-/* $Id: zrpn.c,v 1.177 2005-04-25 11:54:08 adam Exp $
+/* $Id: zrpn.c,v 1.188 2005-05-03 14:04:31 adam Exp $
Copyright (C) 1995-2005
Index Data ApS
02111-1307, USA.
*/
-
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#endif
#include <ctype.h>
+#include <yaz/diagbib1.h>
#include "index.h"
#include <zebra_xpath.h>
#include <charmap.h>
#include <rset.h>
-/* maximum number of terms in an and/or/phrase item */
-#define TERM_LIST_LENGTH_MAX 256
-
-static const struct key_control it_ctrl =
-{
- sizeof(struct it_key),
- 2, /* we have sysnos and seqnos in this key, nothing more */
- key_compare_it,
- key_logdump_txt, /* FIXME - clean up these functions */
- key_get_seq,
-};
-
-
-const struct key_control *key_it_ctrl = &it_ctrl;
-
struct rpn_char_map_info
{
ZebraMaps zm;
}
static int term_pre(ZebraMaps zebra_maps, int reg_type, const char **src,
- const char *ct1, const char *ct2, int first)
+ const char *ct1, const char *ct2, int first)
{
const char *s1, *s0 = *src;
const char **map;
/* 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,
- char *dst_term)
+ const char **src, char *dst, int space_split,
+ char *dst_term)
{
const char *s0;
const char **map;
/* term_101: handle term, where trunc = Process # */
static int term_101(ZebraMaps zebra_maps, int reg_type,
- const char **src, char *dst, int space_split,
- char *dst_term)
+ const char **src, char *dst, int space_split,
+ char *dst_term)
{
const char *s0;
const char **map;
/* term_103: handle term, where trunc = re-2 (regular expressions) */
static int term_103(ZebraMaps zebra_maps, int reg_type, const char **src,
- char *dst, int *errors, int space_split,
- char *dst_term)
+ char *dst, int *errors, int space_split,
+ char *dst_term)
{
int i = 0;
int j = 0;
}
/* term_103: handle term, where trunc = re-1 (regular expressions) */
-static int term_102 (ZebraMaps zebra_maps, int reg_type, const char **src,
- char *dst, int space_split, char *dst_term)
+static int term_102(ZebraMaps zebra_maps, int reg_type, const char **src,
+ char *dst, int space_split, char *dst_term)
{
return term_103(zebra_maps, reg_type, src, dst, NULL, space_split,
- dst_term);
+ dst_term);
}
/* term_104: handle term, where trunc = Process # and ! */
static int term_104(ZebraMaps zebra_maps, int reg_type,
- const char **src, char *dst, int space_split,
- char *dst_term)
+ const char **src, char *dst, int space_split,
+ char *dst_term)
{
- const char *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
}
else
{
- s1 = s0;
- map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0);
+ const char *s1 = s0;
+ int q_map_match = 0;
+ map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0),
+ &q_map_match);
if (space_split && **map == *CHR_SPACE)
break;
- while (s1 < s0)
- {
- if (strchr(REGEX_CHARS, *s1))
- dst[i++] = '\\';
- dst_term[j++] = *s1;
- dst[i++] = *s1++;
- }
+
+ /* add non-space char */
+ memcpy(dst_term+j, s1, s0 - s1);
+ j += (s0 - s1);
+ if (!q_map_match)
+ {
+ while (s1 < s0)
+ {
+ if (strchr(REGEX_CHARS, *s1))
+ dst[i++] = '\\';
+ dst[i++] = *s1++;
+ }
+ }
+ else
+ {
+ char tmpbuf[80];
+ esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0]));
+
+ strcpy(dst + i, map[0]);
+ i += strlen(map[0]);
+ }
}
}
dst[i] = '\0';
}
/* term_105/106: handle term, where trunc = Process * and ! and right trunc */
-static int term_105 (ZebraMaps zebra_maps, int reg_type,
- const char **src, char *dst, int space_split,
- char *dst_term, int right_truncate)
+static int term_105(ZebraMaps zebra_maps, int reg_type,
+ const char **src, char *dst, int space_split,
+ char *dst_term, int right_truncate)
{
- const char *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
}
else
{
- s1 = s0;
- map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0);
+ const char *s1 = s0;
+ int q_map_match = 0;
+ map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0),
+ &q_map_match);
if (space_split && **map == *CHR_SPACE)
break;
- while (s1 < s0)
- {
- if (strchr(REGEX_CHARS, *s1))
- dst[i++] = '\\';
- dst_term[j++] = *s1;
- dst[i++] = *s1++;
- }
+
+ /* add non-space char */
+ memcpy(dst_term+j, s1, s0 - s1);
+ j += (s0 - s1);
+ if (!q_map_match)
+ {
+ while (s1 < s0)
+ {
+ if (strchr(REGEX_CHARS, *s1))
+ dst[i++] = '\\';
+ dst[i++] = *s1++;
+ }
+ }
+ else
+ {
+ char tmpbuf[80];
+ esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0]));
+
+ strcpy(dst + i, map[0]);
+ i += strlen(map[0]);
+ }
}
}
if (right_truncate)
switch (relation_value)
{
case 1:
- if (!term_100 (zh->reg->zebra_maps, reg_type,
- term_sub, term_component,
- space_split, term_dst))
+ if (!term_100(zh->reg->zebra_maps, reg_type,
+ term_sub, term_component,
+ space_split, term_dst))
return 0;
yaz_log(log_level_rpn, "Relation <");
if (i)
*term_tmp++ = '|';
while (j < i)
- string_rel_add_char (&term_tmp, term_component, &j);
+ string_rel_add_char(&term_tmp, term_component, &j);
*term_tmp++ = '[';
*term_tmp++ = '^';
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = '-';
*term_tmp++ = ']';
*term_tmp = '\0';
break;
case 2:
- if (!term_100 (zh->reg->zebra_maps, reg_type,
- term_sub, term_component,
- space_split, term_dst))
+ if (!term_100(zh->reg->zebra_maps, reg_type,
+ term_sub, term_component,
+ space_split, term_dst))
return 0;
yaz_log(log_level_rpn, "Relation <=");
int j = 0;
while (j < i)
- string_rel_add_char (&term_tmp, term_component, &j);
+ string_rel_add_char(&term_tmp, term_component, &j);
*term_tmp++ = '[';
*term_tmp++ = '^';
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = '-';
*term_tmp++ = ']';
break;
}
for (i = 0; term_component[i]; )
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = ')';
*term_tmp = '\0';
break;
int j = 0;
while (j < i)
- string_rel_add_char (&term_tmp, term_component, &j);
+ string_rel_add_char(&term_tmp, term_component, &j);
*term_tmp++ = '[';
*term_tmp++ = '^';
*term_tmp++ = '-';
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = ']';
*term_tmp++ = '.';
break;
}
for (i = 0; term_component[i];)
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = '.';
*term_tmp++ = '+';
*term_tmp++ = ')';
*term_tmp = '\0';
break;
case 4:
- if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub,
- term_component, space_split, term_dst))
+ if (!term_100(zh->reg->zebra_maps, reg_type, term_sub,
+ term_component, space_split, term_dst))
return 0;
yaz_log(log_level_rpn, "Relation >=");
if (i)
*term_tmp++ = '|';
while (j < i)
- string_rel_add_char (&term_tmp, term_component, &j);
+ string_rel_add_char(&term_tmp, term_component, &j);
*term_tmp++ = '[';
if (term_component[i+1])
{
*term_tmp++ = '^';
*term_tmp++ = '-';
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
}
else
{
- string_rel_add_char (&term_tmp, term_component, &i);
+ string_rel_add_char(&term_tmp, term_component, &i);
*term_tmp++ = '-';
}
*term_tmp++ = ']';
case 102:
case -1:
yaz_log(log_level_rpn, "Relation =");
- if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub,
- term_component, space_split, term_dst))
+ if (!term_100(zh->reg->zebra_maps, reg_type, term_sub,
+ term_component, space_split, term_dst))
return 0;
strcat(term_tmp, "(");
strcat(term_tmp, term_component);
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,
- NMEM rset_nmem)
+static ZEBRA_RES 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,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
ZEBRA_RES res;
+ *rset = 0;
grep_info->isam_p_indx = 0;
res = string_term(zh, zapt, term_sub, attributeSet, stream, grep_info,
reg_type, complete_flag, num_bases, basenames,
term_dst, xpath_use);
if (res != ZEBRA_OK)
- return 0;
+ return res;
if (!*term_sub) /* no more terms ? */
- return 0;
+ return res;
yaz_log(log_level_rpn, "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, rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
+ *rset = 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, rset_nmem,
+ kc, kc->scope);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
}
+
static char *nmem_strdup_i(NMEM nmem, int v)
{
char val_str[64];
- sprintf (val_str, "%d", v);
+ sprintf(val_str, "%d", v);
return nmem_strdup(nmem, val_str);
}
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);
+ attr_init(&use, zapt, 1);
+ use_value = attr_find_ex(&use, &curAttributeSet, &use_string);
yaz_log(log_level_rpn, "string_term, use value %d", use_value);
- attr_init (&truncation, zapt, 5);
- truncation_value = attr_find (&truncation, NULL);
+ attr_init(&truncation, zapt, 5);
+ truncation_value = attr_find(&truncation, NULL);
yaz_log(log_level_rpn, "truncation value %d", truncation_value);
if (use_value == -1) /* no attribute - assumy "any" */
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
- zh->errCode = 109; /* Database unavailable */
+ zh->errCode = YAZ_BIB1_DATABASE_UNAVAILABLE;
zh->errString = basenames[base_no];
return ZEBRA_FAIL;
}
if (r == -1)
{
/* set was found, but value wasn't defined */
- errCode = 114;
+ errCode = YAZ_BIB1_UNSUPP_USE_ATTRIBUTE;
if (use_string)
errString = nmem_strdup(stream, use_string);
else
oident.value = curAttributeSet;
oid_ent_to_oid (&oident, oid);
- errCode = 121;
- errString = nmem_strdup (stream, oident.desc);
+ errCode = YAZ_BIB1_UNSUPP_ATTRIBUTE_SET;
+ errString = nmem_strdup(stream, oident.desc);
}
continue;
}
strcat(term_dict, ")");
break;
case 103: /* Regexp-2 */
- r = 1;
+ regex_range = 1;
term_dict[j++] = '(';
init_pos = 2;
- if (!term_103 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, ®ex_range,
- space_split, term_dst))
+ if (!term_103(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, ®ex_range,
+ space_split, term_dst))
{
*term_sub = 0;
return ZEBRA_OK;
}
strcat(term_dict, ")");
+ break;
case 104: /* process # and ! in term */
term_dict[j++] = '(';
- if (!term_104 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
+ if (!term_104(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
{
*term_sub = 0;
return ZEBRA_OK;
break;
case 105: /* process * and ! in term */
term_dict[j++] = '(';
- if (!term_105 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst, 1))
+ if (!term_105(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst, 1))
{
*term_sub = 0;
return ZEBRA_OK;
break;
case 106: /* process * and ! in term */
term_dict[j++] = '(';
- if (!term_105 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst, 0))
+ if (!term_105(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst, 0))
{
*term_sub = 0;
return ZEBRA_OK;
strcat(term_dict, ")");
break;
default:
- zh->errCode = 120;
+ zh->errCode = YAZ_BIB1_UNSUPP_TRUNCATION_ATTRIBUTE;
zh->errString = nmem_strdup_i(stream, truncation_value);
return ZEBRA_FAIL;
}
/* convert APT search term to UTF8 */
-static int zapt_term_to_utf8 (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- char *termz)
+static ZEBRA_RES zapt_term_to_utf8(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ char *termz)
{
size_t sizez;
Z_Term *term = zapt->term;
if (ret == (size_t)(-1))
{
ret = yaz_iconv(zh->iconv_to_utf8, 0, 0, 0, 0);
- zh->errCode = 125;
+ zh->errCode =
+ YAZ_BIB1_QUERY_TERM_INCLUDES_CHARS_THAT_DO_NOT_TRANSLATE_INTO_;
return -1;
}
*outbuf = 0;
termz[sizez] = '\0';
break;
default:
- zh->errCode = 124;
- return -1;
+ zh->errCode = YAZ_BIB1_UNSUPP_CODED_VALUE_FOR_TERM;
+ return ZEBRA_FAIL;
}
- return 0;
+ return ZEBRA_OK;
}
/* convert APT SCAN term to internal cmap */
-static int trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- char *termz, int reg_type)
+static ZEBRA_RES trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ char *termz, int reg_type)
{
char termz0[IT_MAX_WORD];
- if (zapt_term_to_utf8(zh, zapt, termz0))
- return -1; /* error */
+ if (zapt_term_to_utf8(zh, zapt, termz0) == ZEBRA_FAIL)
+ return ZEBRA_FAIL; /* error */
else
{
const char **map;
while ((len = (cp_end - cp)) > 0)
{
- map = zebra_maps_input (zh->reg->zebra_maps, reg_type, &cp, len, 0);
+ map = zebra_maps_input(zh->reg->zebra_maps, reg_type, &cp, len, 0);
if (**map == *CHR_SPACE)
space_map = *map;
else
}
termz[i] = '\0';
}
- return 0;
+ return ZEBRA_OK;
}
char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
int truncation_value;
char *ex_list = 0;
- attr_init (&truncation, zapt, 5);
- truncation_value = attr_find (&truncation, NULL);
+ attr_init(&truncation, zapt, 5);
+ truncation_value = attr_find(&truncation, NULL);
switch (truncation_value)
{
}
}
-static void grep_info_delete (struct grep_info *grep_info)
+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);
+ xfree(grep_info->isam_p_buf);
}
-static int grep_info_prepare (ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- struct grep_info *grep_info,
- int reg_type,
- NMEM stream)
+static int grep_info_prepare(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ struct grep_info *grep_info,
+ int reg_type,
+ NMEM stream)
{
AttrType termset;
int termset_value_numeric;
if (!zapt)
return 0;
- attr_init (&termset, zapt, 8);
+ attr_init(&termset, zapt, 8);
termset_value_numeric =
- attr_find_ex (&termset, NULL, &termset_value_string);
+ attr_find_ex(&termset, NULL, &termset_value_string);
if (termset_value_numeric != -1)
{
char resname[32];
if (termset_value_numeric != -2)
{
- sprintf (resname, "%d", termset_value_numeric);
+ sprintf(resname, "%d", termset_value_numeric);
termset_name = resname;
}
else
termset_name = termset_value_string;
yaz_log(log_level_rpn, "creating termset set %s", termset_name);
- grep_info->termset = resultSetAdd (zh, termset_name, 1);
+ grep_info->termset = resultSetAdd(zh, termset_name, 1);
if (!grep_info->termset)
{
- zh->errCode = 128;
- zh->errString = nmem_strdup (stream, termset_name);
+ zh->errCode = YAZ_BIB1_ILLEGAL_RESULT_SET_NAME;
+ zh->errString = nmem_strdup(stream, termset_name);
return -1;
}
}
}
-static RSET rpn_search_APT_phrase(ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type, int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static ZEBRA_RES term_list_trunc(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type, int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem,
+ RSET **result_sets, int *num_result_sets,
+ struct rset_key_control *kc)
{
char term_dst[IT_MAX_WORD+1];
- RSET rset[TERM_LIST_LENGTH_MAX], result;
- size_t rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
+ int alloc_sets = 0;
+ *num_result_sets = 0;
*term_dst = 0;
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
- for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
+ return ZEBRA_FAIL;
+ while(1)
{
- yaz_log(log_level_rpn, "APT_phrase termp=%s", termp);
- 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, rset_nmem);
- if (!rset[rset_no]) /* error or no more terms */
+ ZEBRA_RES res;
+
+ if (alloc_sets == *num_result_sets)
+ {
+ int add = 10;
+ RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) *
+ sizeof(*rnew));
+ if (alloc_sets)
+ memcpy(rnew, *result_sets, alloc_sets * sizeof(*rnew));
+ alloc_sets = alloc_sets + add;
+ *result_sets = rnew;
+ }
+ res = term_trunc(zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use, rset_nmem,
+ &(*result_sets)[*num_result_sets],
+ kc);
+ if (res != ZEBRA_OK)
+ {
+ int i;
+ for (i = 0; i < *num_result_sets; i++)
+ rset_delete((*result_sets)[i]);
+ grep_info_delete (&grep_info);
+ return res;
+ }
+ if ((*result_sets)[*num_result_sets] == 0)
break;
+ (*num_result_sets)++;
}
- grep_info_delete (&grep_info);
- if (termp)
- {
- int i;
- for (i = 0; i<rset_no; i++)
- rset_delete(rset[i]);
- return 0;
- }
- if (rset_no == 0)
- return rsnull_create (rset_nmem, key_it_ctrl);
- else if (rset_no == 1)
- return (rset[0]);
- else
- result = rsprox_create( rset_nmem, key_it_ctrl, key_it_ctrl->scope,
- rset_no, rset,
- 1 /* ordered */, 0 /* exclusion */,
- 3 /* relation */, 1 /* distance */);
- return result;
+ grep_info_delete(&grep_info);
+ return ZEBRA_OK;
}
-static RSET rpn_search_APT_or_list(ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type,
- int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static ZEBRA_RES rpn_search_APT_phrase(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type, int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
- char term_dst[IT_MAX_WORD+1];
- RSET rset[TERM_LIST_LENGTH_MAX];
- size_t rset_no = 0;
- struct grep_info grep_info;
- char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
- const char *termp = termz;
-
- if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
- return 0;
- for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
- {
- yaz_log(log_level_rpn, "APT_or_list termp=%s", termp);
- 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, rset_nmem);
- if (!rset[rset_no]) /* error or no more terms */
- break;
- }
- grep_info_delete (&grep_info);
- if (termp)
- {
- int i;
- for (i = 0; i<rset_no; i++)
- rset_delete(rset[i]);
- return 0;
- }
- if (rset_no == 0)
- return rsnull_create (rset_nmem, key_it_ctrl);
- return rsmulti_or_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
- rset_no, rset);
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ ZEBRA_RES res =
+ term_list_trunc(zh, zapt, termz_org, attributeSet,
+ stream, reg_type, complete_flag,
+ rank_type, xpath_use,
+ num_bases, basenames,
+ rset_nmem,
+ &result_sets, &num_result_sets, kc);
+ if (res != ZEBRA_OK)
+ return res;
+ if (num_result_sets == 0)
+ *rset = rsnull_create (rset_nmem, kc);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rsprox_create(rset_nmem, kc, kc->scope,
+ num_result_sets, result_sets,
+ 1 /* ordered */, 0 /* exclusion */,
+ 3 /* relation */, 1 /* distance */);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
}
-static RSET rpn_search_APT_and_list(ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type,
- int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static ZEBRA_RES rpn_search_APT_or_list(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type,
+ int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
- char term_dst[IT_MAX_WORD+1];
- RSET rset[TERM_LIST_LENGTH_MAX];
- size_t rset_no = 0;
- struct grep_info grep_info;
- char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
- const char *termp = termz;
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ ZEBRA_RES res =
+ term_list_trunc(zh, zapt, termz_org, attributeSet,
+ stream, reg_type, complete_flag,
+ rank_type, xpath_use,
+ num_bases, basenames,
+ rset_nmem,
+ &result_sets, &num_result_sets, kc);
+ if (res != ZEBRA_OK)
+ return res;
+ if (num_result_sets == 0)
+ *rset = rsnull_create (rset_nmem, kc);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rsmulti_or_create(rset_nmem, kc, kc->scope,
+ num_result_sets, result_sets);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
+}
- if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
- return 0;
- for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
- {
- yaz_log(log_level_rpn, "APT_and_list termp=%s", termp);
- 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, rset_nmem);
- if (!rset[rset_no]) /* error or no more terms */
- break;
- }
- grep_info_delete (&grep_info);
- if (termp)
- {
- int i;
- for (i = 0; i<rset_no; i++)
- rset_delete(rset[i]);
- return 0;
- }
- if (rset_no == 0)
- return rsnull_create(rset_nmem, key_it_ctrl);
-
- return rsmulti_and_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
- rset_no, rset);
+static ZEBRA_RES rpn_search_APT_and_list(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type,
+ int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
+{
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ ZEBRA_RES res =
+ term_list_trunc(zh, zapt, termz_org, attributeSet,
+ stream, reg_type, complete_flag,
+ rank_type, xpath_use,
+ num_bases, basenames,
+ rset_nmem,
+ &result_sets, &num_result_sets,
+ kc);
+ if (res != ZEBRA_OK)
+ return res;
+ if (num_result_sets == 0)
+ *rset = rsnull_create (rset_nmem, kc);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rsmulti_and_create(rset_nmem, kc, kc->scope,
+ num_result_sets, result_sets);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
}
static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
char *term_tmp = term_dict + strlen(term_dict);
*error_code = 0;
- attr_init (&relation, zapt, 2);
- relation_value = attr_find (&relation, NULL);
+ attr_init(&relation, zapt, 2);
+ relation_value = attr_find(&relation, NULL);
yaz_log(log_level_rpn, "numeric relation value=%d", relation_value);
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);
+ attr_init(&use, zapt, 1);
+ use_value = attr_find_ex(&use, &curAttributeSet, &use_string);
if (use_value == -1)
use_value = 1016;
curAttributeSet, use_value, r);
if (r == -1)
{
- errCode = 114;
+ errCode = YAZ_BIB1_UNSUPP_USE_ATTRIBUTE;
if (use_string)
errString = nmem_strdup(stream, use_string);
else
errString = nmem_strdup_i (stream, use_value);
}
else
- errCode = 121;
+ errCode = YAZ_BIB1_UNSUPP_ATTRIBUTE_SET;
continue;
}
}
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
- zh->errCode = 109; /* Database unavailable */
+ zh->errCode = YAZ_BIB1_DATABASE_UNAVAILABLE;
zh->errString = basenames[base_no];
return -1;
}
}
if (!prefix_len)
{
- errCode = 114;
+ errCode = YAZ_BIB1_UNSUPP_USE_ATTRIBUTE;
errString = nmem_strdup_i(stream, use_value);
continue;
}
return ZEBRA_OK;
}
-static RSET rpn_search_APT_numeric(ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type, int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type, int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
char term_dst[IT_MAX_WORD+1];
const char *termp = termz;
- RSET rset[TERM_LIST_LENGTH_MAX];
- ZEBRA_RES r;
- size_t rset_no = 0;
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ ZEBRA_RES res;
struct grep_info grep_info;
+ int alloc_sets = 0;
- yaz_log(log_level_rpn, "APT_numeric t='%s'",termz);
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
- for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
+ yaz_log(log_level_rpn, "APT_numeric t='%s'", termz);
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
+ return ZEBRA_FAIL;
+ while (1)
{
+ if (alloc_sets == num_result_sets)
+ {
+ int add = 10;
+ RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) *
+ sizeof(*rnew));
+ if (alloc_sets)
+ memcpy(rnew, result_sets, alloc_sets * sizeof(*rnew));
+ alloc_sets = alloc_sets + add;
+ result_sets = rnew;
+ }
yaz_log(YLOG_DEBUG, "APT_numeric termp=%s", termp);
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,
- stream);
- if (r == ZEBRA_FAIL || termp == 0)
+ res = numeric_term(zh, zapt, &termp, attributeSet, &grep_info,
+ reg_type, complete_flag, num_bases, basenames,
+ term_dst, xpath_use,
+ stream);
+ if (res == ZEBRA_FAIL || termp == 0)
break;
yaz_log(YLOG_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 */,
- zapt->term->which, rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
- if (!rset[rset_no])
+ result_sets[num_result_sets] =
+ rset_trunc(zh, grep_info.isam_p_buf,
+ grep_info.isam_p_indx, term_dst,
+ strlen(term_dst), rank_type,
+ 0 /* preserve position */,
+ zapt->term->which, rset_nmem,
+ kc, kc->scope);
+ if (!result_sets[num_result_sets])
break;
+ num_result_sets++;
}
- grep_info_delete (&grep_info);
+ grep_info_delete(&grep_info);
if (termp)
{
int i;
- for (i = 0; i<rset_no; i++)
- rset_delete(rset[i]);
- return 0;
+ for (i = 0; i<num_result_sets; i++)
+ rset_delete(result_sets[i]);
+ return ZEBRA_FAIL;
}
- if (rset_no == 0)
- return rsnull_create(rset_nmem,key_it_ctrl);
- if (rset_no == 1)
- return rset[0];
- return rsmulti_and_create(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- rset_no, rset);
+ if (num_result_sets == 0)
+ *rset = rsnull_create(rset_nmem, kc);
+ if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rsmulti_and_create(rset_nmem, kc, kc->scope,
+ num_result_sets, result_sets);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
}
-static RSET rpn_search_APT_local (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- const char *termz,
- oid_value attributeSet,
- NMEM stream,
- const char *rank_type, NMEM rset_nmem)
+static ZEBRA_RES rpn_search_APT_local(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz,
+ oid_value attributeSet,
+ NMEM stream,
+ const char *rank_type, NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
- RSET result;
RSFD rsfd;
struct it_key key;
int sys;
- result = rstemp_create( rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- res_get (zh->res, "setTmpDir"),0 );
- rsfd = rset_open (result, RSETF_WRITE);
-
+ *rset = rstemp_create(rset_nmem, kc, kc->scope,
+ res_get (zh->res, "setTmpDir"),0 );
+ rsfd = rset_open(*rset, RSETF_WRITE);
+
sys = atoi(termz);
if (sys <= 0)
sys = 1;
key.len = 2;
rset_write (rsfd, &key);
rset_close (rsfd);
- return result;
+ return ZEBRA_OK;
}
-static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- oid_value attributeSet, NMEM stream,
- Z_SortKeySpecList *sort_sequence,
- const char *rank_type)
+static ZEBRA_RES rpn_sort_spec(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ oid_value attributeSet, NMEM stream,
+ Z_SortKeySpecList *sort_sequence,
+ const char *rank_type,
+ RSET *rset,
+ struct rset_key_control *kc)
{
int i;
int sort_relation_value;
oident oe;
char termz[20];
- attr_init (&sort_relation_type, zapt, 7);
- sort_relation_value = attr_find (&sort_relation_type, &attributeSet);
+ attr_init(&sort_relation_type, zapt, 7);
+ sort_relation_value = attr_find(&sort_relation_type, &attributeSet);
- attr_init (&use_type, zapt, 1);
- use_value = attr_find (&use_type, &attributeSet);
+ attr_init(&use_type, zapt, 1);
+ use_value = attr_find(&use_type, &attributeSet);
if (!sort_sequence->specs)
{
zapt->term->u.general->len);
if (i >= sort_sequence->num_specs)
i = 0;
- sprintf (termz, "%d", i);
+ sprintf(termz, "%d", i);
oe.proto = PROTO_Z3950;
oe.oclass = CLASS_ATTSET;
oe.value = attributeSet;
if (!oid_ent_to_oid (&oe, oid))
- return 0;
+ return ZEBRA_FAIL;
sks = (Z_SortKeySpec *) nmem_malloc(stream, sizeof(*sks));
sks->sortElement = (Z_SortElement *)
sks->which = Z_SortKeySpec_null;
sks->u.null = odr_nullval ();
sort_sequence->specs[i] = sks;
- return rsnull_create (NULL,key_it_ctrl);
- /* FIXME - nmem?? */
+ *rset = rsnull_create (NULL, kc);
+ return ZEBRA_OK;
}
AttrType use;
const char *use_string = 0;
- attr_init (&use, zapt, 1);
- attr_find_ex (&use, &curAttributeSet, &use_string);
+ attr_init(&use, zapt, 1);
+ attr_find_ex(&use, &curAttributeSet, &use_string);
if (!use_string || *use_string != '/')
return -1;
static RSET xpath_trunc(ZebraHandle zh, NMEM stream,
int reg_type, const char *term, int use,
- oid_value curAttributeSet, NMEM rset_nmem)
+ oid_value curAttributeSet, NMEM rset_nmem,
+ struct rset_key_control *kc)
{
RSET rset;
struct grep_info grep_info;
int term_type = Z_Term_characterString;
const char *flags = "void";
- if (grep_info_prepare (zh, 0 /* zapt */, &grep_info, '0', stream))
- return rsnull_create (rset_nmem,key_it_ctrl);
-
+ if (grep_info_prepare(zh, 0 /* zapt */, &grep_info, '0', stream))
+ return rsnull_create(rset_nmem, kc);
+
if (ord < 0)
- return rsnull_create (rset_nmem,key_it_ctrl);
+ return rsnull_create(rset_nmem, kc);
if (prefix_len)
term_dict[prefix_len++] = '|';
else
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 (YLOG_DEBUG, "%s %d positions", term,
+ yaz_log(YLOG_DEBUG, "%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,rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
- grep_info_delete (&grep_info);
+ kc, kc->scope);
+ 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,
- NMEM rset_nmem)
+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,
+ NMEM rset_nmem,
+ struct rset_key_control *kc)
{
oid_value curAttributeSet = attributeSet;
int base_no;
if (xpath_len < 0)
return rset;
- yaz_log (YLOG_DEBUG, "xpath len=%d", xpath_len);
+ yaz_log(YLOG_DEBUG, "xpath len=%d", xpath_len);
for (i = 0; i<xpath_len; i++)
{
- yaz_log (log_level_rpn, "XPATH %d %s", i, xpath[i].part);
+ yaz_log(log_level_rpn, "XPATH %d %s", i, xpath[i].part);
}
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
- zh->errCode = 109; /* Database unavailable */
+ zh->errCode = YAZ_BIB1_DATABASE_UNAVAILABLE;
zh->errString = basenames[base_no];
return rset;
}
wrbuf_puts(wbuf, "");
rset_attr = xpath_trunc(
zh, stream, '0', wrbuf_buf(wbuf), 3,
- curAttributeSet,rset_nmem);
+ curAttributeSet, rset_nmem, kc);
wrbuf_free(wbuf, 1);
}
else
if (!first_path)
continue;
}
- yaz_log (log_level_rpn, "xpath_rev (%d) = %s", level, xpath_rev);
+ yaz_log(log_level_rpn, "xpath_rev (%d) = %s", level, xpath_rev);
if (strlen(xpath_rev))
{
rset_start_tag = xpath_trunc(zh, stream, '0',
- xpath_rev, 1, curAttributeSet, rset_nmem);
+ xpath_rev, 1, curAttributeSet, rset_nmem, kc);
rset_end_tag = xpath_trunc(zh, stream, '0',
- xpath_rev, 2, curAttributeSet, rset_nmem);
+ xpath_rev, 2, curAttributeSet, rset_nmem, kc);
- rset = rsbetween_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope,
+ rset = rsbetween_create(rset_nmem, kc, kc->scope,
rset_start_tag, rset,
rset_end_tag, rset_attr);
}
return rset;
}
-
-
static ZEBRA_RES rpn_search_APT(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
oid_value attributeSet, NMEM stream,
Z_SortKeySpecList *sort_sequence,
int num_bases, char **basenames,
NMEM rset_nmem,
- RSET *rset)
+ RSET *rset,
+ struct rset_key_control *kc)
{
+ ZEBRA_RES res = ZEBRA_OK;
unsigned reg_id;
char *search_type = NULL;
char rank_type[128];
yaz_log(YLOG_DEBUG, "search_type=%s", search_type);
yaz_log(YLOG_DEBUG, "rank_type=%s", rank_type);
- if (zapt_term_to_utf8(zh, zapt, termz))
+ if (zapt_term_to_utf8(zh, zapt, termz) == ZEBRA_FAIL)
return ZEBRA_FAIL;
if (sort_flag)
- {
- *rset = rpn_sort_spec(zh, zapt, attributeSet, stream, sort_sequence,
- rank_type);
- if (!*rset)
- return ZEBRA_FAIL;
- return ZEBRA_OK;
- }
+ return rpn_sort_spec(zh, zapt, attributeSet, stream, sort_sequence,
+ rank_type, rset, kc);
xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream);
if (xpath_len >= 0)
{
xpath_use = 1015;
}
- if (!strcmp (search_type, "phrase"))
+ 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, rset_nmem);
+ res = rpn_search_APT_phrase(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem,
+ rset, kc);
}
- else if (!strcmp (search_type, "and-list"))
+ 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, rset_nmem);
+ res = rpn_search_APT_and_list(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem,
+ rset, kc);
}
- else if (!strcmp (search_type, "or-list"))
+ 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, rset_nmem);
+ res = rpn_search_APT_or_list(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem,
+ rset, kc);
}
- else if (!strcmp (search_type, "local"))
+ else if (!strcmp(search_type, "local"))
{
- *rset = rpn_search_APT_local(zh, zapt, termz, attributeSet, stream,
- rank_type, rset_nmem);
+ res = rpn_search_APT_local(zh, zapt, termz, attributeSet, stream,
+ rank_type, rset_nmem, rset, kc);
}
- else if (!strcmp (search_type, "numeric"))
+ else if (!strcmp(search_type, "numeric"))
{
- *rset = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, stream,
- reg_id, complete_flag, rank_type,
- xpath_use,
- num_bases, basenames, rset_nmem);
+ res = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem,
+ rset, kc);
}
else
{
- zh->errCode = 118;
+ zh->errCode = YAZ_BIB1_UNSUPP_STRUCTURE_ATTRIBUTE;
return ZEBRA_FAIL;
}
+ if (res != ZEBRA_OK)
+ return res;
if (!*rset)
return ZEBRA_FAIL;
- *rset = rpn_search_xpath (zh, attributeSet, num_bases, basenames,
- stream, rank_type, *rset,
- xpath_len, xpath, rset_nmem);
+ *rset = rpn_search_xpath(zh, attributeSet, num_bases, basenames,
+ stream, rank_type, *rset,
+ xpath_len, xpath, rset_nmem, kc);
if (!*rset)
return ZEBRA_FAIL;
return ZEBRA_OK;
Z_SortKeySpecList *sort_sequence,
int num_bases, char **basenames,
RSET **result_sets, int *num_result_sets,
- Z_Operator *parent_op);
+ Z_Operator *parent_op,
+ struct rset_key_control *kc);
ZEBRA_RES rpn_search_top(ZebraHandle zh, Z_RPNStructure *zs,
oid_value attributeSet,
{
RSET *result_sets = 0;
int num_result_sets = 0;
- ZEBRA_RES res = rpn_search_structure(zh, zs, attributeSet,
- stream, rset_nmem,
- sort_sequence,
- num_bases, basenames,
- &result_sets, &num_result_sets,
- 0 /* no op */);
+ ZEBRA_RES res;
+ struct rset_key_control *kc = zebra_key_control_create(zh);
+
+ res = rpn_search_structure(zh, zs, attributeSet,
+ stream, rset_nmem,
+ sort_sequence,
+ num_bases, basenames,
+ &result_sets, &num_result_sets,
+ 0 /* no parent op */,
+ kc);
if (res != ZEBRA_OK)
{
int i;
assert(result_sets);
assert(*result_sets);
*result_set = *result_sets;
+
+ (*kc->dec)(kc);
return ZEBRA_OK;
}
Z_SortKeySpecList *sort_sequence,
int num_bases, char **basenames,
RSET **result_sets, int *num_result_sets,
- Z_Operator *parent_op)
+ Z_Operator *parent_op,
+ struct rset_key_control *kc)
{
*num_result_sets = 0;
if (zs->which == Z_RPNStructure_complex)
sort_sequence,
num_bases, basenames,
&result_sets_l, &num_result_sets_l,
- zop);
+ zop, kc);
if (res != ZEBRA_OK)
{
int i;
sort_sequence,
num_bases, basenames,
&result_sets_r, &num_result_sets_r,
- zop);
+ zop, kc);
if (res != ZEBRA_OK)
{
int i;
switch (zop->which)
{
case Z_Operator_and:
- rset = rsmulti_and_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope,
+ rset = rsmulti_and_create(rset_nmem, kc,
+ kc->scope,
*num_result_sets, *result_sets);
break;
case Z_Operator_or:
- rset = rsmulti_or_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope,
+ rset = rsmulti_or_create(rset_nmem, kc,
+ kc->scope,
*num_result_sets, *result_sets);
break;
case Z_Operator_and_not:
- rset = rsbool_create_not(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope,
+ rset = rsbool_create_not(rset_nmem, kc,
+ kc->scope,
(*result_sets)[0],
(*result_sets)[1]);
break;
case Z_Operator_prox:
if (zop->u.prox->which != Z_ProximityOperator_known)
{
- zh->errCode = 132;
+ zh->errCode = YAZ_BIB1_UNSUPP_PROX_UNIT_CODE;
return ZEBRA_FAIL;
}
if (*zop->u.prox->u.known != Z_ProxUnit_word)
{
char *val = (char *) nmem_malloc(stream, 16);
- zh->errCode = 132;
+ zh->errCode = YAZ_BIB1_UNSUPP_PROX_UNIT_CODE;
zh->errString = val;
- sprintf (val, "%d", *zop->u.prox->u.known);
+ sprintf(val, "%d", *zop->u.prox->u.known);
return ZEBRA_FAIL;
}
else
{
- rset = rsprox_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope,
+ rset = rsprox_create(rset_nmem, kc,
+ kc->scope,
*num_result_sets, *result_sets,
*zop->u.prox->ordered,
(!zop->u.prox->exclusion ?
}
break;
default:
- zh->errCode = 110;
+ zh->errCode = YAZ_BIB1_OPERATOR_UNSUPP;
return ZEBRA_FAIL;
}
*num_result_sets = 1;
yaz_log(YLOG_DEBUG, "rpn_search_APT");
res = rpn_search_APT(zh, zs->u.simple->u.attributesPlusTerm,
attributeSet, stream, sort_sequence,
- num_bases, basenames, rset_nmem, &rset);
+ num_bases, basenames, rset_nmem, &rset,
+ kc);
if (res != ZEBRA_OK)
return res;
}
rset = resultSetRef(zh, zs->u.simple->u.resultSetId);
if (!rset)
{
- zh->errCode = 30;
+ zh->errCode = YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST;
zh->errString =
nmem_strdup(stream, zs->u.simple->u.resultSetId);
return ZEBRA_FAIL;
}
else
{
- zh->errCode = 3;
+ zh->errCode = YAZ_BIB1_UNSUPP_SEARCH;
return ZEBRA_FAIL;
}
*num_result_sets = 1;
}
else
{
- zh->errCode = 3;
+ zh->errCode = YAZ_BIB1_UNSUPP_SEARCH;
return ZEBRA_FAIL;
}
return ZEBRA_OK;
len_prefix = strlen(scan_info->prefix);
if (memcmp (name, scan_info->prefix, len_prefix))
return 1;
- if (pos > 0) idx = scan_info->after - pos + scan_info->before;
+ if (pos > 0)
+ idx = scan_info->after - pos + scan_info->before;
else
idx = - pos - 1;
+
+ if (idx < 0)
+ return 0;
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);
int complete_flag;
int sort_flag;
NMEM rset_nmem = NULL;
+ struct rset_key_control *kc = 0;
*list = 0;
*is_partial = 0;
AttrType termset;
int termset_value_numeric;
const char *termset_value_string;
- attr_init (&termset, zapt, 8);
+ attr_init(&termset, zapt, 8);
termset_value_numeric =
- attr_find_ex (&termset, NULL, &termset_value_string);
+ attr_find_ex(&termset, NULL, &termset_value_string);
if (termset_value_numeric != -1)
{
char resname[32];
if (termset_value_numeric != -2)
{
- sprintf (resname, "%d", termset_value_numeric);
+ sprintf(resname, "%d", termset_value_numeric);
termset_name = resname;
}
else
}
}
- yaz_log (YLOG_DEBUG, "position = %d, num = %d set=%d",
- pos, num, attributeset);
+ yaz_log(YLOG_DEBUG, "position = %d, num = %d set=%d",
+ pos, num, attributeset);
- attr_init (&use, zapt, 1);
- use_value = attr_find_ex (&use, &attributeset, &use_string);
+ attr_init(&use, zapt, 1);
+ 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))
+ if (zebra_maps_attr(zh->reg->zebra_maps, zapt, ®_id, &search_type,
+ rank_type, &complete_flag, &sort_flag))
{
*num_entries = 0;
- zh->errCode = 113;
+ zh->errCode = YAZ_BIB1_UNSUPP_ATTRIBUTE_TYPE;
return ZEBRA_FAIL;
}
- yaz_log (YLOG_DEBUG, "use_value = %d", use_value);
+ yaz_log(YLOG_DEBUG, "use_value = %d", use_value);
if (use_value == -1)
use_value = 1016;
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
zh->errString = basenames[base_no];
- zh->errCode = 109; /* Database unavailable */
+ zh->errCode = YAZ_BIB1_DATABASE_UNAVAILABLE;
*num_entries = 0;
return ZEBRA_FAIL;
}
attributeset, use_value);
if (r == -1)
{
- errCode = 114;
+ errCode = YAZ_BIB1_UNSUPP_USE_ATTRIBUTE;
if (use_string)
errString = odr_strdup(stream, use_string);
else
{
char val_str[32];
- sprintf (val_str, "%d", use_value);
+ sprintf(val_str, "%d", use_value);
errString = odr_strdup(stream, val_str);
}
}
else
- errCode = 121;
+ errCode = YAZ_BIB1_UNSUPP_ATTRIBUTE_SET;
continue;
}
}
return ZEBRA_OK;
}
/* prepare dictionary scanning */
+ if (num < 1)
+ {
+ *num_entries = 0;
+ return ZEBRA_OK;
+ }
before = pos-1;
+ if (before < 0)
+ before = 0;
after = 1+num-pos;
+ if (after < 0)
+ after = 0;
+ yaz_log(YLOG_DEBUG, "rpn_scan pos=%d num=%d before=%d "
+ "after=%d before+after=%d",
+ pos, num, before, after, before+after);
scan_info_array = (struct scan_info *)
odr_malloc(stream, ord_no * sizeof(*scan_info_array));
for (i = 0; i < ord_no; i++)
termz[prefix_len] = 0;
strcpy(scan_info->prefix, termz);
- if (trans_scan_term(zh, zapt, termz+prefix_len, reg_id))
+ if (trans_scan_term(zh, zapt, termz+prefix_len, reg_id) == ZEBRA_FAIL)
return ZEBRA_FAIL;
dict_scan(zh->reg->dict, termz, &before_tmp, &after_tmp,
odr_malloc(stream, (before+after)*sizeof(*glist));
rset_nmem = nmem_create();
+ kc = zebra_key_control_create(zh);
/* consider terms after main term */
for (i = 0; i < ord_no; i++)
int j, j0 = -1;
const char *mterm = NULL;
const char *tst;
- RSET rset;
-
+ RSET rset = 0;
+ int lo = i + pos-1; /* offset in result list */
+
+ /* find: j0 is the first of the minimal values */
for (j = 0; j < ord_no; j++)
{
- if (ptr[j] < before+after &&
+ if (ptr[j] < before+after && ptr[j] >= 0 &&
(tst = scan_info_array[j].list[ptr[j]].term) &&
(!mterm || strcmp (tst, mterm) < 0))
{
}
}
if (j0 == -1)
- break;
- scan_term_untrans (zh, stream->mem, reg_id,
- &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, zapt->term->which, rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
- ptr[j0]++;
+ break; /* no value found, stop */
+
+ /* get result set for first one , but only if it's within bounds */
+ if (lo >= 0)
+ {
+ /* get result set for first term */
+ scan_term_untrans(zh, stream->mem, reg_id,
+ &glist[lo].term, mterm);
+ rset = rset_trunc(zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
+ glist[lo].term, strlen(glist[lo].term),
+ NULL, 0, zapt->term->which, rset_nmem,
+ kc, kc->scope);
+ }
+ ptr[j0]++; /* move index for this set .. */
+ /* get result set for remaining scan terms */
for (j = j0+1; j<ord_no; j++)
{
- if (ptr[j] < before+after &&
+ if (ptr[j] < before+after && ptr[j] >= 0 &&
(tst = scan_info_array[j].list[ptr[j]].term) &&
!strcmp (tst, mterm))
{
- RSET rsets[2];
-
- rsets[0] = rset;
- rsets[1] =
- rset_trunc(zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
- glist[i+before].term,
- strlen(glist[i+before].term), NULL, 0,
- zapt->term->which,rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
- rset = rsmulti_or_create(rset_nmem, key_it_ctrl,
- 2, key_it_ctrl->scope, rsets);
+ if (lo >= 0)
+ {
+ RSET rsets[2];
+
+ rsets[0] = rset;
+ rsets[1] =
+ rset_trunc(
+ zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
+ glist[lo].term,
+ strlen(glist[lo].term), NULL, 0,
+ zapt->term->which,rset_nmem,
+ kc, kc->scope);
+ rset = rsmulti_or_create(rset_nmem, kc,
+ 2, kc->scope, rsets);
+ }
ptr[j]++;
}
}
- if (limit_set)
+ if (lo >= 0)
{
- RSET rsets[2];
- rsets[0] = rset;
- rsets[1] = rset_dup(limit_set);
-
- rset = rsmulti_and_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope, 2, rsets);
+ /* merge with limit_set if given */
+ if (limit_set)
+ {
+ RSET rsets[2];
+ rsets[0] = rset;
+ rsets[1] = rset_dup(limit_set);
+
+ rset = rsmulti_and_create(rset_nmem, kc,
+ kc->scope, 2, rsets);
+ }
+ /* count it */
+ count_set(rset, &glist[lo].occurrences);
+ rset_delete(rset);
}
- count_set(rset, &glist[i+before].occurrences);
- rset_delete(rset);
}
if (i < after)
{
*num_entries -= (after-i);
*is_partial = 1;
+ if (*num_entries < 0)
+ {
+ (*kc->dec)(kc);
+ nmem_destroy(rset_nmem);
+ *num_entries = 0;
+ return ZEBRA_OK;
+ }
}
-
/* consider terms before main term */
for (i = 0; i<ord_no; i++)
ptr[i] = 0;
const char *mterm = NULL;
const char *tst;
RSET rset;
+ int lo = before-1-i; /* offset in result list */
for (j = 0; j <ord_no; j++)
{
- if (ptr[j] < before &&
+ if (ptr[j] < before && ptr[j] >= 0 &&
(tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
(!mterm || strcmp (tst, mterm) > 0))
{
break;
scan_term_untrans (zh, stream->mem, reg_id,
- &glist[before-1-i].term, mterm);
+ &glist[lo].term, mterm);
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),
+ glist[lo].term, strlen(glist[lo].term),
NULL, 0, zapt->term->which,rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
+ kc, kc->scope);
ptr[j0]++;
for (j = j0+1; j<ord_no; j++)
{
- if (ptr[j] < before &&
+ if (ptr[j] < before && ptr[j] >= 0 &&
(tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
!strcmp (tst, mterm))
{
rsets[1] = 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,
+ glist[lo].term,
+ strlen(glist[lo].term), NULL, 0,
zapt->term->which, rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
- rset = rsmulti_or_create(rset_nmem, key_it_ctrl,
- 2, key_it_ctrl->scope, rsets);
+ kc, kc->scope);
+ rset = rsmulti_or_create(rset_nmem, kc,
+ 2, kc->scope, rsets);
ptr[j]++;
}
rsets[0] = rset;
rsets[1] = rset_dup(limit_set);
- rset = rsmulti_and_create(rset_nmem, key_it_ctrl,
- key_it_ctrl->scope, 2, rsets);
+ rset = rsmulti_and_create(rset_nmem, kc,
+ kc->scope, 2, rsets);
}
- count_set (rset, &glist[before-1-i].occurrences);
+ count_set (rset, &glist[lo].occurrences);
rset_delete (rset);
}
+ (*kc->dec)(kc);
+ nmem_destroy(rset_nmem);
i = before-i;
if (i)
{
*is_partial = 1;
*position -= i;
*num_entries -= i;
+ if (*num_entries <= 0)
+ {
+ *num_entries = 0;
+ return ZEBRA_OK;
+ }
}
- nmem_destroy(rset_nmem);
*list = glist + i; /* list is set to first 'real' entry */
yaz_log(YLOG_DEBUG, "position = %d, num_entries = %d",
projects / idzebra-moved-to-github.git / blobdiff