-/* $Id: zrpn.c,v 1.125 2002-10-03 10:16:23 adam Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
+/* $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.
#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;
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)
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);
{
/* 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);
+ sprintf (val_str, "%d", use_value);
+ 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;
if (!prefix_len)
{
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);
+ 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;
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, term_type);
parms.rset_term->nn = min_nn;
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);
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,
return rset_create (rset_kind_null, &parms);
}
+/* pop - moved to xpath.c */
+#if 0
+
struct xpath_predicate {
int which;
union {
struct xpath_predicate *predicate;
};
+#endif
+
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,
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;
{
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"))
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
{
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 */
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);
}
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