-/*
- * Copyright (C) 1995-2002, Index Data
- * All rights reserved.
- * Sebastian Hammer, Adam Dickmeiss
- *
- * $Id: zrpn.c,v 1.116 2002-07-03 10:05:19 adam Exp $
- */
+/* $Id: zrpn.c,v 1.135 2004-01-15 13:31:31 adam Exp $
+ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
+ Index Data Aps
+
+This file is part of the Zebra server.
+
+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
+Software Foundation; either version 2, or (at your option) any later
+version.
+
+Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
+WARRANTY; without even the implied warranty of MERCHANTABILITY or
+FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+for more details.
+
+You should have received a copy of the GNU General Public License
+along with Zebra; see the file LICENSE.zebra. If not, write to the
+Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
+02111-1307, USA.
+*/
+
+
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#include <ctype.h>
#include "index.h"
+#include <zebra_xpath.h>
#include <charmap.h>
#include <rstemp.h>
int reg_type;
};
+typedef struct {
+ int type;
+ int major;
+ int minor;
+ Z_AttributesPlusTerm *zapt;
+} AttrType;
+
+
static const char **rpn_char_map_handler (void *vp, const char **from, int len)
{
struct rpn_char_map_info *p = (struct rpn_char_map_info *) vp;
- return zebra_maps_input (p->zm, p->reg_type, from, len);
+ const char **out = zebra_maps_input (p->zm, p->reg_type, from, len);
+#if 0
+ if (out && *out)
+ {
+ const char *outp = *out;
+ yaz_log (LOG_LOG, "---");
+ while (*outp)
+ {
+ yaz_log (LOG_LOG, "%02X", *outp);
+ outp++;
+ }
+ }
+#endif
+ return out;
}
static void rpn_char_map_prepare (struct zebra_register *reg, int reg_type,
dict_grep_cmap (reg->dict, map_info, rpn_char_map_handler);
}
-typedef struct {
- int type;
- int major;
- int minor;
- Z_AttributesPlusTerm *zapt;
-} AttrType;
-
static int attr_find_ex (AttrType *src, oid_value *attributeSetP,
const char **string_value)
{
int num_attributes;
-#ifdef ASN_COMPILED
num_attributes = src->zapt->attributes->num_attributes;
-#else
- num_attributes = src->zapt->num_attributes;
-#endif
while (src->major < num_attributes)
{
Z_AttributeElement *element;
-#ifdef ASN_COMPILED
element = src->zapt->attributes->attributes[src->major];
-#else
- element = src->zapt->attributeList[src->major];
-#endif
if (src->type == *element->attributeType)
{
switch (element->which)
static void term_untrans (ZebraHandle zh, int reg_type,
char *dst, const char *src)
{
+ int len = 0;
while (*src)
{
const char *cp = zebra_maps_output (zh->reg->zebra_maps,
reg_type, &src);
- if (!cp)
- *dst++ = *src++;
+ if (!cp && len < IT_MAX_WORD-1)
+ dst[len++] = *src++;
else
- while (*cp)
- *dst++ = *cp++;
+ while (*cp && len < IT_MAX_WORD-1)
+ dst[len++] = *cp++;
}
- *dst = '\0';
+ dst[len] = '\0';
}
static void add_isam_p (const char *name, const char *info,
{
const char *db;
int set, use;
- char term_tmp[512];
+ char term_tmp[IT_MAX_WORD];
int su_code = 0;
int len = key_SU_decode (&su_code, name);
return *s0;
}
+#define REGEX_CHARS " []()|.*+?!"
+
/* term_100: handle term, where trunc=none (no operators at all) */
static int term_100 (ZebraMaps zebra_maps, int reg_type,
const char **src, char *dst, int space_split,
{ /* reload last space */
while (space_start < space_end)
{
- if (!isalnum (*space_start) && *space_start != '-')
+ if (strchr (REGEX_CHARS, *space_start))
dst[i++] = '\\';
dst_term[j++] = *space_start;
dst[i++] = *space_start++;
/* add non-space char */
while (s1 < s0)
{
- if (!isalnum (*s1) && *s1 != '-')
+ if (strchr(REGEX_CHARS, *s1))
dst[i++] = '\\';
dst_term[j++] = *s1;
dst[i++] = *s1++;
break;
while (s1 < s0)
{
- if (!isalnum (*s1))
+ if (strchr(REGEX_CHARS, *s1))
dst[i++] = '\\';
dst_term[j++] = *s1;
dst[i++] = *s1++;
break;
while (s1 < s0)
{
- if (!isalnum (*s1))
+ if (strchr(REGEX_CHARS, *s1))
dst[i++] = '\\';
dst_term[j++] = *s1;
dst[i++] = *s1++;
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "#!", "#!"))
+ if (!term_pre (zebra_maps, reg_type, src, "?*#", "?*#"))
return 0;
s0 = *src;
while (*s0)
{
- if (*s0 == '#')
+ if (*s0 == '?')
+ {
+ dst_term[j++] = *s0++;
+ if (*s0 >= '0' && *s0 <= '9')
+ {
+ int limit = 0;
+ while (*s0 >= '0' && *s0 <= '9')
+ {
+ limit = limit * 10 + (*s0 - '0');
+ dst_term[j++] = *s0++;
+ }
+ if (limit > 20)
+ limit = 20;
+ while (--limit >= 0)
+ {
+ dst[i++] = '.';
+ dst[i++] = '?';
+ }
+ }
+ else
+ {
+ dst[i++] = '.';
+ dst[i++] = '*';
+ }
+ }
+ else if (*s0 == '*')
{
dst[i++] = '.';
dst[i++] = '*';
dst_term[j++] = *s0++;
}
- else if (*s0 == '!')
+ else if (*s0 == '#')
{
dst[i++] = '.';
dst_term[j++] = *s0++;
break;
while (s1 < s0)
{
- if (!isalnum (*s1))
+ if (strchr(REGEX_CHARS, *s1))
dst[i++] = '\\';
dst_term[j++] = *s1;
dst[i++] = *s1++;
break;
while (s1 < s0)
{
- if (!isalnum (*s1))
+ if (strchr(REGEX_CHARS, *s1))
dst[i++] = '\\';
dst_term[j++] = *s1;
dst[i++] = *s1++;
int relation_value;
int i;
char *term_tmp = term_dict + strlen(term_dict);
- char term_component[256];
+ char term_component[2*IT_MAX_WORD+20];
attr_init (&relation, zapt, 2);
relation_value = attr_find (&relation, NULL);
*term_tmp++ = ']';
*term_tmp++ = '.';
*term_tmp++ = '*';
+
+ if ((term_tmp - term_dict) > IT_MAX_WORD)
+ break;
}
*term_tmp++ = ')';
*term_tmp = '\0';
*term_tmp++ = '*';
*term_tmp++ = '|';
+
+ if ((term_tmp - term_dict) > IT_MAX_WORD)
+ break;
}
for (i = 0; term_component[i]; )
string_rel_add_char (&term_tmp, term_component, &i);
*term_tmp++ = '*';
*term_tmp++ = '|';
+
+ if ((term_tmp - term_dict) > IT_MAX_WORD)
+ break;
}
for (i = 0; term_component[i];)
string_rel_add_char (&term_tmp, term_component, &i);
*term_tmp++ = ']';
*term_tmp++ = '.';
*term_tmp++ = '*';
+
+ if ((term_tmp - term_dict) > IT_MAX_WORD)
+ break;
}
*term_tmp++ = ')';
*term_tmp = '\0';
struct grep_info *grep_info,
int reg_type, int complete_flag,
int num_bases, char **basenames,
- char *term_dst);
+ char *term_dst, int xpath_use);
static RSET term_trunc (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
const char **term_sub,
int reg_type, int complete_flag,
int num_bases, char **basenames,
char *term_dst,
- const char *rank_type)
+ const char *rank_type, int xpath_use)
{
int r;
grep_info->isam_p_indx = 0;
r = string_term (zh, zapt, term_sub, attributeSet, stream, grep_info,
reg_type, complete_flag, num_bases, basenames,
- term_dst);
+ term_dst, xpath_use);
if (r < 1)
return 0;
logf (LOG_DEBUG, "term: %s", term_dst);
return rset_trunc (zh, grep_info->isam_p_buf,
grep_info->isam_p_indx, term_dst,
- strlen(term_dst), rank_type, 1 /* preserve pos */);
+ strlen(term_dst), rank_type, 1 /* preserve pos */,
+ zapt->term->which);
}
struct grep_info *grep_info,
int reg_type, int complete_flag,
int num_bases, char **basenames,
- char *term_dst)
+ char *term_dst, int xpath_use)
{
char term_dict[2*IT_MAX_WORD+4000];
int j, r, base_no;
struct rpn_char_map_info rcmi;
int space_split = complete_flag ? 0 : 1;
+ int bases_ok = 0; /* no of databases with OK attribute */
+ int errCode = 0; /* err code (if any is not OK) */
+ char *errString = 0; /* addinfo */
+
rpn_char_map_prepare (zh->reg, reg_type, &rcmi);
attr_init (&use, zapt, 1);
use_value = attr_find_ex (&use, &curAttributeSet, &use_string);
if (use_value == -1) /* no attribute - assumy "any" */
use_value = 1016;
- if (use_value == -2) /* string attribute - assumy "any" */
- use_value = 1016;
-
for (base_no = 0; base_no < num_bases; base_no++)
{
attent attp;
zh->errString = basenames[base_no];
return -1;
}
- if (curAttributeSet == VAL_IDXPATH)
+ if (use_value == -2) /* string attribute (assume IDXPATH/any) */
+ {
+ use_value = xpath_use;
+ attp.local_attributes = &id_xpath_attr;
+ attp.attset_ordinal = VAL_IDXPATH;
+ id_xpath_attr.next = 0;
+ id_xpath_attr.local = use_value;
+ }
+ else if (curAttributeSet == VAL_IDXPATH)
{
attp.local_attributes = &id_xpath_attr;
- attp.attset_ordinal = curAttributeSet;
+ attp.attset_ordinal = VAL_IDXPATH;
id_xpath_attr.next = 0;
id_xpath_attr.local = use_value;
}
/* set was found, but value wasn't defined */
char val_str[32];
sprintf (val_str, "%d", use_value);
- zh->errCode = 114;
- zh->errString = nmem_strdup (stream, val_str);
+ errCode = 114;
+ errString = nmem_strdup (stream, val_str);
}
else
{
oident.value = curAttributeSet;
oid_ent_to_oid (&oident, oid);
- zh->errCode = 121;
- zh->errString = nmem_strdup (stream, oident.desc);
+ errCode = 121;
+ errString = nmem_strdup (stream, oident.desc);
}
- return -1;
+ continue;
}
}
for (local_attr = attp.local_attributes; local_attr;
{
char val_str[32];
sprintf (val_str, "%d", use_value);
- zh->errCode = 114;
- zh->errString = nmem_strdup (stream, val_str);
- return -1;
+ errCode = 114;
+ errString = nmem_strdup (stream, val_str);
+ continue;
}
+ bases_ok++; /* this has OK attributes */
+
term_dict[prefix_len++] = ')';
term_dict[prefix_len++] = 1;
term_dict[prefix_len++] = reg_type;
attributeSet,
reg_type, space_split, term_dst))
return 0;
- logf (LOG_DEBUG, "dict_lookup_grep: %s", term_dict+prefix_len);
+ logf (LOG_LOG, "dict_lookup_grep: %s", term_dict+prefix_len);
r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
grep_info, &max_pos, 0, grep_handle);
if (r)
- logf (LOG_WARN, "dict_lookup_grep fail, rel=gt: %d", r);
+ logf (LOG_WARN, "dict_lookup_grep fail %d", r);
break;
case 1: /* right truncation */
term_dict[j++] = '(';
break;
}
}
+ if (!bases_ok)
+ {
+ zh->errCode = errCode;
+ zh->errString = errString;
+ return -1;
+ }
*term_sub = termp;
logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
return 1;
}
-static void trans_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- char *termz)
+
+/* convert APT search term to UTF8 */
+static int zapt_term_to_utf8 (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ char *termz)
{
size_t sizez;
Z_Term *term = zapt->term;
- sizez = term->u.general->len;
- if (sizez > IT_MAX_WORD-1)
- sizez = IT_MAX_WORD-1;
- memcpy (termz, term->u.general->buf, sizez);
- termz[sizez] = '\0';
+ switch (term->which)
+ {
+ case Z_Term_general:
+ if (zh->iconv_to_utf8 != 0)
+ {
+ char *inbuf = term->u.general->buf;
+ size_t inleft = term->u.general->len;
+ char *outbuf = termz;
+ size_t outleft = IT_MAX_WORD-1;
+ size_t ret;
+
+ ret = yaz_iconv(zh->iconv_to_utf8, &inbuf, &inleft,
+ &outbuf, &outleft);
+ if (ret == (size_t)(-1))
+ {
+ ret = yaz_iconv(zh->iconv_to_utf8, 0, 0, 0, 0);
+ zh->errCode = 125;
+ return -1;
+ }
+ *outbuf = 0;
+ }
+ else
+ {
+ sizez = term->u.general->len;
+ if (sizez > IT_MAX_WORD-1)
+ sizez = IT_MAX_WORD-1;
+ memcpy (termz, term->u.general->buf, sizez);
+ termz[sizez] = '\0';
+ }
+ break;
+ case Z_Term_characterString:
+ sizez = strlen(term->u.characterString);
+ if (sizez > IT_MAX_WORD-1)
+ sizez = IT_MAX_WORD-1;
+ memcpy (termz, term->u.characterString, sizez);
+ termz[sizez] = '\0';
+ break;
+ default:
+ zh->errCode = 124;
+ return -1;
+ }
+ return 0;
}
-static void trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- char *termz, int reg_type)
+/* convert APT SCAN term to internal cmap */
+static int trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ char *termz, int reg_type)
{
- Z_Term *term = zapt->term;
- const char **map;
- const char *cp = (const char *) term->u.general->buf;
- const char *cp_end = cp + term->u.general->len;
- const char *src;
- int i = 0;
- const char *space_map = NULL;
- int len;
-
- while ((len = (cp_end - cp)) > 0)
+ char termz0[IT_MAX_WORD];
+
+ if (zapt_term_to_utf8(zh, zapt, termz0))
+ return -1; /* error */
+ else
{
- map = zebra_maps_input (zh->reg->zebra_maps, reg_type, &cp, len);
- if (**map == *CHR_SPACE)
- space_map = *map;
- else
+ const char **map;
+ const char *cp = (const char *) termz0;
+ const char *cp_end = cp + strlen(cp);
+ const char *src;
+ int i = 0;
+ const char *space_map = NULL;
+ int len;
+
+ while ((len = (cp_end - cp)) > 0)
{
- if (i && space_map)
- for (src = space_map; *src; src++)
+ map = zebra_maps_input (zh->reg->zebra_maps, reg_type, &cp, len);
+ if (**map == *CHR_SPACE)
+ space_map = *map;
+ else
+ {
+ if (i && space_map)
+ for (src = space_map; *src; src++)
+ termz[i++] = *src;
+ space_map = NULL;
+ for (src = *map; *src; src++)
termz[i++] = *src;
- space_map = NULL;
- for (src = *map; *src; src++)
- termz[i++] = *src;
+ }
}
+ termz[i] = '\0';
}
- termz[i] = '\0';
+ return 0;
}
static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no,
int length_prox_term = 0;
int min_nn = 10000000;
int term_index;
+ int term_type = Z_Term_characterString;
const char *flags = NULL;
rsfd = (RSFD *) xmalloc (sizeof(*rsfd)*rset_no);
if (min_nn > rset[i]->rset_terms[j]->nn)
min_nn = rset[i]->rset_terms[j]->nn;
flags = nflags;
+ term_type = rset[i]->rset_terms[j]->type;
+
+ /* only if all term types are of type characterString .. */
+ /* the resulting term is of that type */
+ if (term_type != Z_Term_characterString)
+ term_type = Z_Term_general;
}
}
for (i = 0; i<rset_no; i++)
rset_null_parms parms;
parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags);
+ flags, term_type);
parms.rset_term->nn = 0;
result = rset_create (rset_kind_null, &parms);
}
RSFD rsfd_result;
parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags);
+ flags, term_type);
parms.rset_term->nn = min_nn;
parms.cmp = key_compare_it;
parms.key_size = sizeof (struct it_key);
parms.temp_path = res_get (zh->res, "setTmpDir");
result = rset_create (rset_kind_temp, &parms);
rsfd_result = rset_open (result, RSETF_WRITE);
-
+
while (*more)
{
for (i = 1; i<rset_no; i++)
rset_temp_parms parms;
RSFD rsfd_result;
- logf (LOG_LOG, "generic prox, dist = %d, relation = %d, ordered =%d, exclusion=%d",
- distance, relation, ordered, exclusion);
+ yaz_log (LOG_LOG, "generic prox, dist=%d, relation=%d, ordered=%d"
+ ", exclusion=%d",
+ distance, relation, ordered, exclusion);
parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags);
+ flags, term_type);
parms.rset_term->nn = min_nn;
parms.cmp = key_compare_it;
parms.key_size = sizeof (struct it_key);
rset_null_parms parms;
parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags);
+ flags, term_type);
parms.rset_term->nn = 0;
result = rset_create (rset_kind_null, &parms);
}
grep_info->reg_type = reg_type;
grep_info->termset = 0;
+ if (!zapt)
+ return 0;
attr_init (&termset, zapt, 8);
termset_value_numeric =
attr_find_ex (&termset, NULL, &termset_value_string);
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
- const char *rank_type,
+ const char *rank_type, int xpath_use,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
stream, &grep_info,
reg_type, complete_flag,
num_bases, basenames,
- term_dst, rank_type);
+ term_dst, rank_type,
+ xpath_use);
if (!rset[rset_no])
break;
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
{
rset_null_parms parms;
- parms.rset_term = rset_term_create (termz, -1, rank_type);
+ parms.rset_term = rset_term_create (termz, -1, rank_type,
+ zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
else if (rset_no == 1)
NMEM stream,
int reg_type, int complete_flag,
const char *rank_type,
+ int xpath_use,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
stream, &grep_info,
reg_type, complete_flag,
num_bases, basenames,
- term_dst, rank_type);
+ term_dst, rank_type,
+ xpath_use);
if (!rset[rset_no])
break;
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
{
rset_null_parms parms;
- parms.rset_term = rset_term_create (termz, -1, rank_type);
+ parms.rset_term = rset_term_create (termz, -1, rank_type,
+ zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
- const char *rank_type,
+ const char *rank_type,
+ int xpath_use,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
stream, &grep_info,
reg_type, complete_flag,
num_bases, basenames,
- term_dst, rank_type);
+ term_dst, rank_type,
+ xpath_use);
if (!rset[rset_no])
break;
assert (rset[rset_no]);
{
rset_null_parms parms;
- parms.rset_term = rset_term_create (termz, -1, rank_type);
+ parms.rset_term = rset_term_create (termz, -1, rank_type,
+ zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
oid_value attributeSet, struct grep_info *grep_info,
int reg_type, int complete_flag,
int num_bases, char **basenames,
- char *term_dst)
+ char *term_dst, int xpath_use, NMEM stream)
{
char term_dict[2*IT_MAX_WORD+2];
int r, base_no;
AttrType use;
int use_value;
+ const char *use_string = 0;
oid_value curAttributeSet = attributeSet;
const char *termp;
struct rpn_char_map_info rcmi;
+ int bases_ok = 0; /* no of databases with OK attribute */
+ int errCode = 0; /* err code (if any is not OK) */
+ char *errString = 0; /* addinfo */
+
rpn_char_map_prepare (zh->reg, reg_type, &rcmi);
attr_init (&use, zapt, 1);
- use_value = attr_find (&use, &curAttributeSet);
- logf (LOG_DEBUG, "numeric_term, use value %d", use_value);
+ use_value = attr_find_ex (&use, &curAttributeSet, &use_string);
if (use_value == -1)
use_value = 1016;
for (base_no = 0; base_no < num_bases; base_no++)
{
attent attp;
+ data1_local_attribute id_xpath_attr;
data1_local_attribute *local_attr;
int max_pos, prefix_len = 0;
termp = *term_sub;
- if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value)))
+ if (use_value == -2) /* string attribute (assume IDXPATH/any) */
{
- logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d",
- curAttributeSet, use_value, r);
- if (r == -1)
- zh->errCode = 114;
- else
- zh->errCode = 121;
- return -1;
+ use_value = xpath_use;
+ attp.local_attributes = &id_xpath_attr;
+ attp.attset_ordinal = VAL_IDXPATH;
+ id_xpath_attr.next = 0;
+ id_xpath_attr.local = use_value;
+ }
+ else if (curAttributeSet == VAL_IDXPATH)
+ {
+ attp.local_attributes = &id_xpath_attr;
+ attp.attset_ordinal = VAL_IDXPATH;
+ id_xpath_attr.next = 0;
+ id_xpath_attr.local = use_value;
+ }
+ else
+ {
+ if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value)))
+ {
+ logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d",
+ curAttributeSet, use_value, r);
+ if (r == -1)
+ {
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ errString = nmem_strdup (stream, val_str);
+ errCode = 114;
+ }
+ else
+ errCode = 121;
+ continue;
+ }
}
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
}
if (!prefix_len)
{
- zh->errCode = 114;
- return -1;
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ errCode = 114;
+ errString = nmem_strdup (stream, val_str);
+ continue;
}
+ bases_ok++;
term_dict[prefix_len++] = ')';
term_dict[prefix_len++] = 1;
term_dict[prefix_len++] = reg_type;
term_dst))
return 0;
}
+ if (!bases_ok)
+ {
+ zh->errCode = errCode;
+ zh->errString = errString;
+ return -1;
+ }
*term_sub = termp;
logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
return 1;
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
- const char *rank_type,
+ const char *rank_type, int xpath_use,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
grep_info.isam_p_indx = 0;
r = numeric_term (zh, zapt, &termp, attributeSet, &grep_info,
reg_type, complete_flag, num_bases, basenames,
- term_dst);
+ term_dst, xpath_use,
+ stream);
if (r < 1)
break;
logf (LOG_DEBUG, "term: %s", term_dst);
rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf,
grep_info.isam_p_indx, term_dst,
strlen(term_dst), rank_type,
- 0 /* preserve position */);
+ 0 /* preserve position */,
+ zapt->term->which);
assert (rset[rset_no]);
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
break;
{
rset_null_parms parms;
- parms.rset_term = rset_term_create (term_dst, -1, rank_type);
+ parms.rset_term = rset_term_create (term_dst, -1, rank_type,
+ zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
struct it_key key;
rset_temp_parms parms;
- parms.rset_term = rset_term_create (termz, -1, rank_type);
+ parms.rset_term = rset_term_create (termz, -1, rank_type,
+ zapt->term->which);
parms.cmp = key_compare_it;
parms.key_size = sizeof (struct it_key);
parms.temp_path = res_get (zh->res, "setTmpDir");
nmem_malloc (stream, sizeof(*sks->caseSensitivity));
*sks->caseSensitivity = 0;
-#ifdef ASN_COMPILED
sks->which = Z_SortKeySpec_null;
sks->u.null = odr_nullval ();
-#else
- sks->missingValueAction = 0;
-#endif
-
sort_sequence->specs[i] = sks;
- parms.rset_term = rset_term_create (termz, -1, rank_type);
+ parms.rset_term = rset_term_create (termz, -1, rank_type,
+ zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
-static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- oid_value attributeSet,
- int num_bases, char **basenames,
- NMEM stream, const char *rank_type, RSET rset)
+/* pop - moved to xpath.c */
+#if 0
+
+struct xpath_predicate {
+ int which;
+ union {
+#define XPATH_PREDICATE_RELATION 1
+ struct {
+ char *name;
+ char *op;
+ char *value;
+ } relation;
+#define XPATH_PREDICATE_BOOLEAN 2
+ struct {
+ const char *op;
+ struct xpath_predicate *left;
+ struct xpath_predicate *right;
+ } boolean;
+ } u;
+};
+
+struct xpath_location_step {
+ char *part;
+ struct xpath_predicate *predicate;
+};
+
+#endif
+
+static int parse_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ oid_value attributeSet,
+ struct xpath_location_step *xpath, int max, NMEM mem)
{
+ oid_value curAttributeSet = attributeSet;
AttrType use;
const char *use_string = 0;
- oid_value curAttributeSet = attributeSet;
- char term_dict[2048];
- int base_no;
- int reg_type = '0';
- struct grep_info grep_info;
-
- yaz_log (LOG_LOG, "rpn_search_xpath 1");
+
attr_init (&use, zapt, 1);
attr_find_ex (&use, &curAttributeSet, &use_string);
- if (curAttributeSet != VAL_IDXPATH)
+ if (!use_string || *use_string != '/')
+ return -1;
+
+ return zebra_parse_xpath_str(use_string, xpath, max, mem);
+}
+
+
+
+static RSET xpath_trunc(ZebraHandle zh, NMEM stream,
+ int reg_type, const char *term, int use,
+ oid_value curAttributeSet)
+{
+ RSET rset;
+ struct grep_info grep_info;
+ char term_dict[2048];
+ char ord_buf[32];
+ int prefix_len = 0;
+ int ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, use);
+ int ord_len, i, r, max_pos;
+ int term_type = Z_Term_characterString;
+ const char *flags = "void";
+
+ if (grep_info_prepare (zh, 0 /* zapt */, &grep_info, '0', stream))
{
- yaz_log (LOG_LOG, "rpn_search_xpath - not 1");
- return rset;
+ rset_null_parms parms;
+
+ parms.rset_term = rset_term_create (term, strlen(term),
+ flags, term_type);
+ parms.rset_term->nn = 0;
+ return rset_create (rset_kind_null, &parms);
+ }
+
+ if (ord < 0)
+ {
+ rset_null_parms parms;
+
+ parms.rset_term = rset_term_create (term, strlen(term),
+ flags, term_type);
+ parms.rset_term->nn = 0;
+ return rset_create (rset_kind_null, &parms);
}
- if (!use_string)
+ if (prefix_len)
+ term_dict[prefix_len++] = '|';
+ else
+ term_dict[prefix_len++] = '(';
+
+ ord_len = key_SU_encode (ord, ord_buf);
+ for (i = 0; i<ord_len; i++)
{
- yaz_log (LOG_LOG, "rpn_search_xpath - not 2");
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = ord_buf[i];
+ }
+ term_dict[prefix_len++] = ')';
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = reg_type;
+
+ strcpy (term_dict+prefix_len, term);
+
+ grep_info.isam_p_indx = 0;
+ r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
+ &grep_info, &max_pos, 0, grep_handle);
+ yaz_log (LOG_LOG, "%s %d positions", term,
+ grep_info.isam_p_indx);
+ rset = rset_trunc (zh, grep_info.isam_p_buf,
+ grep_info.isam_p_indx, term, strlen(term),
+ flags, 1, term_type);
+ grep_info_delete (&grep_info);
+ return rset;
+}
+
+static RSET rpn_search_xpath (ZebraHandle zh,
+ oid_value attributeSet,
+ int num_bases, char **basenames,
+ NMEM stream, const char *rank_type, RSET rset,
+ int xpath_len, struct xpath_location_step *xpath)
+{
+ oid_value curAttributeSet = attributeSet;
+ int base_no;
+ int i;
+
+ if (xpath_len < 0)
return rset;
+
+ yaz_log (LOG_LOG, "len=%d", xpath_len);
+ for (i = 0; i<xpath_len; i++)
+ {
+ yaz_log (LOG_LOG, "XPATH %d %s", i, xpath[i].part);
+
}
+ curAttributeSet = VAL_IDXPATH;
+
+ /*
+ //a -> a/.*
+ //a/b -> b/a/.*
+ /a -> a/
+ /a/b -> b/a/
+
+ / -> none
+
+ a[@attr=value]/b[@other=othervalue]
+
+ /e/@a val range(e/,range(@a,freetext(w,1015,val),@a),e/)
+ /a/b val range(b/a/,freetext(w,1016,val),b/a/)
+ /a/b/@c val range(b/a/,range(@c,freetext(w,1016,val),@c),b/a/)
+ /a/b[@c=y] val range(b/a/,freetext(w,1016,val),b/a/,@c=y)
+ /a[@c=y]/b val range(a/,range(b/a/,freetext(w,1016,val),b/a/),a/,@c=y)
+ /a[@c=x]/b[@c=y] range(a/,range(b/a/,freetext(w,1016,val),b/a/,@c=y),a/,@c=x)
+
+ */
+
dict_grep_cmap (zh->reg->dict, 0, 0);
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
- yaz_log (LOG_LOG, "rpn_search_xpath 2");
for (base_no = 0; base_no < num_bases; base_no++)
{
- const char *termp = use_string;
- rset_between_parms parms;
- RSET rset_start_tag, rset_end_tag;
- int ord, ord_len, i, r, max_pos;
- int prefix_len ;
- char ord_buf[32];
+ int level = xpath_len;
+ int first_path = 1;
+
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
zh->errCode = 109; /* Database unavailable */
zh->errString = basenames[base_no];
return rset;
}
-
- prefix_len = 0;
- ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, 1);
- if (ord < 0)
- continue;
- if (prefix_len)
- term_dict[prefix_len++] = '|';
- else
- term_dict[prefix_len++] = '(';
-
- ord_len = key_SU_encode (ord, ord_buf);
- for (i = 0; i<ord_len; i++)
+ while (--level >= 0)
{
- term_dict[prefix_len++] = 1;
- term_dict[prefix_len++] = ord_buf[i];
- }
- term_dict[prefix_len++] = ')';
- term_dict[prefix_len++] = 1;
- term_dict[prefix_len++] = reg_type;
+ char xpath_rev[128];
+ int i, len;
+ rset_between_parms parms;
+ RSET rset_start_tag = 0, rset_end_tag = 0, rset_attr = 0;
+
+ *xpath_rev = 0;
+ len = 0;
+ for (i = level; i >= 1; --i)
+ {
+ const char *cp = xpath[i].part;
+ if (*cp)
+ {
+ for (;*cp; cp++)
+ if (*cp == '*')
+ {
+ memcpy (xpath_rev + len, "[^/]*", 5);
+ len += 5;
+ }
+ else if (*cp == ' ')
+ {
+
+ xpath_rev[len++] = 1;
+ xpath_rev[len++] = ' ';
+ }
+
+ else
+ xpath_rev[len++] = *cp;
+ xpath_rev[len++] = '/';
+ }
+ else if (i == 1) /* // case */
+ {
+ xpath_rev[len++] = '.';
+ xpath_rev[len++] = '*';
+ }
+ }
+ xpath_rev[len] = 0;
- termp = use_string;
- strcpy (term_dict+prefix_len, use_string);
-
- grep_info.isam_p_indx = 0;
- yaz_log (LOG_LOG, "rpn_search_xpath 3 %s", term_dict+prefix_len);
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
- &grep_info, &max_pos, 0, grep_handle);
- yaz_log (LOG_LOG, "%s %d positions", use_string,
- grep_info.isam_p_indx);
- rset_start_tag =
- rset_trunc (zh, grep_info.isam_p_buf,
- grep_info.isam_p_indx, use_string, strlen(use_string),
- rank_type, 1);
-
- prefix_len = 0;
- ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, 2);
- if (ord < 0)
- continue;
- if (prefix_len)
- term_dict[prefix_len++] = '|';
- else
- term_dict[prefix_len++] = '(';
-
- ord_len = key_SU_encode (ord, ord_buf);
- for (i = 0; i<ord_len; i++)
- {
- term_dict[prefix_len++] = 1;
- term_dict[prefix_len++] = ord_buf[i];
+ if (xpath[level].predicate &&
+ xpath[level].predicate->which == XPATH_PREDICATE_RELATION &&
+ xpath[level].predicate->u.relation.name[0])
+ {
+ WRBUF wbuf = wrbuf_alloc();
+ wrbuf_puts(wbuf, xpath[level].predicate->u.relation.name+1);
+ if (xpath[level].predicate->u.relation.value)
+ {
+ const char *cp = xpath[level].predicate->u.relation.value;
+ wrbuf_putc(wbuf, '=');
+
+ while (*cp)
+ {
+ if (strchr(REGEX_CHARS, *cp))
+ wrbuf_putc(wbuf, '\\');
+ wrbuf_putc(wbuf, *cp);
+ cp++;
+ }
+ }
+ wrbuf_puts(wbuf, "");
+ rset_attr = xpath_trunc (
+ zh, stream, '0', wrbuf_buf(wbuf), 3, curAttributeSet);
+ wrbuf_free(wbuf, 1);
+ }
+ else
+ {
+ if (!first_path)
+ continue;
+ }
+ yaz_log (LOG_LOG, "xpath_rev (%d) = %s", level, xpath_rev);
+ if (strlen(xpath_rev))
+ {
+ rset_start_tag = xpath_trunc(zh, stream,
+ '0', xpath_rev, 1, curAttributeSet);
+
+ rset_end_tag = xpath_trunc(zh, stream,
+ '0', xpath_rev, 2, curAttributeSet);
+
+ parms.key_size = sizeof(struct it_key);
+ parms.cmp = key_compare_it;
+ parms.rset_l = rset_start_tag;
+ parms.rset_m = rset;
+ parms.rset_r = rset_end_tag;
+ parms.rset_attr = rset_attr;
+ parms.printer = key_print_it;
+ rset = rset_create (rset_kind_between, &parms);
+ }
+ first_path = 0;
}
- term_dict[prefix_len++] = ')';
- term_dict[prefix_len++] = 1;
- term_dict[prefix_len++] = reg_type;
-
- termp = use_string;
-
- strcpy (term_dict+prefix_len, use_string);
-
- grep_info.isam_p_indx = 0;
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
- &grep_info, &max_pos, 0, grep_handle);
-
- yaz_log (LOG_LOG, "%s %d positions", use_string,
- grep_info.isam_p_indx);
- rset_end_tag =
- rset_trunc (zh, grep_info.isam_p_buf,
- grep_info.isam_p_indx, use_string, strlen(use_string),
- rank_type, 1);
-
- parms.key_size = sizeof(struct it_key);
- parms.cmp = key_compare_it;
- parms.rset_l = rset_start_tag;
- parms.rset_m = rset;
- parms.rset_r = rset_end_tag;
- parms.printer = key_print_it;
- yaz_log (LOG_LOG, "rpn_search_xpath 4");
- rset = rset_create (rset_kind_between, &parms);
}
- grep_info_delete (&grep_info);
return rset;
}
int sort_flag;
char termz[IT_MAX_WORD+1];
RSET rset = 0;
+ int xpath_len;
+ int xpath_use = 0;
+ struct xpath_location_step xpath[10];
zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type,
rank_type, &complete_flag, &sort_flag);
logf (LOG_DEBUG, "search_type=%s", search_type);
logf (LOG_DEBUG, "rank_type=%s", rank_type);
- if (zapt->term->which != Z_Term_general)
- {
- zh->errCode = 124;
- return NULL;
- }
- trans_term (zh, zapt, termz);
+ if (zapt_term_to_utf8(zh, zapt, termz))
+ return 0;
if (sort_flag)
return rpn_sort_spec (zh, zapt, attributeSet, stream, sort_sequence,
rank_type);
+ xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream);
+ if (xpath_len >= 0)
+ {
+ xpath_use = 1016;
+ if (xpath[xpath_len-1].part[0] == '@')
+ xpath_use = 1015;
+ }
if (!strcmp (search_type, "phrase"))
{
rset = rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream,
reg_id, complete_flag, rank_type,
+ xpath_use,
num_bases, basenames);
}
else if (!strcmp (search_type, "and-list"))
{
rset = rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream,
reg_id, complete_flag, rank_type,
+ xpath_use,
num_bases, basenames);
}
else if (!strcmp (search_type, "or-list"))
{
rset = rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream,
reg_id, complete_flag, rank_type,
+ xpath_use,
num_bases, basenames);
}
else if (!strcmp (search_type, "local"))
{
rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream,
reg_id, complete_flag, rank_type,
+ xpath_use,
num_bases, basenames);
}
+ else if (!strcmp (search_type, "always"))
+ {
+ rset = 0;
+ }
else
zh->errCode = 118;
- return rpn_search_xpath (zh, zapt, attributeSet, num_bases, basenames,
- stream, rank_type, rset);
+ return rpn_search_xpath (zh, attributeSet, num_bases, basenames,
+ stream, rank_type, rset, xpath_len, xpath);
}
static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs,
r = rset_create (rset_kind_not, &bool_parms);
break;
case Z_Operator_prox:
-#ifdef ASN_COMPILED
if (zop->u.prox->which != Z_ProximityOperator_known)
{
zh->errCode = 132;
return NULL;
}
-#else
- if (zop->u.prox->which != Z_ProxCode_known)
- {
- zh->errCode = 132;
- return NULL;
- }
-#endif
-
-#ifdef ASN_COMPILED
if (*zop->u.prox->u.known != Z_ProxUnit_word)
{
char *val = (char *) nmem_malloc (stream, 16);
sprintf (val, "%d", *zop->u.prox->u.known);
return NULL;
}
-#else
- if (*zop->u.prox->proximityUnitCode != Z_ProxUnit_word)
- {
- char *val = (char *) nmem_malloc (stream, 16);
- zh->errCode = 132;
- zh->errString = val;
- sprintf (val, "%d", *zop->u.prox->proximityUnitCode);
- return NULL;
- }
-#endif
else
{
RSET rsets[2];
nmem_strdup (stream, zs->u.simple->u.resultSetId);
return 0;
}
+ else
+ rset_dup(r);
}
else
{
static void scan_term_untrans (ZebraHandle zh, NMEM stream, int reg_type,
char **dst, const char *src)
{
- char term_dst[1024];
+ char term_src[IT_MAX_WORD];
+ char term_dst[IT_MAX_WORD];
- term_untrans (zh, reg_type, term_dst, src);
-
- *dst = (char *) nmem_malloc (stream, strlen(term_dst)+1);
- strcpy (*dst, term_dst);
+ term_untrans (zh, reg_type, term_src, src);
+
+ if (zh->iconv_from_utf8 != 0)
+ {
+ int len;
+ char *inbuf = term_src;
+ size_t inleft = strlen(term_src);
+ char *outbuf = term_dst;
+ size_t outleft = sizeof(term_dst)-1;
+ size_t ret;
+
+ ret = yaz_iconv (zh->iconv_from_utf8, &inbuf, &inleft,
+ &outbuf, &outleft);
+ if (ret == (size_t)(-1))
+ len = 0;
+ else
+ len = outbuf - term_dst;
+ *dst = nmem_malloc (stream, len + 1);
+ if (len > 0)
+ memcpy (*dst, term_dst, len);
+ (*dst)[len] = '\0';
+ }
+ else
+ *dst = nmem_strdup (stream, term_src);
}
static void count_set (RSET r, int *count)
oid_value attributeset,
int num_bases, char **basenames,
int *position, int *num_entries, ZebraScanEntry **list,
- int *is_partial)
+ int *is_partial, RSET limit_set, int return_zero)
{
int i;
int pos = *position;
int ords[32], ord_no = 0;
int ptr[32];
+ int bases_ok = 0; /* no of databases with OK attribute */
+ int errCode = 0; /* err code (if any is not OK) */
+ char *errString = 0; /* addinfo */
+
unsigned reg_id;
char *search_type = NULL;
char rank_type[128];
int complete_flag;
int sort_flag;
+
*list = 0;
if (attributeset == VAL_NONE)
attributeset = VAL_BIB1;
+ if (!limit_set)
+ {
+ AttrType termset;
+ int termset_value_numeric;
+ const char *termset_value_string;
+ attr_init (&termset, zapt, 8);
+ termset_value_numeric =
+ attr_find_ex (&termset, NULL, &termset_value_string);
+ if (termset_value_numeric != -1)
+ {
+ char resname[32];
+ const char *termset_name = 0;
+
+ if (termset_value_numeric != -2)
+ {
+
+ sprintf (resname, "%d", termset_value_numeric);
+ termset_name = resname;
+ }
+ else
+ termset_name = termset_value_string;
+
+ limit_set = resultSetRef (zh, termset_name);
+ }
+ }
+
yaz_log (LOG_DEBUG, "position = %d, num = %d set=%d",
pos, num, attributeset);
logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d",
attributeset, use_value);
if (r == -1)
- zh->errCode = 114;
+ {
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ errCode = 114;
+ errString = odr_strdup (stream, val_str);
+ }
else
- zh->errCode = 121;
- *num_entries = 0;
- return;
+ errCode = 121;
+ continue;
}
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
*num_entries = 0;
return;
}
+ bases_ok++;
for (local_attr = attp.local_attributes; local_attr && ord_no < 32;
local_attr = local_attr->next)
{
ords[ord_no++] = ord;
}
}
+ if (!bases_ok && errCode)
+ {
+ zh->errCode = errCode;
+ zh->errString = errString;
+ *num_entries = 0;
+ }
if (ord_no == 0)
{
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ zh->errCode = 114;
+ zh->errString = odr_strdup (stream, val_str);
+
*num_entries = 0;
- zh->errCode = 113;
return;
}
/* prepare dictionary scanning */
termz[prefix_len] = 0;
strcpy (scan_info->prefix, termz);
- trans_scan_term (zh, zapt, termz+prefix_len, reg_id);
+ if (trans_scan_term (zh, zapt, termz+prefix_len, reg_id))
+ return ;
dict_scan (zh->reg->dict, termz, &before_tmp, &after_tmp,
scan_info, scan_handle);
&glist[i+before].term, mterm);
rset = rset_trunc (zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
glist[i+before].term, strlen(glist[i+before].term),
- NULL, 0);
+ NULL, 0, zapt->term->which);
ptr[j0]++;
for (j = j0+1; j<ord_no; j++)
rset2 =
rset_trunc (zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
glist[i+before].term,
- strlen(glist[i+before].term), NULL, 0);
+ strlen(glist[i+before].term), NULL, 0,
+ zapt->term->which);
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
ptr[j]++;
}
}
+ if (limit_set)
+ {
+ rset_bool_parms bool_parms;
+
+ bool_parms.key_size = sizeof(struct it_key);
+ bool_parms.cmp = key_compare_it;
+ bool_parms.rset_l = rset;
+ bool_parms.rset_r = rset_dup(limit_set);
+
+ rset = rset_create (rset_kind_and, &bool_parms);
+ }
count_set (rset, &glist[i+before].occurrences);
rset_delete (rset);
}
rset = rset_trunc
(zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1,
glist[before-1-i].term, strlen(glist[before-1-i].term),
- NULL, 0);
+ NULL, 0, zapt->term->which);
ptr[j0]++;
rset2 = rset_trunc (zh,
&scan_info_array[j].list[before-1-ptr[j]].isam_p, 1,
glist[before-1-i].term,
- strlen(glist[before-1-i].term), NULL, 0);
+ strlen(glist[before-1-i].term), NULL, 0,
+ zapt->term->which);
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
ptr[j]++;
}
}
+ if (limit_set)
+ {
+ rset_bool_parms bool_parms;
+
+ bool_parms.key_size = sizeof(struct it_key);
+ bool_parms.cmp = key_compare_it;
+ bool_parms.rset_l = rset;
+ bool_parms.rset_r = rset_dup(limit_set);
+
+ rset = rset_create (rset_kind_and, &bool_parms);
+ }
count_set (rset, &glist[before-1-i].occurrences);
rset_delete (rset);
}
projects / idzebra-moved-to-github.git / blobdiff