 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
#include <yaz/xmalloc.h>
#include <yaz/wrbuf.h>
#include <yaz/log.h>
+#include <yaz/yaz-version.h>
#include <ctype.h>
#include <assert.h>
+#include <string.h>
#include "charsets.h"
#include "normalize7bit.h"
#include <yaz/icu.h>
#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;
};
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 */
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
};
}
}
-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));
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)
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)
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;
{
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);
#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