/* This file is part of the Zebra server.
- Copyright (C) 1994-2010 Index Data
+ Copyright (C) 1994-2011 Index Data
Zebra 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
*/
+#if HAVE_CONFIG_H
+#include <config.h>
+#endif
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
}
static int term_pre(zebra_map_t zm, const char **src,
- const char *ct1, const char *ct2, int first)
+ const char *ct1, int first)
{
const char *s1, *s0 = *src;
const char **map;
{
if (ct1 && strchr(ct1, *s0))
break;
- if (ct2 && strchr(ct2, *s0))
- break;
s1 = s0;
map = zebra_maps_input(zm, &s1, strlen(s1), first);
if (**map != *CHR_SPACE)
}
+static int term_102_icu(zebra_map_t zm,
+ const char **src, WRBUF term_dict, int space_split,
+ WRBUF display_term)
+{
+ int no_terms = 0;
+ const char *s0 = *src, *s1;
+ while (*s0 == ' ')
+ s0++;
+ s1 = s0;
+ for (;;)
+ {
+ if (*s1 == ' ' && space_split)
+ break;
+ else if (*s1 && !strchr(REGEX_CHARS "-", *s1))
+ s1++;
+ else
+ {
+ /* EOF or regex reserved char */
+ if (s0 != s1)
+ {
+ const char *res_buf = 0;
+ size_t res_len = 0;
+ const char *display_buf;
+ size_t display_len;
+
+ zebra_map_tokenize_start(zm, s0, s1 - s0);
+
+ if (zebra_map_tokenize_next(zm, &res_buf, &res_len,
+ &display_buf, &display_len))
+ {
+ size_t i = res_len;
+ while (--i >= 0 && res_buf[i] != '\x01')
+ ;
+ if (i > 0)
+ {
+ while (--i >= 0 && res_buf[i] != '\x01')
+ ;
+ }
+ res_len = i; /* reduce res_len */
+ for (i = 0; i < res_len; i++)
+ {
+ if (strchr(REGEX_CHARS "\\", res_buf[i]))
+ wrbuf_putc(term_dict, '\\');
+ if (res_buf[i] < 32)
+ wrbuf_putc(term_dict, '\x01');
+
+ wrbuf_putc(term_dict, res_buf[i]);
+ }
+ wrbuf_write(display_term, display_buf, display_len);
+
+ no_terms++;
+ }
+ }
+ if (*s1 == '\0')
+ break;
+
+ wrbuf_putc(term_dict, *s1);
+ wrbuf_putc(display_term, *s1);
+
+ s1++;
+ s0 = s1;
+ }
+ }
+ if (no_terms)
+ wrbuf_puts(term_dict, "\x01\x01.*");
+ *src = s1;
+ return no_terms;
+}
+
static int term_100_icu(zebra_map_t zm,
const char **src, WRBUF term_dict, int space_split,
WRBUF display_term,
- int right_trunc)
+ int mode)
{
- int i;
+ size_t i;
const char *res_buf = 0;
size_t res_len = 0;
const char *display_buf;
size_t display_len;
+ const char *s0 = *src, *s1;
+
+ while (*s0 == ' ')
+ s0++;
+
+ if (*s0 == '\0')
+ return 0;
+
+ if (space_split)
+ {
+ s1 = s0;
+ while (*s1 && *s1 != ' ')
+ s1++;
+ }
+ else
+ s1 = s0 + strlen(s0);
+
+ *src = s1;
+
+ zebra_map_tokenize_start(zm, s0, s1 - s0);
+
if (!zebra_map_tokenize_next(zm, &res_buf, &res_len,
&display_buf, &display_len))
{
- *src += strlen(*src);
return 0;
}
wrbuf_write(display_term, display_buf, display_len);
- if (right_trunc)
+ if (mode)
{
/* ICU sort keys seem to be of the form
basechars \x01 accents \x01 length
}
res_len = i; /* reduce res_len */
}
+ if (mode & 2)
+ wrbuf_puts(term_dict, ".*");
for (i = 0; i < res_len; i++)
{
if (strchr(REGEX_CHARS "\\", res_buf[i]))
wrbuf_putc(term_dict, '\\');
if (res_buf[i] < 32)
- wrbuf_putc(term_dict, 1);
-
+ wrbuf_putc(term_dict, '\x01');
+
wrbuf_putc(term_dict, res_buf[i]);
}
- if (right_trunc)
+ if (mode & 1)
wrbuf_puts(term_dict, ".*");
+ else if (mode)
+ wrbuf_puts(term_dict, "\x01\x01.*");
return 1;
}
const char *space_start = 0;
const char *space_end = 0;
- if (!term_pre(zm, src, NULL, NULL, !space_split))
+ if (!term_pre(zm, src, 0, !space_split))
return 0;
s0 = *src;
while (*s0)
const char **map;
int i = 0;
- if (!term_pre(zm, src, "#", "#", !space_split))
+ if (!term_pre(zm, src, "#", !space_split))
return 0;
s0 = *src;
while (*s0)
const char *s0;
const char **map;
- if (!term_pre(zm, src, "^\\()[].*+?|", "(", !space_split))
+ if (!term_pre(zm, src, "^\\()[].*+?|", !space_split))
return 0;
s0 = *src;
if (errors && *s0 == '+' && s0[1] && s0[2] == '+' && s0[3] &&
}
-/* term_104: handle term, process # and ! */
+/* term_104: handle term, process ?n * # */
static int term_104(zebra_map_t zm, const char **src,
WRBUF term_dict, int space_split, WRBUF display_term)
{
const char **map;
int i = 0;
- if (!term_pre(zm, src, "?*#", "?*#", !space_split))
+ if (!term_pre(zm, src, "?*#", !space_split))
return 0;
s0 = *src;
while (*s0)
return i;
}
-/* term_105/106: handle term, where trunc = Process * and ! and right trunc */
+/* term_105/106: handle term, process * ! and possibly right_truncate */
static int term_105(zebra_map_t zm, const char **src,
WRBUF term_dict, int space_split,
WRBUF display_term, int right_truncate)
const char **map;
int i = 0;
- if (!term_pre(zm, src, "*!", "*!", !space_split))
+ if (!term_pre(zm, src, "\\*!", !space_split))
return 0;
s0 = *src;
while (*s0)
wrbuf_putc(display_term, *s0);
s0++;
}
+ else if (*s0 == '\\')
+ {
+ i++;
+ wrbuf_puts(term_dict, "\\\\");
+ wrbuf_putc(display_term, *s0);
+ s0++;
+ }
else
{
const char *s1 = s0;
return ZEBRA_OK;
}
break;
+ case 102:
+ if (!term_102_icu(zm, &termp, term_dict, space_split, display_term))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ break;
case 1: /* right truncation */
if (!term_100_icu(zm, &termp, term_dict, space_split, display_term, 1))
{
return ZEBRA_OK;
}
break;
+ case 2:
+ if (!term_100_icu(zm, &termp, term_dict, space_split, display_term, 2))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ break;
+ case 3:
+ if (!term_100_icu(zm, &termp, term_dict, space_split, display_term, 3))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ break;
default:
zebra_setError_zint(zh,
YAZ_BIB1_UNSUPP_TRUNCATION_ATTRIBUTE,
}
wrbuf_putc(term_dict, ')');
break;
- case 104: /* process # and ! in term */
+ case 104: /* process ?n * # term */
wrbuf_putc(term_dict, '(');
if (!term_104(zm, &termp, term_dict, space_split, display_term))
{
}
wrbuf_putc(term_dict, ')');
break;
- case 105: /* process * and ! in term */
+ case 105: /* process * ! in term and right truncate */
wrbuf_putc(term_dict, '(');
if (!term_105(zm, &termp, term_dict, space_split, display_term, 1))
{
}
wrbuf_putc(term_dict, ')');
break;
- case 106: /* process * and ! in term */
+ case 106: /* process * ! in term */
wrbuf_putc(term_dict, '(');
if (!term_105(zm, &termp, term_dict, space_split, display_term, 0))
{
struct rset_key_control *kc)
{
zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type);
- if (zebra_maps_is_icu(zm))
- zebra_map_tokenize_start(zm, termz, strlen(termz));
return search_terms_chrmap(zh, zapt, termz, attributeSet, hits_limit,
stream, index_type, complete_flag,
rank_type, xpath_use,
Z_AttributesPlusTerm *zapt,
const char *termz,
const Odr_oid *attributeSet,
+ zint hits_limit,
NMEM stream,
const char *index_type,
int complete_flag,
ZEBRA_RES res;
struct grep_info grep_info;
int alloc_sets = 0;
- zint hits_limit_value;
+ zint hits_limit_value = hits_limit;
const char *term_ref_id_str = 0;
zebra_term_limits_APT(zh, zapt, &hits_limit_value, &term_ref_id_str,
sk->u.sortAttributes->id = odr_oiddup_nmem(stream, attributeSet);
sk->u.sortAttributes->list = zapt->attributes;
- sks->sortRelation = (int *)
+ sks->sortRelation = (Odr_int *)
nmem_malloc(stream, sizeof(*sks->sortRelation));
if (sort_relation_value == 1)
*sks->sortRelation = Z_SortKeySpec_ascending;
else
*sks->sortRelation = Z_SortKeySpec_ascending;
- sks->caseSensitivity = (int *)
+ sks->caseSensitivity = (Odr_int *)
nmem_malloc(stream, sizeof(*sks->caseSensitivity));
*sks->caseSensitivity = 0;
return rset_create_null(rset_nmem, kc, 0);
else
{
- int i, r, max_pos;
+ int i, max_pos;
char ord_buf[32];
RSET rset;
WRBUF term_dict = wrbuf_alloc();
wrbuf_puts(term_dict, term);
grep_info.isam_p_indx = 0;
- r = dict_lookup_grep(zh->reg->dict, wrbuf_cstr(term_dict), 0,
- &grep_info, &max_pos, 0, grep_handle);
+ dict_lookup_grep(zh->reg->dict, wrbuf_cstr(term_dict), 0,
+ &grep_info, &max_pos, 0, grep_handle);
yaz_log(YLOG_DEBUG, "%s %d positions", term,
grep_info.isam_p_indx);
rset = rset_trunc(zh, grep_info.isam_p_buf,
}
else if (!strcmp(search_type, "numeric"))
{
- res = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, stream,
+ res = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, hits_limit,
+ stream,
index_type, complete_flag, rank_type,
xpath_use,
rset_nmem,
projects / idzebra-moved-to-github.git / blobdiff