-/*
- * Copyright (C) 1995-2002, Index Data
- * All rights reserved.
- * Sebastian Hammer, Adam Dickmeiss
- *
- * $Id: zrpn.c,v 1.112 2002-04-04 14:14:13 adam Exp $
- */
+/* $Id: zrpn.c,v 1.121 2002-08-23 14:30:51 adam Exp $
+ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
+ 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 <rstemp.h>
#include <rsnull.h>
#include <rsbool.h>
+#include <rsbetween.h>
struct rpn_char_map_info {
ZebraMaps zm;
case Z_AttributeValue_complex:
if (src->minor >= element->value.complex->num_list)
break;
+ if (element->attributeSet && attributeSetP)
+ {
+ oident *attrset;
+
+ attrset = oid_getentbyoid (element->attributeSet);
+ *attributeSetP = attrset->value;
+ }
if (element->value.complex->list[src->minor]->which ==
Z_StringOrNumeric_numeric)
{
++(src->minor);
- if (element->attributeSet && attributeSetP)
- {
- oident *attrset;
-
- attrset = oid_getentbyoid (element->attributeSet);
- *attributeSetP = attrset->value;
- }
return
*element->value.complex->list[src->minor-1]->u.numeric;
}
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++;
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++;
}
static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- const char **term_sub,
- oid_value attributeSet, NMEM stream,
- struct grep_info *grep_info,
- int reg_type, int complete_flag,
- int num_bases, char **basenames,
- char *term_dst)
+ const char **term_sub,
+ oid_value attributeSet, NMEM stream,
+ struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst, int xpath_use);
+
+static RSET term_trunc (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ oid_value attributeSet, NMEM stream,
+ struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst,
+ 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, 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 */,
+ zapt->term->which);
+}
+
+
+static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ oid_value attributeSet, NMEM stream,
+ struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst, int xpath_use)
{
char term_dict[2*IT_MAX_WORD+4000];
int j, r, base_no;
int truncation_value;
AttrType use;
int use_value;
+ const char *use_string = 0;
oid_value curAttributeSet = attributeSet;
const char *termp;
struct rpn_char_map_info rcmi;
rpn_char_map_prepare (zh->reg, reg_type, &rcmi);
attr_init (&use, zapt, 1);
- use_value = attr_find (&use, &curAttributeSet);
+ use_value = attr_find_ex (&use, &curAttributeSet, &use_string);
logf (LOG_DEBUG, "string_term, use value %d", use_value);
attr_init (&truncation, zapt, 5);
truncation_value = attr_find (&truncation, NULL);
logf (LOG_DEBUG, "truncation value %d", truncation_value);
- if (use_value == -1)
+ if (use_value == -1) /* no attribute - assumy "any" */
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)))
- {
- 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);
- zh->errCode = 114;
- zh->errString = nmem_strdup (stream, val_str);
- }
- else
- {
- int oid[OID_SIZE];
- struct oident oident;
-
- oident.proto = PROTO_Z3950;
- oident.oclass = CLASS_ATTSET;
- oident.value = curAttributeSet;
- oid_ent_to_oid (&oident, oid);
- zh->errCode = 121;
- zh->errString = nmem_strdup (stream, oident.desc);
- }
- return -1;
- }
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
zh->errCode = 109; /* Database unavailable */
zh->errString = basenames[base_no];
return -1;
}
+ 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 = 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)
+ {
+ /* set was found, but value wasn't defined */
+ char val_str[32];
+ sprintf (val_str, "%d (1)", use_value);
+ zh->errCode = 114;
+ zh->errString = nmem_strdup (stream, val_str);
+ }
+ else
+ {
+ int oid[OID_SIZE];
+ struct oident oident;
+
+ oident.proto = PROTO_Z3950;
+ oident.oclass = CLASS_ATTSET;
+ oident.value = curAttributeSet;
+ oid_ent_to_oid (&oident, oid);
+
+ zh->errCode = 121;
+ zh->errString = nmem_strdup (stream, oident.desc);
+ }
+ return -1;
+ }
+ }
for (local_attr = attp.local_attributes; local_attr;
local_attr = local_attr->next)
{
int ord;
- char ord_buf[32];
- int i, ord_len;
-
+ char ord_buf[32];
+ int i, ord_len;
+
ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- local_attr->local);
+ local_attr->local);
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];
- }
+
+ 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 (!prefix_len)
{
- char val_str[32];
- sprintf (val_str, "%d", use_value);
- zh->errCode = 114;
- zh->errString = nmem_strdup (stream, val_str);
+ char val_str[32];
+ sprintf (val_str, "%d (2)", use_value);
+ zh->errCode = 114;
+ zh->errString = nmem_strdup (stream, val_str);
return -1;
}
- term_dict[prefix_len++] = ')';
+ term_dict[prefix_len++] = ')';
term_dict[prefix_len++] = 1;
term_dict[prefix_len++] = reg_type;
logf (LOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]);
return 1;
}
-static void trans_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- char *termz)
+
+static int trans_term (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 HAVE_ICONV_H
+ if (zh->iconv_to_utf8 != (iconv_t)(-1))
+ {
+ 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;
+
+ yaz_log (LOG_DEBUG, "converting general from ISO-8859-1");
+ ret = iconv(zh->iconv_to_utf8, &inbuf, &inleft,
+ &outbuf, &outleft);
+ if (ret == (size_t)(-1))
+ {
+ ret = iconv(zh->iconv_to_utf8, 0, 0, 0, 0);
+ zh->errCode = 125;
+ return -1;
+ }
+ *outbuf = 0;
+ return 0;
+ }
+#endif
+ 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 0;
}
static void trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
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);
logf (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);
}
}
}
+static void grep_info_delete (struct grep_info *grep_info)
+{
+#ifdef TERM_COUNT
+ xfree(grep_info->term_no);
+#endif
+ xfree (grep_info->isam_p_buf);
+}
+
static int grep_info_prepare (ZebraHandle zh,
Z_AttributesPlusTerm *zapt,
struct grep_info *grep_info,
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];
RSET rset[60], result;
- int i, r, rset_no = 0;
+ int i, rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
while (1)
{
logf (LOG_DEBUG, "APT_phrase termp=%s", termp);
- grep_info.isam_p_indx = 0;
- r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info,
- reg_type, complete_flag, num_bases, basenames,
- term_dst);
- if (r < 1)
+ rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use);
+ if (!rset[rset_no])
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);
- assert (rset[rset_no]);
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
break;
}
-#ifdef TERM_COUNT
- xfree(grep_info.term_no);
-#endif
- xfree (grep_info.isam_p_buf);
+ grep_info_delete (&grep_info);
if (rset_no == 0)
{
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];
RSET rset[60], result;
- int i, r, rset_no = 0;
+ int i, rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
while (1)
{
logf (LOG_DEBUG, "APT_or_list termp=%s", termp);
- grep_info.isam_p_indx = 0;
- r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info,
- reg_type, complete_flag, num_bases, basenames,
- term_dst);
- if (r < 1)
+ rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use);
+ if (!rset[rset_no])
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);
- assert (rset[rset_no]);
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
break;
}
-#ifdef TERM_COUNT
- xfree(grep_info.term_no);
-#endif
- xfree (grep_info.isam_p_buf);
+ grep_info_delete (&grep_info);
if (rset_no == 0)
{
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];
RSET rset[60], result;
- int i, r, rset_no = 0;
+ int i, rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
while (1)
{
logf (LOG_DEBUG, "APT_and_list termp=%s", termp);
- grep_info.isam_p_indx = 0;
- r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info,
- reg_type, complete_flag, num_bases, basenames,
- term_dst);
- if (r < 1)
+ rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use);
+ if (!rset[rset_no])
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);
assert (rset[rset_no]);
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
break;
}
-#ifdef TERM_COUNT
- xfree(grep_info.term_no);
-#endif
- xfree (grep_info.isam_p_buf);
+ grep_info_delete (&grep_info);
if (rset_no == 0)
{
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];
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);
+ strlen(term_dst), rank_type,
+ 0 /* preserve position */,
+ zapt->term->which);
assert (rset[rset_no]);
if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
break;
}
-#ifdef TERM_COUNT
- xfree(grep_info.term_no);
-#endif
- xfree (grep_info.isam_p_buf);
+ grep_info_delete (&grep_info);
if (rset_no == 0)
{
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");
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);
}
+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;
+};
+
+static int parse_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ oid_value attributeSet,
+ struct xpath_location_step *xpath, NMEM mem)
+{
+ oid_value curAttributeSet = attributeSet;
+ AttrType use;
+ const char *use_string = 0;
+ const char *cp;
+ int no = 0;
+
+ attr_init (&use, zapt, 1);
+ attr_find_ex (&use, &curAttributeSet, &use_string);
+
+ if (!use_string || *use_string != '/')
+ return -1;
+ cp = use_string;
+ while (*cp)
+ {
+ int i = 0;
+ while (*cp && !strchr("/[",*cp))
+ {
+ i++;
+ cp++;
+ }
+ xpath[no].predicate = 0;
+ xpath[no].part = nmem_malloc (mem, i+1);
+ memcpy (xpath[no].part, cp - i, i);
+ xpath[no].part[i] = 0;
+
+ if (*cp == '[')
+ {
+ struct xpath_predicate *p = xpath[no].predicate =
+ nmem_malloc (mem, sizeof(struct xpath_predicate));
+
+ p->which = XPATH_PREDICATE_RELATION;
+ cp++;
+ while (*cp == ' ')
+ cp++;
+
+ for (i = 0; *cp && !strchr("><=] ", *cp); i++)
+ cp++;
+ p->u.relation.name = nmem_malloc (mem, i+1);
+ memcpy (p->u.relation.name, cp - i, i);
+ p->u.relation.name[i] = 0;
+ while (*cp == ' ')
+ cp++;
+ if (*cp != ']')
+ {
+ for (i = 0; *cp && strchr(">=<!", *cp); i++)
+ cp++;
+
+ p->u.relation.op = nmem_malloc (mem, i+1);
+ if (i)
+ memcpy (p->u.relation.op, cp - i, i);
+ p->u.relation.op[i] = 0;
+
+ while (*cp == ' ')
+ cp++;
+
+ if (strchr("\"'", *cp))
+ {
+ cp++;
+ for (i = 0; *cp && !strchr("\"'", *cp); i++)
+ cp++;
+
+ p->u.relation.value = nmem_malloc (mem, i+1);
+ if (i)
+ memcpy (p->u.relation.value, cp - i, i);
+ p->u.relation.value[i] = 0;
+ yaz_log (LOG_LOG, "value=%s", p->u.relation.value);
+
+ cp++;
+ }
+ else
+ {
+ for (i = 0; *cp && !strchr(" ]", *cp); i++)
+ cp++;
+ p->u.relation.value = nmem_malloc (mem, i+1);
+ if (i)
+ memcpy (p->u.relation.value, cp - i, i);
+ p->u.relation.value[i] = 0;
+ }
+ while (*cp == ' ')
+ cp++;
+ }
+ if (*cp == ']')
+ cp++;
+ } /* end of ] predicate */
+ no++;
+ if (*cp != '/')
+ break;
+ cp++;
+ }
+ return no;
+}
+
+
+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))
+ {
+ 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 (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];
+ }
+ 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);
+
+ for (base_no = 0; base_no < num_bases; base_no++)
+ {
+ 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;
+ }
+ while (--level >= 0)
+ {
+ 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;
+
+ if (xpath[level].predicate &&
+ xpath[level].predicate->which == XPATH_PREDICATE_RELATION &&
+ xpath[level].predicate->u.relation.name[0])
+ {
+ char predicate_str[128];
+
+ strcpy (predicate_str,
+ xpath[level].predicate->u.relation.name+1);
+ if (xpath[level].predicate->u.relation.value)
+ {
+ strcat (predicate_str, "=");
+ strcat (predicate_str,
+ xpath[level].predicate->u.relation.value);
+ }
+ rset_attr = xpath_trunc (
+ zh, stream, '0', predicate_str, 3, curAttributeSet);
+ }
+ 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;
+ }
+ }
+
+ return rset;
+}
+
+
static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
oid_value attributeSet, NMEM stream,
int complete_flag;
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 (trans_term (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, stream);
+ if (xpath_len >= 0)
+ {
+ xpath_use = 1016;
+ if (xpath[xpath_len-1].part[0] == '@')
+ xpath_use = 1015;
+ }
if (!strcmp (search_type, "phrase"))
{
- return rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream,
+ 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"))
{
- return rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream,
+ 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"))
{
- return rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream,
+ 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"))
{
- return rpn_search_APT_local (zh, zapt, termz, attributeSet, stream,
+ rset = rpn_search_APT_local (zh, zapt, termz, attributeSet, stream,
rank_type);
}
else if (!strcmp (search_type, "numeric"))
{
- return rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream,
+ rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream,
reg_id, complete_flag, rank_type,
num_bases, basenames);
}
- zh->errCode = 118;
- return NULL;
+ else if (!strcmp (search_type, "always"))
+ {
+ rset = 0;
+ }
+ else
+ zh->errCode = 118;
+ 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,
&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);
+ 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);
+ strlen(glist[i+before].term), NULL, 0,
+ zapt->term->which);
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
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);
+ 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);
+ 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;
projects / idzebra-moved-to-github.git / blobdiff