-/* $Id: zrpn.c,v 1.126 2002-12-16 22:59:34 adam Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
+/* $Id: zrpn.c,v 1.144 2004-08-20 14:44:46 heikki Exp $
+ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
Index Data Aps
This file is part of the Zebra server.
#include <ctype.h>
#include "index.h"
+#include <zebra_xpath.h>
#include <charmap.h>
#include <rstemp.h>
#include <rsnull.h>
#include <rsbool.h>
#include <rsbetween.h>
+#include <rsprox.h>
struct rpn_char_map_info {
ZebraMaps zm;
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;
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)
{
#ifdef TERM_COUNT
int *term_no;
#endif
- ISAMS_P *isam_p_buf;
+ ISAMC_P *isam_p_buf;
int isam_p_size;
int isam_p_indx;
ZebraHandle zh;
{
if (p->isam_p_indx == p->isam_p_size)
{
- ISAMS_P *new_isam_p_buf;
+ ISAMC_P *new_isam_p_buf;
#ifdef TERM_COUNT
int *new_term_no;
#endif
p->isam_p_size = 2*p->isam_p_size + 100;
- new_isam_p_buf = (ISAMS_P *) xmalloc (sizeof(*new_isam_p_buf) *
+ new_isam_p_buf = (ISAMC_P *) xmalloc (sizeof(*new_isam_p_buf) *
p->isam_p_size);
if (p->isam_p_buf)
{
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++;
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 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);
zh->errString = basenames[base_no];
return -1;
}
- if (use_value == -2) /* string attribute (assume IDXPATH/any) */
+ if (xpath_use > 0 && use_value == -2)
{
use_value = xpath_use;
attp.local_attributes = &id_xpath_attr;
}
else
{
- if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value)))
+ if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value,
+ use_string)))
{
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);
+ errCode = 114;
+ if (use_string)
+ errString = nmem_strdup(stream, use_string);
+ else
+ {
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ 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;
}
if (!prefix_len)
{
+#if 1
+ bases_ok++;
+#else
char val_str[32];
- sprintf (val_str, "%d (2)", use_value);
- zh->errCode = 114;
- zh->errString = nmem_strdup (stream, val_str);
- return -1;
+ sprintf (val_str, "%d", use_value);
+ errCode = 114;
+ errString = nmem_strdup (stream, val_str);
+#endif
+ continue;
}
+ bases_ok++; /* this has OK attributes */
+
term_dict[prefix_len++] = ')';
term_dict[prefix_len++] = 1;
term_dict[prefix_len++] = reg_type;
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;
return 0;
}
-static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no,
- int ordered, int exclusion, int relation, int distance)
-{
- int i;
- RSFD *rsfd;
- int *more;
- struct it_key **buf;
- RSET result;
- char prox_term[1024];
- 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);
- more = (int *) xmalloc (sizeof(*more)*rset_no);
- buf = (struct it_key **) xmalloc (sizeof(*buf)*rset_no);
-
- *prox_term = '\0';
- for (i = 0; i<rset_no; i++)
- {
- int j;
- for (j = 0; j<rset[i]->no_rset_terms; j++)
- {
- const char *nflags = rset[i]->rset_terms[j]->flags;
- char *term = rset[i]->rset_terms[j]->name;
- int lterm = strlen(term);
- if (lterm + length_prox_term < sizeof(prox_term)-1)
- {
- if (length_prox_term)
- prox_term[length_prox_term++] = ' ';
- strcpy (prox_term + length_prox_term, term);
- length_prox_term += lterm;
- }
- 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++)
- {
- buf[i] = 0;
- rsfd[i] = 0;
- }
- for (i = 0; i<rset_no; i++)
- {
- buf[i] = (struct it_key *) xmalloc (sizeof(**buf));
- rsfd[i] = rset_open (rset[i], RSETF_READ);
- if (!(more[i] = rset_read (rset[i], rsfd[i], buf[i], &term_index)))
- break;
- }
- if (i != rset_no)
- {
- /* at least one is empty ... return null set */
- rset_null_parms parms;
-
- parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags, term_type);
- parms.rset_term->nn = 0;
- result = rset_create (rset_kind_null, &parms);
- }
- else if (ordered && relation == 3 && exclusion == 0 && distance == 1)
- {
- /* special proximity case = phrase search ... */
- rset_temp_parms parms;
- RSFD rsfd_result;
-
- parms.rset_term = rset_term_create (prox_term, length_prox_term,
- 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++)
- {
- int cmp;
-
- if (!more[i])
- {
- *more = 0;
- break;
- }
- cmp = key_compare_it (buf[i], buf[i-1]);
- if (cmp > 1)
- {
- more[i-1] = rset_read (rset[i-1], rsfd[i-1],
- buf[i-1], &term_index);
- break;
- }
- else if (cmp == 1)
- {
- if (buf[i-1]->seqno+1 != buf[i]->seqno)
- {
- more[i-1] = rset_read (rset[i-1], rsfd[i-1],
- buf[i-1], &term_index);
- break;
- }
- }
- else
- {
- more[i] = rset_read (rset[i], rsfd[i], buf[i],
- &term_index);
- break;
- }
- }
- if (i == rset_no)
- {
- rset_write (result, rsfd_result, buf[0]);
- more[0] = rset_read (*rset, *rsfd, *buf, &term_index);
- }
- }
- rset_close (result, rsfd_result);
- }
- else if (rset_no == 2)
- {
- /* generic proximity case (two input sets only) ... */
- rset_temp_parms parms;
- RSFD rsfd_result;
-
- 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, 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[0] && more[1])
- {
- int cmp = key_compare_it (buf[0], buf[1]);
- if (cmp < -1)
- more[0] = rset_read (rset[0], rsfd[0], buf[0], &term_index);
- else if (cmp > 1)
- more[1] = rset_read (rset[1], rsfd[1], buf[1], &term_index);
- else
- {
- int sysno = buf[0]->sysno;
- int seqno[500];
- int n = 0;
-
- seqno[n++] = buf[0]->seqno;
- while ((more[0] = rset_read (rset[0], rsfd[0], buf[0],
- &term_index)) &&
- sysno == buf[0]->sysno)
- if (n < 500)
- seqno[n++] = buf[0]->seqno;
- do
- {
- for (i = 0; i<n; i++)
- {
- int diff = buf[1]->seqno - seqno[i];
- int excl = exclusion;
- if (!ordered && diff < 0)
- diff = -diff;
- switch (relation)
- {
- case 1: /* < */
- if (diff < distance && diff >= 0)
- excl = !excl;
- break;
- case 2: /* <= */
- if (diff <= distance && diff >= 0)
- excl = !excl;
- break;
- case 3: /* == */
- if (diff == distance && diff >= 0)
- excl = !excl;
- break;
- case 4: /* >= */
- if (diff >= distance && diff >= 0)
- excl = !excl;
- break;
- case 5: /* > */
- if (diff > distance && diff >= 0)
- excl = !excl;
- break;
- case 6: /* != */
- if (diff != distance && diff >= 0)
- excl = !excl;
- break;
- }
- if (excl)
- {
- rset_write (result, rsfd_result, buf[1]);
- break;
- }
- }
- } while ((more[1] = rset_read (rset[1], rsfd[1], buf[1],
- &term_index)) &&
- sysno == buf[1]->sysno);
- }
- }
- rset_close (result, rsfd_result);
- }
- else
- {
- rset_null_parms parms;
-
- parms.rset_term = rset_term_create (prox_term, length_prox_term,
- flags, term_type);
- parms.rset_term->nn = 0;
- result = rset_create (rset_kind_null, &parms);
- }
- for (i = 0; i<rset_no; i++)
- {
- if (rsfd[i])
- rset_close (rset[i], rsfd[i]);
- xfree (buf[i]);
- }
- xfree (buf);
- xfree (more);
- xfree (rsfd);
- return result;
-}
-
-
char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
const char *termz, NMEM stream, unsigned reg_id)
{
{
char term_dst[IT_MAX_WORD+1];
RSET rset[60], result;
- int i, rset_no = 0;
+ int rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
if (rset_no == 0)
{
rset_null_parms parms;
-
- 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)
return (rset[0]);
- result = rpn_prox (zh, rset, rset_no, 1, 0, 3, 1);
- for (i = 0; i<rset_no; i++)
- rset_delete (rset[i]);
+ else
+ {
+ /* new / old prox */
+ rset_prox_parms parms;
+
+ parms.rset = rset;
+ parms.rset_no = rset_no;
+ parms.ordered = 1;
+ parms.exclusion = 0;
+ parms.relation = 3;
+ parms.distance = 1;
+ parms.key_size = sizeof(struct it_key);
+ parms.cmp = key_compare_it;
+ parms.getseq = key_get_seq;
+ parms.log_item = key_logdump_txt;
+ result = rset_create(rset_kind_prox, &parms);
+ }
return result;
}
if (rset_no == 0)
{
rset_null_parms parms;
-
- parms.rset_term = rset_term_create (termz, -1, rank_type,
- zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
bool_parms.rset_r = rset[i];
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
+ bool_parms.log_item = key_logdump_txt;
result = rset_create (rset_kind_or, &bool_parms);
}
return result;
if (rset_no == 0)
{
rset_null_parms parms;
-
- parms.rset_term = rset_term_create (termz, -1, rank_type,
- zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
bool_parms.rset_r = rset[i];
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
+ bool_parms.log_item = key_logdump_txt;
result = rset_create (rset_kind_and, &bool_parms);
}
return result;
oid_value attributeSet, struct grep_info *grep_info,
int reg_type, int complete_flag,
int num_bases, char **basenames,
- char *term_dst, int xpath_use)
+ char *term_dst, int xpath_use, NMEM stream)
{
char term_dict[2*IT_MAX_WORD+2];
int r, base_no;
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_ex (&use, &curAttributeSet, &use_string);
}
else
{
- if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value)))
+ if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value,
+ use_string)))
{
logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d",
curAttributeSet, use_value, r);
if (r == -1)
- zh->errCode = 114;
+ {
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ errString = nmem_strdup (stream, val_str);
+ errCode = 114;
+ }
else
- zh->errCode = 121;
- return -1;
+ 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;
grep_info.isam_p_indx = 0;
r = numeric_term (zh, zapt, &termp, attributeSet, &grep_info,
reg_type, complete_flag, num_bases, basenames,
- term_dst, xpath_use);
+ term_dst, xpath_use,
+ stream);
if (r < 1)
break;
logf (LOG_DEBUG, "term: %s", term_dst);
if (rset_no == 0)
{
rset_null_parms parms;
-
- parms.rset_term = rset_term_create (term_dst, -1, rank_type,
- zapt->term->which);
return rset_create (rset_kind_null, &parms);
}
result = rset[0];
bool_parms.rset_r = rset[i];
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
+ bool_parms.log_item = key_logdump_txt;
result = rset_create (rset_kind_and, &bool_parms);
}
return result;
RSET result;
RSFD rsfd;
struct it_key key;
+ int sys;
rset_temp_parms parms;
- 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");
result = rset_create (rset_kind_temp, &parms);
rsfd = rset_open (result, RSETF_WRITE);
- key.sysno = atoi (termz);
+ sys = atoi(termz);
+ if (sys <= 0)
+ sys = 1;
+#if IT_KEY_NEW
+ key.mem[0] = sys;
+ key.mem[1] = 1;
+ key.len = 2;
+#else
+ key.sysno = sys;
key.seqno = 1;
if (key.sysno <= 0)
key.sysno = 1;
+#endif
rset_write (result, rsfd, &key);
rset_close (result, rsfd);
return result;
sks->sortRelation = (int *)
nmem_malloc (stream, sizeof(*sks->sortRelation));
if (sort_relation_value == 1)
- *sks->sortRelation = Z_SortRelation_ascending;
+ *sks->sortRelation = Z_SortKeySpec_ascending;
else if (sort_relation_value == 2)
- *sks->sortRelation = Z_SortRelation_descending;
+ *sks->sortRelation = Z_SortKeySpec_descending;
else
- *sks->sortRelation = Z_SortRelation_ascending;
+ *sks->sortRelation = Z_SortKeySpec_ascending;
sks->caseSensitivity = (int *)
nmem_malloc (stream, sizeof(*sks->caseSensitivity));
sks->which = Z_SortKeySpec_null;
sks->u.null = odr_nullval ();
sort_sequence->specs[i] = sks;
-
- 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)
+ struct xpath_location_step *xpath, int max, 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;
+ 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,
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)
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);
+ WRBUF wbuf = wrbuf_alloc();
+ wrbuf_puts(wbuf, 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);
+ 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', predicate_str, 3, curAttributeSet);
+ zh, stream, '0', wrbuf_buf(wbuf), 3, curAttributeSet);
+ wrbuf_free(wbuf, 1);
}
else
{
if (sort_flag)
return rpn_sort_spec (zh, zapt, attributeSet, stream, sort_sequence,
rank_type);
- xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, stream);
+ xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream);
if (xpath_len >= 0)
{
xpath_use = 1016;
}
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
+ bool_parms.log_item = key_logdump_txt;
switch (zop->which)
{
}
else
{
- RSET rsets[2];
-
- rsets[0] = bool_parms.rset_l;
- rsets[1] = bool_parms.rset_r;
+ /* new / old prox */
+ rset_prox_parms parms;
+ RSET twosets[2];
- r = rpn_prox (zh, rsets, 2,
- *zop->u.prox->ordered,
- (!zop->u.prox->exclusion ? 0 :
- *zop->u.prox->exclusion),
- *zop->u.prox->relationType,
- *zop->u.prox->distance);
- rset_delete (rsets[0]);
- rset_delete (rsets[1]);
+ twosets[0] = bool_parms.rset_l;
+ twosets[1] = bool_parms.rset_r;
+ parms.rset = twosets;
+ parms.rset_no = 2;
+ parms.ordered = *zop->u.prox->ordered;
+ parms.exclusion = (!zop->u.prox->exclusion ? 0 :
+ *zop->u.prox->exclusion);
+ parms.relation = *zop->u.prox->relationType;
+ parms.distance = *zop->u.prox->distance;
+ parms.key_size = sizeof(struct it_key);
+ parms.cmp = key_compare_it;
+ parms.getseq = key_get_seq;
+ parms.log_item = key_logdump_txt;
+ r = rset_create(rset_kind_prox, &parms);
}
break;
default:
nmem_strdup (stream, zs->u.simple->u.resultSetId);
return 0;
}
+ else
+ rset_dup(r);
}
else
{
struct scan_info_entry {
char *term;
- ISAMS_P isam_p;
+ ISAMC_P isam_p;
};
struct scan_info {
scan_info->list[idx].term = (char *)
odr_malloc (scan_info->odr, strlen(name + len_prefix)+1);
strcpy (scan_info->list[idx].term, name + len_prefix);
- assert (*info == sizeof(ISAMS_P));
- memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAMS_P));
+ assert (*info == sizeof(ISAMC_P));
+ memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAMC_P));
return 0;
}
static void count_set (RSET r, int *count)
{
- int psysno = 0;
+ zint psysno = 0;
int kno = 0;
struct it_key key;
RSFD rfd;
- int term_index;
logf (LOG_DEBUG, "count_set");
*count = 0;
rfd = rset_open (r, RSETF_READ);
- while (rset_read (r, rfd, &key, &term_index))
+ while (rset_read (r, rfd, &key))
{
+#if IT_KEY_NEW
+ if (key.mem[0] != psysno)
+ {
+ psysno = key.mem[0];
+ (*count)++;
+ }
+#else
if (key.sysno != psysno)
{
psysno = key.sysno;
(*count)++;
}
+#endif
kno++;
}
rset_close (r, rfd);
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;
char termz[IT_MAX_WORD+20];
AttrType use;
int use_value;
+ const char *use_string = 0;
struct scan_info *scan_info_array;
ZebraScanEntry *glist;
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);
attr_init (&use, zapt, 1);
- use_value = attr_find (&use, &attributeset);
+ use_value = attr_find_ex (&use, &attributeset, &use_string);
if (zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type,
rank_type, &complete_flag, &sort_flag))
attent attp;
data1_local_attribute *local_attr;
- if ((r=att_getentbyatt (zh, &attp, attributeset, use_value)))
+ if ((r=att_getentbyatt (zh, &attp, attributeset, use_value,
+ use_string)))
{
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 */
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
- bool_parms.rset_l = rset;
+ bool_parms.log_item = key_logdump_txt;
+ bool_parms.rset_l = rset;
bool_parms.rset_r = rset2;
rset = rset_create (rset_kind_or, &bool_parms);
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.log_item = key_logdump_txt;
+ 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);
}
bool_parms.key_size = sizeof(struct it_key);
bool_parms.cmp = key_compare_it;
- bool_parms.rset_l = rset;
+ bool_parms.log_item = key_logdump_txt;
+ bool_parms.rset_l = rset;
bool_parms.rset_r = rset2;
rset = rset_create (rset_kind_or, &bool_parms);
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.log_item = key_logdump_txt;
+ 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);
}