-/* $Id: zrpn.c,v 1.146 2004-08-30 12:32:24 heikki Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
- Index Data Aps
+/* $Id: zrpn.c,v 1.217 2006-06-07 10:50:08 adam Exp $
+ Copyright (C) 1995-2006
+ Index Data ApS
This file is part of the Zebra server.
02111-1307, USA.
*/
-
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#include <io.h>
-#else
+#endif
+#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <ctype.h>
+#include <yaz/diagbib1.h>
#include "index.h"
#include <zebra_xpath.h>
-
+#include <attrfind.h>
#include <charmap.h>
-#include <rstemp.h>
-#include <rsnull.h>
-#include <rsbool.h>
-#include <rsmultior.h>
-#include <rsbetween.h>
-#include <rsprox.h>
-
-struct rpn_char_map_info {
+#include <rset.h>
+
+struct rpn_char_map_info
+{
ZebraMaps zm;
int reg_type;
};
-typedef struct {
- int type;
- int major;
- int minor;
- Z_AttributesPlusTerm *zapt;
-} AttrType;
-
+static int log_level_set = 0;
+static int log_level_rpn = 0;
-static const char **rpn_char_map_handler (void *vp, const char **from, int len)
+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;
- const char **out = zebra_maps_input (p->zm, p->reg_type, from, len);
+ const char **out = zebra_maps_input(p->zm, p->reg_type, from, len, 0);
#if 0
if (out && *out)
{
const char *outp = *out;
- yaz_log (LOG_LOG, "---");
+ yaz_log(YLOG_LOG, "---");
while (*outp)
{
- yaz_log (LOG_LOG, "%02X", *outp);
+ yaz_log(YLOG_LOG, "%02X", *outp);
outp++;
}
}
return out;
}
-static void rpn_char_map_prepare (struct zebra_register *reg, int reg_type,
+static void rpn_char_map_prepare(struct zebra_register *reg, int reg_type,
struct rpn_char_map_info *map_info)
{
map_info->zm = reg->zebra_maps;
map_info->reg_type = reg_type;
- dict_grep_cmap (reg->dict, map_info, rpn_char_map_handler);
-}
-
-static int attr_find_ex (AttrType *src, oid_value *attributeSetP,
- const char **string_value)
-{
- int num_attributes;
-
- num_attributes = src->zapt->attributes->num_attributes;
- while (src->major < num_attributes)
- {
- Z_AttributeElement *element;
-
- element = src->zapt->attributes->attributes[src->major];
- if (src->type == *element->attributeType)
- {
- switch (element->which)
- {
- case Z_AttributeValue_numeric:
- ++(src->major);
- if (element->attributeSet && attributeSetP)
- {
- oident *attrset;
-
- attrset = oid_getentbyoid (element->attributeSet);
- *attributeSetP = attrset->value;
- }
- return *element->value.numeric;
- break;
- 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);
- return
- *element->value.complex->list[src->minor-1]->u.numeric;
- }
- else if (element->value.complex->list[src->minor]->which ==
- Z_StringOrNumeric_string)
- {
- if (!string_value)
- break;
- ++(src->minor);
- *string_value =
- element->value.complex->list[src->minor-1]->u.string;
- return -2;
- }
- else
- break;
- default:
- assert (0);
- }
- }
- ++(src->major);
- }
- return -1;
-}
-
-static int attr_find (AttrType *src, oid_value *attributeSetP)
-{
- return attr_find_ex (src, attributeSetP, 0);
-}
-
-static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt,
- int type)
-{
- src->zapt = zapt;
- src->type = type;
- src->major = 0;
- src->minor = 0;
+ dict_grep_cmap(reg->dict, map_info, rpn_char_map_handler);
}
#define TERM_COUNT
#ifdef TERM_COUNT
int *term_no;
#endif
- ISAMC_P *isam_p_buf;
+ ISAM_P *isam_p_buf;
int isam_p_size;
int isam_p_indx;
ZebraHandle zh;
ZebraSet termset;
};
-static void term_untrans (ZebraHandle zh, int reg_type,
- char *dst, const char *src)
+void zebra_term_untrans(ZebraHandle zh, int reg_type,
+ char *dst, const char *src)
{
int len = 0;
while (*src)
{
- const char *cp = zebra_maps_output (zh->reg->zebra_maps,
- reg_type, &src);
- if (!cp && len < IT_MAX_WORD-1)
- dst[len++] = *src++;
+ const char *cp = zebra_maps_output(zh->reg->zebra_maps,
+ reg_type, &src);
+ if (!cp)
+ {
+ if (len < IT_MAX_WORD-1)
+ dst[len++] = *src;
+ src++;
+ }
else
while (*cp && len < IT_MAX_WORD-1)
dst[len++] = *cp++;
dst[len] = '\0';
}
-static void add_isam_p (const char *name, const char *info,
- struct grep_info *p)
+static void add_isam_p(const char *name, const char *info,
+ struct grep_info *p)
{
+ if (!log_level_set)
+ {
+ log_level_rpn = yaz_log_module_level("rpn");
+ log_level_set = 1;
+ }
if (p->isam_p_indx == p->isam_p_size)
{
- ISAMC_P *new_isam_p_buf;
+ ISAM_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 = (ISAMC_P *) xmalloc (sizeof(*new_isam_p_buf) *
- p->isam_p_size);
+ new_isam_p_buf = (ISAM_P *) xmalloc(sizeof(*new_isam_p_buf) *
+ p->isam_p_size);
if (p->isam_p_buf)
{
- memcpy (new_isam_p_buf, p->isam_p_buf,
+ memcpy(new_isam_p_buf, p->isam_p_buf,
p->isam_p_indx * sizeof(*p->isam_p_buf));
- xfree (p->isam_p_buf);
+ xfree(p->isam_p_buf);
}
p->isam_p_buf = new_isam_p_buf;
#ifdef TERM_COUNT
- new_term_no = (int *) xmalloc (sizeof(*new_term_no) *
- p->isam_p_size);
+ new_term_no = (int *) xmalloc(sizeof(*new_term_no) * p->isam_p_size);
if (p->term_no)
{
- memcpy (new_term_no, p->isam_p_buf,
+ memcpy(new_term_no, p->isam_p_buf,
p->isam_p_indx * sizeof(*p->term_no));
- xfree (p->term_no);
+ xfree(p->term_no);
}
p->term_no = new_term_no;
#endif
}
- assert (*info == sizeof(*p->isam_p_buf));
- memcpy (p->isam_p_buf + p->isam_p_indx, info+1, sizeof(*p->isam_p_buf));
+ assert(*info == sizeof(*p->isam_p_buf));
+ memcpy(p->isam_p_buf + p->isam_p_indx, info+1, sizeof(*p->isam_p_buf));
#if 1
if (p->termset)
{
const char *db;
- int set, use;
char term_tmp[IT_MAX_WORD];
- int su_code = 0;
- int len = key_SU_decode (&su_code, name);
+ int ord = 0;
+ const char *index_name;
+ int len = key_SU_decode (&ord, (const unsigned char *) name);
- term_untrans (p->zh, p->reg_type, term_tmp, name+len+1);
- logf (LOG_LOG, "grep: %d %c %s", su_code, name[len], term_tmp);
- zebraExplain_lookup_ord (p->zh->reg->zei,
- su_code, &db, &set, &use);
- logf (LOG_LOG, "grep: set=%d use=%d db=%s", set, use, db);
+ zebra_term_untrans (p->zh, p->reg_type, term_tmp, name+len+1);
+ yaz_log(log_level_rpn, "grep: %d %c %s", ord, name[len], term_tmp);
+ zebraExplain_lookup_ord(p->zh->reg->zei,
+ ord, 0 /* index_type */, &db, &index_name);
+ yaz_log(log_level_rpn, "grep: db=%s index=%s", db, index_name);
- resultSetAddTerm (p->zh, p->termset, name[len], db,
- set, use, term_tmp);
+ resultSetAddTerm(p->zh, p->termset, name[len], db,
+ index_name, term_tmp);
}
#endif
(p->isam_p_indx)++;
}
-static int grep_handle (char *name, const char *info, void *p)
+static int grep_handle(char *name, const char *info, void *p)
{
- add_isam_p (name, info, (struct grep_info *) p);
+ add_isam_p(name, info, (struct grep_info *) p);
return 0;
}
-static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src,
- const char *ct1, const char *ct2)
+static int term_pre(ZebraMaps zebra_maps, int reg_type, const char **src,
+ const char *ct1, const char *ct2, int first)
{
const char *s1, *s0 = *src;
const char **map;
/* skip white space */
while (*s0)
{
- if (ct1 && strchr (ct1, *s0))
+ if (ct1 && strchr(ct1, *s0))
break;
- if (ct2 && strchr (ct2, *s0))
+ if (ct2 && strchr(ct2, *s0))
break;
s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s1, strlen(s1));
+ map = zebra_maps_input(zebra_maps, reg_type, &s1, strlen(s1), first);
if (**map != *CHR_SPACE)
break;
s0 = s1;
return *s0;
}
+
+static void esc_str(char *out_buf, size_t out_size,
+ const char *in_buf, int in_size)
+{
+ int k;
+
+ assert(out_buf);
+ assert(in_buf);
+ assert(out_size > 20);
+ *out_buf = '\0';
+ for (k = 0; k<in_size; k++)
+ {
+ int c = in_buf[k] & 0xff;
+ int pc;
+ if (c < 32 || c > 126)
+ pc = '?';
+ else
+ pc = c;
+ sprintf(out_buf +strlen(out_buf), "%02X:%c ", c, pc);
+ if (strlen(out_buf) > out_size-20)
+ {
+ strcat(out_buf, "..");
+ break;
+ }
+ }
+}
+
#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,
- char *dst_term)
+/* 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 *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
const char *space_start = 0;
const char *space_end = 0;
- if (!term_pre (zebra_maps, reg_type, src, NULL, NULL))
+ if (!term_pre(zebra_maps, reg_type, src, NULL, NULL, !space_split))
return 0;
s0 = *src;
while (*s0)
{
- s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ 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)
{
if (**map == *CHR_SPACE)
{ /* reload last space */
while (space_start < space_end)
{
- if (strchr (REGEX_CHARS, *space_start))
+ if (strchr(REGEX_CHARS, *space_start))
dst[i++] = '\\';
dst_term[j++] = *space_start;
dst[i++] = *space_start++;
space_start = space_end = 0;
}
}
- /* add non-space char */
- 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';
dst_term[j] = '\0';
return i;
}
-/* 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)
+/* 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 *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "#", "#"))
+ if (!term_pre(zebra_maps, reg_type, src, "#", "#", !space_split))
return 0;
s0 = *src;
while (*s0)
}
else
{
- s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ 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';
return i;
}
-/* 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)
+/* 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)
{
int i = 0;
int j = 0;
- const char *s0, *s1;
+ const char *s0;
const char **map;
- if (!term_pre (zebra_maps, reg_type, src, "^\\()[].*+?|", "("))
+ if (!term_pre(zebra_maps, reg_type, src, "^\\()[].*+?|", "(", !space_split))
return 0;
s0 = *src;
if (errors && *s0 == '+' && s0[1] && s0[2] == '+' && s0[3] &&
- isdigit (s0[1]))
+ isdigit(((const unsigned char *)s0)[1]))
{
*errors = s0[1] - '0';
s0 += 3;
}
while (*s0)
{
- if (strchr ("^\\()[].*+?|-", *s0))
+ if (strchr("^\\()[].*+?|-", *s0))
{
dst_term[j++] = *s0;
dst[i++] = *s0++;
}
else
{
- s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
- if (**map == *CHR_SPACE)
+ 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';
dst_term[j] = '\0';
*src = s0;
+
return i;
}
-/* 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)
+/* 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)
{
- return term_103 (zebra_maps, reg_type, src, dst, NULL, space_split,
- dst_term);
+ return term_103(zebra_maps, reg_type, src, dst, NULL, space_split,
+ 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)
+/* 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 *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "?*#", "?*#"))
+ if (!term_pre(zebra_maps, reg_type, src, "?*#", "?*#", !space_split))
return 0;
s0 = *src;
while (*s0)
dst[i++] = '.';
dst_term[j++] = *s0++;
}
+ else
{
- s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ 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';
return i;
}
-/* 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)
+/* 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)
{
- const char *s0, *s1;
+ const char *s0;
const char **map;
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "*!", "*!"))
+ if (!term_pre(zebra_maps, reg_type, src, "*!", "*!", !space_split))
return 0;
s0 = *src;
while (*s0)
dst[i++] = '.';
dst_term[j++] = *s0++;
}
+ else
{
- s1 = s0;
- map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ 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)
* val: border value (inclusive)
* islt: 1 if <=; 0 if >=.
*/
-static void gen_regular_rel (char *dst, int val, int islt)
+static void gen_regular_rel(char *dst, int val, int islt)
{
int dst_p;
int w, d, i;
int pos = 0;
char numstr[20];
- logf (LOG_DEBUG, "gen_regular_rel. val=%d, islt=%d", val, islt);
+ yaz_log(YLOG_DEBUG, "gen_regular_rel. val=%d, islt=%d", val, islt);
if (val >= 0)
{
if (islt)
- strcpy (dst, "(-[0-9]+|(");
+ strcpy(dst, "(-[0-9]+|(");
else
- strcpy (dst, "((");
+ strcpy(dst, "((");
}
else
{
if (!islt)
{
- strcpy (dst, "([0-9]+|-(");
- dst_p = strlen (dst);
+ strcpy(dst, "([0-9]+|-(");
+ dst_p = strlen(dst);
islt = 1;
}
else
{
- strcpy (dst, "(-(");
+ strcpy(dst, "(-(");
islt = 0;
}
val = -val;
}
- dst_p = strlen (dst);
- sprintf (numstr, "%d", val);
+ dst_p = strlen(dst);
+ sprintf(numstr, "%d", val);
for (w = strlen(numstr); --w >= 0; pos++)
{
d = numstr[w];
}
}
- strcpy (dst + dst_p, numstr);
+ strcpy(dst + dst_p, numstr);
dst_p = strlen(dst) - pos - 1;
if (islt)
if (islt)
{
/* match everything less than 10^(pos-1) */
- strcat (dst, "0*");
- for (i=1; i<pos; i++)
- strcat (dst, "[0-9]?");
+ strcat(dst, "0*");
+ for (i = 1; i<pos; i++)
+ strcat(dst, "[0-9]?");
}
else
{
/* match everything greater than 10^pos */
for (i = 0; i <= pos; i++)
- strcat (dst, "[0-9]");
- strcat (dst, "[0-9]*");
+ strcat(dst, "[0-9]");
+ strcat(dst, "[0-9]*");
}
- strcat (dst, "))");
+ strcat(dst, "))");
}
-void string_rel_add_char (char **term_p, const char *src, int *indx)
+void string_rel_add_char(char **term_p, const char *src, int *indx)
{
if (src[*indx] == '\\')
*(*term_p)++ = src[(*indx)++];
* <= abc ([-0].*|a[-a].*|ab[-b].*|abc)
* ([^a-].*|a[^b-].*|ab[^c-].*|abc)
*/
-static int string_relation (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- const char **term_sub, char *term_dict,
- oid_value attributeSet,
- int reg_type, int space_split, char *term_dst)
+static int string_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub, char *term_dict,
+ oid_value attributeSet,
+ int reg_type, int space_split, char *term_dst,
+ int *error_code)
{
AttrType relation;
int relation_value;
char *term_tmp = term_dict + strlen(term_dict);
char term_component[2*IT_MAX_WORD+20];
- attr_init (&relation, zapt, 2);
- relation_value = attr_find (&relation, NULL);
+ attr_init_APT(&relation, zapt, 2);
+ relation_value = attr_find(&relation, NULL);
- logf (LOG_DEBUG, "string relation value=%d", relation_value);
+ *error_code = 0;
+ yaz_log(YLOG_DEBUG, "string relation value=%d", relation_value);
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;
- logf (LOG_DEBUG, "Relation <");
+ yaz_log(log_level_rpn, "Relation <");
*term_tmp++ = '(';
for (i = 0; term_component[i]; )
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;
- logf (LOG_DEBUG, "Relation <=");
+ yaz_log(log_level_rpn, "Relation <=");
*term_tmp++ = '(';
for (i = 0; term_component[i]; )
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;
if (!term_100 (zh->reg->zebra_maps, reg_type,
term_sub, term_component, space_split, term_dst))
return 0;
- logf (LOG_DEBUG, "Relation >");
+ yaz_log(log_level_rpn, "Relation >");
*term_tmp++ = '(';
for (i = 0; term_component[i];)
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;
- logf (LOG_DEBUG, "Relation >=");
+ yaz_log(log_level_rpn, "Relation >=");
*term_tmp++ = '(';
for (i = 0; term_component[i];)
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++ = ']';
*term_tmp = '\0';
break;
case 3:
- default:
- logf (LOG_DEBUG, "Relation =");
- if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub,
- term_component, space_split, term_dst))
+ case 102:
+ case 103:
+ case -1:
+ if (!**term_sub)
+ return 1;
+ yaz_log(log_level_rpn, "Relation =");
+ 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);
- strcat (term_tmp, ")");
+ strcat(term_tmp, "(");
+ strcat(term_tmp, term_component);
+ strcat(term_tmp, ")");
+ break;
+ default:
+ *error_code = YAZ_BIB1_UNSUPP_RELATION_ATTRIBUTE;
+ return 0;
}
return 1;
}
-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);
-
-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)
+static ZEBRA_RES 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 *xpath_use,
+ struct ord_list **ol);
+
+static ZEBRA_RES term_limits_APT(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ zint *hits_limit_value,
+ const char **term_ref_id_str,
+ NMEM nmem)
{
- 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);
+ AttrType term_ref_id_attr;
+ AttrType hits_limit_attr;
+ int term_ref_id_int;
+
+ attr_init_APT(&hits_limit_attr, zapt, 9);
+ *hits_limit_value = attr_find(&hits_limit_attr, NULL);
+
+ attr_init_APT(&term_ref_id_attr, zapt, 10);
+ term_ref_id_int = attr_find_ex(&term_ref_id_attr, NULL, term_ref_id_str);
+ if (term_ref_id_int >= 0)
+ {
+ char *res = nmem_malloc(nmem, 20);
+ sprintf(res, "%d", term_ref_id_int);
+ *term_ref_id_str = res;
+ }
+
+ /* no limit given ? */
+ if (*hits_limit_value == -1)
+ {
+ if (*term_ref_id_str)
+ {
+ /* use global if term_ref is present */
+ *hits_limit_value = zh->approx_limit;
+ }
+ else
+ {
+ /* no counting if term_ref is not present */
+ *hits_limit_value = 0;
+ }
+ }
+ else if (*hits_limit_value == 0)
+ {
+ /* 0 is the same as global limit */
+ *hits_limit_value = zh->approx_limit;
+ }
+ yaz_log(YLOG_DEBUG, "term_limits_APT ref_id=%s limit=" ZINT_FORMAT,
+ *term_ref_id_str ? *term_ref_id_str : "none",
+ *hits_limit_value);
+ return ZEBRA_OK;
}
+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,
+ const char *xpath_use,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
+{
+ ZEBRA_RES res;
+ struct ord_list *ol;
+ zint hits_limit_value;
+ const char *term_ref_id_str = 0;
+ *rset = 0;
+
+ term_limits_APT(zh, zapt, &hits_limit_value, &term_ref_id_str,
+ stream);
+ 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, &ol);
+ if (res != ZEBRA_OK)
+ return res;
+ if (!*term_sub) /* no more terms ? */
+ return res;
+ yaz_log(log_level_rpn, "term: %s", term_dst);
+ *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, ol, reg_type, hits_limit_value,
+ term_ref_id_str);
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return ZEBRA_OK;
+}
-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)
+static ZEBRA_RES 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 *xpath_use,
+ struct ord_list **ol)
{
char term_dict[2*IT_MAX_WORD+4000];
int j, r, base_no;
AttrType truncation;
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;
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 */
+
+ *ol = ord_list_create(stream);
rpn_char_map_prepare (zh->reg, reg_type, &rcmi);
- attr_init (&use, zapt, 1);
- 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) /* no attribute - assumy "any" */
- use_value = 1016;
+ attr_init_APT(&truncation, zapt, 5);
+ truncation_value = attr_find(&truncation, NULL);
+ yaz_log(log_level_rpn, "truncation value %d", truncation_value);
+
for (base_no = 0; base_no < num_bases; base_no++)
{
+ int ord = -1;
+ int attr_ok = 0;
+ int regex_range = 0;
+ int init_pos = 0;
+#if 0
attent attp;
data1_local_attribute id_xpath_attr;
data1_local_attribute *local_attr;
+#endif
int max_pos, prefix_len = 0;
+ int relation_error;
+ char ord_buf[32];
+ int ord_len, i;
termp = *term_sub;
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
- zh->errCode = 109; /* Database unavailable */
- zh->errString = basenames[base_no];
- return -1;
+ zebra_setError(zh, YAZ_BIB1_DATABASE_UNAVAILABLE,
+ basenames[base_no]);
+ return ZEBRA_FAIL;
}
- if (xpath_use > 0 && use_value == -2)
- {
- 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)
+
+ if (zebra_apt_get_ord(zh, zapt, reg_type, xpath_use,
+ curAttributeSet, &ord)
+ != ZEBRA_OK)
{
- attp.local_attributes = &id_xpath_attr;
- attp.attset_ordinal = VAL_IDXPATH;
- id_xpath_attr.next = 0;
- id_xpath_attr.local = use_value;
+ break;
}
+ *ol = ord_list_append(stream, *ol, ord);
+
+ 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++)
{
- 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 */
- 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
- {
- int oid[OID_SIZE];
- struct oident oident;
-
- oident.proto = PROTO_Z3950;
- oident.oclass = CLASS_ATTSET;
- oident.value = curAttributeSet;
- oid_ent_to_oid (&oident, oid);
-
- errCode = 121;
- errString = nmem_strdup (stream, oident.desc);
- }
- continue;
- }
- }
- for (local_attr = attp.local_attributes; local_attr;
- local_attr = local_attr->next)
- {
- int ord;
- char ord_buf[32];
- int i, ord_len;
-
- ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- 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];
- }
- }
- if (!prefix_len)
- {
-#if 1
- bases_ok++;
-#else
- char val_str[32];
- sprintf (val_str, "%d", use_value);
- errCode = 114;
- errString = nmem_strdup (stream, val_str);
-#endif
- continue;
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = ord_buf[i];
}
- bases_ok++; /* this has OK attributes */
+ if (ord_len > init_pos)
+ init_pos = ord_len;
+
+ bases_ok++;
+ if (prefix_len)
+ attr_ok = 1;
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]);
term_dict[prefix_len] = '\0';
j = prefix_len;
switch (truncation_value)
case 100: /* do not truncate */
if (!string_relation (zh, zapt, &termp, term_dict,
attributeSet,
- reg_type, space_split, term_dst))
- return 0;
- logf (LOG_LOG, "dict_lookup_grep: %s", term_dict+prefix_len);
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
- grep_info, &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep fail %d", r);
- break;
+ reg_type, space_split, term_dst,
+ &relation_error))
+ {
+ if (relation_error)
+ {
+ zebra_setError(zh, relation_error, 0);
+ return ZEBRA_FAIL;
+ }
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ break;
case 1: /* right truncation */
term_dict[j++] = '(';
- if (!term_100 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
- return 0;
- strcat (term_dict, ".*)");
- dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
+ if (!term_100(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ".*)");
break;
case 2: /* keft truncation */
term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*';
- if (!term_100 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
- return 0;
- strcat (term_dict, ")");
- dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
+ if (!term_100(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
break;
case 3: /* left&right truncation */
term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*';
- if (!term_100 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
- return 0;
- strcat (term_dict, ".*)");
- dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
+ if (!term_100(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ".*)");
break;
- zh->errCode = 120;
- return -1;
case 101: /* process # in term */
term_dict[j++] = '(';
- if (!term_101 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
- return 0;
- strcat (term_dict, ")");
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=#: %d", r);
+ if (!term_101(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
break;
case 102: /* Regexp-1 */
term_dict[j++] = '(';
- if (!term_102 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, space_split, term_dst))
- return 0;
- strcat (term_dict, ")");
- logf (LOG_DEBUG, "Regexp-1 tolerance=%d", r);
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=regular: %d",
- r);
+ if (!term_102(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
break;
case 103: /* Regexp-2 */
- r = 1;
+ regex_range = 1;
term_dict[j++] = '(';
- if (!term_103 (zh->reg->zebra_maps, reg_type,
- &termp, term_dict + j, &r, space_split, term_dst))
- return 0;
- strcat (term_dict, ")");
- logf (LOG_DEBUG, "Regexp-2 tolerance=%d", r);
- r = dict_lookup_grep (zh->reg->dict, term_dict, r, grep_info,
- &max_pos, 2, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=eregular: %d",
- r);
- break;
+ 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))
- return 0;
- strcat (term_dict, ")");
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=#/!: %d", r);
+ if (!term_104(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
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))
- return 0;
- strcat (term_dict, ")");
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=*/!: %d", r);
+ if (!term_105(zh->reg->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst, 1))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
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))
- return 0;
- strcat (term_dict, ")");
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info,
- &max_pos, 0, grep_handle);
- if (r)
- logf (LOG_WARN, "dict_lookup_grep err, trunc=*/!: %d", r);
+ 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:
+ zebra_setError_zint(zh,
+ YAZ_BIB1_UNSUPP_TRUNCATION_ATTRIBUTE,
+ truncation_value);
+ return ZEBRA_FAIL;
+ }
+ if (attr_ok)
+ {
+ char buf[80];
+ const char *input = term_dict + prefix_len;
+ esc_str(buf, sizeof(buf), input, strlen(input));
+ }
+ if (attr_ok)
+ {
+ yaz_log(log_level_rpn, "dict_lookup_grep: %s", term_dict+prefix_len);
+ r = dict_lookup_grep(zh->reg->dict, term_dict, regex_range,
+ grep_info, &max_pos, init_pos,
+ grep_handle);
+ if (r)
+ yaz_log(YLOG_WARN, "dict_lookup_grep fail %d", r);
+ }
}
if (!bases_ok)
- {
- zh->errCode = errCode;
- zh->errString = errString;
- return -1;
- }
+ return ZEBRA_FAIL;
*term_sub = termp;
- logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
- return 1;
+ yaz_log(YLOG_DEBUG, "%d positions", grep_info->isam_p_indx);
+ return ZEBRA_OK;
}
/* 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;
case Z_Term_general:
if (zh->iconv_to_utf8 != 0)
{
- char *inbuf = term->u.general->buf;
+ char *inbuf = (char *) term->u.general->buf;
size_t inleft = term->u.general->len;
char *outbuf = termz;
size_t outleft = IT_MAX_WORD-1;
if (ret == (size_t)(-1))
{
ret = yaz_iconv(zh->iconv_to_utf8, 0, 0, 0, 0);
- zh->errCode = 125;
- return -1;
+ zebra_setError(
+ zh,
+ YAZ_BIB1_QUERY_TERM_INCLUDES_CHARS_THAT_DO_NOT_TRANSLATE_INTO_,
+ 0);
+ return ZEBRA_FAIL;
}
*outbuf = 0;
}
termz[sizez] = '\0';
break;
default:
- zh->errCode = 124;
- return -1;
+ zebra_setError(zh, YAZ_BIB1_UNSUPP_CODED_VALUE_FOR_TERM, 0);
+ 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);
+ 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,
- const char *termz, NMEM stream, unsigned reg_id)
-{
- WRBUF wrbuf = 0;
- AttrType truncation;
- int truncation_value;
- char *ex_list = 0;
-
- attr_init (&truncation, zapt, 5);
- truncation_value = attr_find (&truncation, NULL);
-
- switch (truncation_value)
- {
- default:
- ex_list = "";
- break;
- case 101:
- ex_list = "#";
- break;
- case 102:
- case 103:
- ex_list = 0;
- break;
- case 104:
- ex_list = "!#";
- break;
- case 105:
- ex_list = "!*";
- break;
- }
- if (ex_list)
- wrbuf = zebra_replace(zh->reg->zebra_maps, reg_id, ex_list,
- termz, strlen(termz));
- if (!wrbuf)
- return nmem_strdup(stream, termz);
- else
- {
- char *buf = (char*) nmem_malloc (stream, wrbuf_len(wrbuf)+1);
- memcpy (buf, wrbuf_buf(wrbuf), wrbuf_len(wrbuf));
- buf[wrbuf_len(wrbuf)] = '\0';
- return buf;
- }
-}
-
-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 ZEBRA_RES grep_info_prepare(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ struct grep_info *grep_info,
+ int reg_type)
{
AttrType termset;
int termset_value_numeric;
grep_info->termset = 0;
if (!zapt)
- return 0;
- attr_init (&termset, zapt, 8);
+ return ZEBRA_OK;
+ attr_init_APT(&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;
- logf (LOG_LOG, "creating termset set %s", termset_name);
- grep_info->termset = resultSetAdd (zh, termset_name, 1);
+ yaz_log(log_level_rpn, "creating termset set %s", termset_name);
+ grep_info->termset = resultSetAdd(zh, termset_name, 1);
if (!grep_info->termset)
{
- zh->errCode = 128;
- zh->errString = nmem_strdup (stream, termset_name);
- return -1;
+ zebra_setError(zh, YAZ_BIB1_ILLEGAL_RESULT_SET_NAME, termset_name);
+ return ZEBRA_FAIL;
}
}
- return 0;
+ return ZEBRA_OK;
}
-
-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)
+/**
+ \brief Create result set(s) for list of terms
+ \param zh Zebra Handle
+ \param termz term as used in query but converted to UTF-8
+ \param attributeSet default attribute set
+ \param stream memory for result
+ \param reg_type register type ('w', 'p',..)
+ \param complete_flag whether it's phrases or not
+ \param rank_type term flags for ranking
+ \param xpath_use use attribute for X-Path (-1 for no X-path)
+ \param num_bases number of databases
+ \param basenames array of databases
+ \param rset_mem memory for result sets
+ \param result_sets output result set for each term in list (output)
+ \param number number of output result sets
+ \param kc rset key control to be used for created result sets
+*/
+static ZEBRA_RES term_list_trunc(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type,
+ const char *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[60], result;
- int 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;
+ int empty_term = *termz ? 0 : 1;
+ empty_term = 0;
+ *num_result_sets = 0;
*term_dst = 0;
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
- while (1)
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type) == ZEBRA_FAIL)
+ return ZEBRA_FAIL;
+ while(1)
{
- logf (LOG_DEBUG, "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);
- if (!rset[rset_no])
+ 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)++;
+
+ if (empty_term)
break;
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
+ if (!*termp)
break;
}
- grep_info_delete (&grep_info);
- if (rset_no == 0)
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
- else if (rset_no == 1)
- return (rset[0]);
- else
- result = rsprox_create( NULL, /* FIXME - use a proper nmem */
- sizeof(struct it_key), key_compare_it, key_get_seq,
- 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)
+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,
+ const char *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[60];
- 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 (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
- while (1)
- {
- logf (LOG_DEBUG, "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);
- if (!rset[rset_no])
- break;
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
- }
- grep_info_delete (&grep_info);
- if (rset_no == 0)
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
- return rsmultior_create(NULL, /* FIXME */
- sizeof(struct it_key),
- key_compare_it,
- rset_no, rset);
-
-#if 0
- /* The old code, kept for reference. May be deleted soon */
- result = rset[0];
- for (i = 1; i<rset_no; i++)
- {
- rset_bool_parms bool_parms;
- bool_parms.rset_l = result;
- 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;
-#endif
+ 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 = rset_create_null(rset_nmem, kc, 0);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rset_create_prox(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)
+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,
+ const char *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[60], result;
- 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;
+ 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 = rset_create_null(rset_nmem, kc, 0);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rset_create_or(rset_nmem, kc, kc->scope, 0 /* termid */,
+ 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;
- while (1)
- {
- logf (LOG_DEBUG, "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);
- if (!rset[rset_no])
- break;
- assert (rset[rset_no]);
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
- }
- grep_info_delete (&grep_info);
- if (rset_no == 0)
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
- result = rset[0];
- /* FIXME - Use a proper rsmultiand */
- for (i = 1; i<rset_no; i++)
- {
- /*
- rset_bool_parms bool_parms;
- bool_parms.rset_l = result;
- 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);
- */
- result= rsbool_create_and(NULL, /* FIXME - use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- result, rset[i], key_logdump_txt );
- }
- return result;
+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,
+ const char *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 = rset_create_null(rset_nmem, kc, 0);
+ else if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rset_create_and(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,
- const char **term_sub,
- char *term_dict,
- oid_value attributeSet,
- struct grep_info *grep_info,
- int *max_pos,
- int reg_type,
- char *term_dst)
+static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ char *term_dict,
+ oid_value attributeSet,
+ struct grep_info *grep_info,
+ int *max_pos,
+ int reg_type,
+ char *term_dst,
+ int *error_code)
{
AttrType relation;
int relation_value;
int r;
char *term_tmp = term_dict + strlen(term_dict);
- attr_init (&relation, zapt, 2);
- relation_value = attr_find (&relation, NULL);
+ *error_code = 0;
+ attr_init_APT(&relation, zapt, 2);
+ relation_value = attr_find(&relation, NULL);
- logf (LOG_DEBUG, "numeric relation value=%d", relation_value);
+ yaz_log(log_level_rpn, "numeric relation value=%d", relation_value);
- if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub, term_tmp, 1,
- term_dst))
+ if (!term_100(zh->reg->zebra_maps, reg_type, term_sub, term_tmp, 1,
+ term_dst))
return 0;
term_value = atoi (term_tmp);
switch (relation_value)
{
case 1:
- logf (LOG_DEBUG, "Relation <");
- gen_regular_rel (term_tmp, term_value-1, 1);
+ yaz_log(log_level_rpn, "Relation <");
+ gen_regular_rel(term_tmp, term_value-1, 1);
break;
case 2:
- logf (LOG_DEBUG, "Relation <=");
- gen_regular_rel (term_tmp, term_value, 1);
+ yaz_log(log_level_rpn, "Relation <=");
+ gen_regular_rel(term_tmp, term_value, 1);
break;
case 4:
- logf (LOG_DEBUG, "Relation >=");
- gen_regular_rel (term_tmp, term_value, 0);
+ yaz_log(log_level_rpn, "Relation >=");
+ gen_regular_rel(term_tmp, term_value, 0);
break;
case 5:
- logf (LOG_DEBUG, "Relation >");
- gen_regular_rel (term_tmp, term_value+1, 0);
+ yaz_log(log_level_rpn, "Relation >");
+ gen_regular_rel(term_tmp, term_value+1, 0);
break;
+ case -1:
case 3:
+ yaz_log(log_level_rpn, "Relation =");
+ sprintf(term_tmp, "(0*%d)", term_value);
+ break;
default:
- logf (LOG_DEBUG, "Relation =");
- sprintf (term_tmp, "(0*%d)", term_value);
+ *error_code = YAZ_BIB1_UNSUPP_RELATION_ATTRIBUTE;
+ return 0;
}
- logf (LOG_DEBUG, "dict_lookup_grep: %s", term_tmp);
- r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, max_pos,
+ yaz_log(log_level_rpn, "dict_lookup_grep: %s", term_tmp);
+ r = dict_lookup_grep(zh->reg->dict, term_dict, 0, grep_info, max_pos,
0, grep_handle);
if (r)
- logf (LOG_WARN, "dict_lookup_grep fail, rel=gt: %d", r);
- logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
+ yaz_log(YLOG_WARN, "dict_lookup_grep fail, rel = gt: %d", r);
+ yaz_log(log_level_rpn, "%d positions", grep_info->isam_p_indx);
return 1;
}
-static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- const char **term_sub,
- 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, NMEM stream)
+static ZEBRA_RES numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ oid_value attributeSet,
+ struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst,
+ const char *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;
+ int base_no;
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_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;
+ int relation_error = 0;
+ int ord, ord_len, i;
+ char ord_buf[32];
termp = *term_sub;
- 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,
- use_string)))
- {
- 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]))
{
- zh->errCode = 109; /* Database unavailable */
- zh->errString = basenames[base_no];
- return -1;
+ zebra_setError(zh, YAZ_BIB1_DATABASE_UNAVAILABLE,
+ basenames[base_no]);
+ return ZEBRA_FAIL;
}
- for (local_attr = attp.local_attributes; local_attr;
- local_attr = local_attr->next)
- {
- int ord;
- char ord_buf[32];
- int i, ord_len;
-
- ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- 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];
- }
+ if (zebra_apt_get_ord(zh, zapt, reg_type, xpath_use,
+ curAttributeSet, &ord)
+ != ZEBRA_OK)
+ {
+ break;
}
- if (!prefix_len)
+
+ 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++)
{
- char val_str[32];
- sprintf (val_str, "%d", use_value);
- errCode = 114;
- errString = nmem_strdup (stream, val_str);
- continue;
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = ord_buf[i];
}
bases_ok++;
- 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]);
+ term_dict[prefix_len++] = ')';
term_dict[prefix_len] = '\0';
- if (!numeric_relation (zh, zapt, &termp, term_dict,
- attributeSet, grep_info, &max_pos, reg_type,
- term_dst))
- return 0;
+ if (!numeric_relation(zh, zapt, &termp, term_dict,
+ attributeSet, grep_info, &max_pos, reg_type,
+ term_dst, &relation_error))
+ {
+ if (relation_error)
+ {
+ zebra_setError(zh, relation_error, 0);
+ return ZEBRA_FAIL;
+ }
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
}
if (!bases_ok)
- {
- zh->errCode = errCode;
- zh->errString = errString;
- return -1;
- }
+ return ZEBRA_FAIL;
*term_sub = termp;
- logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
- return 1;
+ yaz_log(YLOG_DEBUG, "%d positions", grep_info->isam_p_indx);
+ 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)
+
+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,
+ const char *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[60], result;
- int i, r, rset_no = 0;
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ ZEBRA_RES res;
struct grep_info grep_info;
+ int alloc_sets = 0;
+ zint hits_limit_value;
+ const char *term_ref_id_str = 0;
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
- return 0;
+ term_limits_APT(zh, zapt, &hits_limit_value, &term_ref_id_str, stream);
+
+ yaz_log(log_level_rpn, "APT_numeric t='%s'", termz);
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type) == ZEBRA_FAIL)
+ return ZEBRA_FAIL;
while (1)
{
- logf (LOG_DEBUG, "APT_numeric termp=%s", termp);
+ 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 < 1)
- break;
- logf (LOG_DEBUG, "term: %s", term_dst);
- rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf,
- grep_info.isam_p_indx, term_dst,
- strlen(term_dst), rank_type,
- 0 /* preserve position */,
- zapt->term->which);
- assert (rset[rset_no]);
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
+ 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);
+ 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, 0, reg_type,
+ hits_limit_value,
+ term_ref_id_str);
+ if (!result_sets[num_result_sets])
+ break;
+ num_result_sets++;
}
- grep_info_delete (&grep_info);
- if (rset_no == 0)
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
- result = rset[0];
- for (i = 1; i<rset_no; i++)
+ grep_info_delete(&grep_info);
+ if (termp)
{
- /*
- rset_bool_parms bool_parms;
- bool_parms.rset_l = result;
- 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);
- */
- /* FIXME - Use a proper multi-and */
- result= rsbool_create_and(NULL, /* FIXME - use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- result, rset[i], key_logdump_txt );
+ int i;
+ for (i = 0; i<num_result_sets; i++)
+ rset_delete(result_sets[i]);
+ return ZEBRA_FAIL;
}
- return result;
+ if (num_result_sets == 0)
+ *rset = rset_create_null(rset_nmem, kc, 0);
+ if (num_result_sets == 1)
+ *rset = result_sets[0];
+ else
+ *rset = rset_create_and(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)
+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;
- /*
- rset_temp_parms parms;
-
- 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);
- */
- result = rstemp_create( NULL, /* FIXME - use a proper nmem */
- sizeof (struct it_key), key_compare_it,
- res_get (zh->res, "setTmpDir") );
- rsfd = rset_open (result, RSETF_WRITE);
-
+ *rset = rset_create_temp(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;
-#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;
+ rset_write (rsfd, &key);
+ rset_close (rsfd);
+ 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,
+ NMEM rset_nmem,
+ RSET *rset,
+ struct rset_key_control *kc)
{
int i;
int sort_relation_value;
AttrType sort_relation_type;
- int use_value;
- AttrType use_type;
Z_SortKeySpec *sks;
Z_SortKey *sk;
- Z_AttributeElement *ae;
int oid[OID_SIZE];
oident oe;
char termz[20];
- 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_APT(&sort_relation_type, zapt, 7);
+ sort_relation_value = attr_find(&sort_relation_type, &attributeSet);
if (!sort_sequence->specs)
{
sort_sequence->num_specs = 10;
sort_sequence->specs = (Z_SortKeySpec **)
- nmem_malloc (stream, sort_sequence->num_specs *
+ nmem_malloc(stream, sort_sequence->num_specs *
sizeof(*sort_sequence->specs));
for (i = 0; i<sort_sequence->num_specs; i++)
sort_sequence->specs[i] = 0;
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 = (Z_SortKeySpec *) nmem_malloc(stream, sizeof(*sks));
sks->sortElement = (Z_SortElement *)
- nmem_malloc (stream, sizeof(*sks->sortElement));
+ nmem_malloc(stream, sizeof(*sks->sortElement));
sks->sortElement->which = Z_SortElement_generic;
sk = sks->sortElement->u.generic = (Z_SortKey *)
- nmem_malloc (stream, sizeof(*sk));
+ nmem_malloc(stream, sizeof(*sk));
sk->which = Z_SortKey_sortAttributes;
sk->u.sortAttributes = (Z_SortAttributes *)
- nmem_malloc (stream, sizeof(*sk->u.sortAttributes));
+ nmem_malloc(stream, sizeof(*sk->u.sortAttributes));
sk->u.sortAttributes->id = oid;
- sk->u.sortAttributes->list = (Z_AttributeList *)
- nmem_malloc (stream, sizeof(*sk->u.sortAttributes->list));
- sk->u.sortAttributes->list->num_attributes = 1;
- sk->u.sortAttributes->list->attributes = (Z_AttributeElement **)
- nmem_malloc (stream, sizeof(*sk->u.sortAttributes->list->attributes));
- ae = *sk->u.sortAttributes->list->attributes = (Z_AttributeElement *)
- nmem_malloc (stream, sizeof(**sk->u.sortAttributes->list->attributes));
- ae->attributeSet = 0;
- ae->attributeType = (int *)
- nmem_malloc (stream, sizeof(*ae->attributeType));
- *ae->attributeType = 1;
- ae->which = Z_AttributeValue_numeric;
- ae->value.numeric = (int *)
- nmem_malloc (stream, sizeof(*ae->value.numeric));
- *ae->value.numeric = use_value;
+ sk->u.sortAttributes->list = zapt->attributes;
sks->sortRelation = (int *)
- nmem_malloc (stream, sizeof(*sks->sortRelation));
+ nmem_malloc(stream, sizeof(*sks->sortRelation));
if (sort_relation_value == 1)
*sks->sortRelation = Z_SortKeySpec_ascending;
else if (sort_relation_value == 2)
*sks->sortRelation = Z_SortKeySpec_ascending;
sks->caseSensitivity = (int *)
- nmem_malloc (stream, sizeof(*sks->caseSensitivity));
+ nmem_malloc(stream, sizeof(*sks->caseSensitivity));
*sks->caseSensitivity = 0;
sks->which = Z_SortKeySpec_null;
sks->u.null = odr_nullval ();
sort_sequence->specs[i] = sks;
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
+ *rset = rset_create_null(rset_nmem, kc, 0);
+ return ZEBRA_OK;
}
AttrType use;
const char *use_string = 0;
- attr_init (&use, zapt, 1);
- attr_find_ex (&use, &curAttributeSet, &use_string);
+ attr_init_APT(&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)
+ int reg_type, const char *term,
+ const char *xpath_use,
+ oid_value curAttributeSet, NMEM rset_nmem,
+ struct rset_key_control *kc)
{
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 = zebraExplain_lookup_attr_str(zh->reg->zei, reg_type,
+ xpath_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))
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
-
+ if (grep_info_prepare(zh, 0 /* zapt */, &grep_info, '0') == ZEBRA_FAIL)
+ return rset_create_null(rset_nmem, kc, 0);
+
if (ord < 0)
- return rsnull_create (NULL); /* FIXME - Use a proper nmem */
+ return rset_create_null(rset_nmem, kc, 0);
if (prefix_len)
term_dict[prefix_len++] = '|';
else
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);
+ strcpy(term_dict+prefix_len, term);
grep_info.isam_p_indx = 0;
- r = dict_lookup_grep (zh->reg->dict, term_dict, 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,
+ 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);
- grep_info_delete (&grep_info);
+ rset = rset_trunc(zh, grep_info.isam_p_buf,
+ grep_info.isam_p_indx, term, strlen(term),
+ flags, 1, term_type,rset_nmem,
+ kc, kc->scope, 0, reg_type, 0 /* hits_limit */,
+ 0 /* term_ref_id_str */);
+ 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)
+static
+ZEBRA_RES 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,
+ RSET *rset_out,
+ struct rset_key_control *kc)
{
oid_value curAttributeSet = attributeSet;
int base_no;
int i;
if (xpath_len < 0)
- return rset;
+ {
+ *rset_out = rset;
+ return ZEBRA_OK;
+ }
- yaz_log (LOG_LOG, "len=%d", xpath_len);
+ yaz_log(YLOG_DEBUG, "xpath len=%d", xpath_len);
for (i = 0; i<xpath_len; i++)
{
- yaz_log (LOG_LOG, "XPATH %d %s", i, xpath[i].part);
+ yaz_log(log_level_rpn, "XPATH %d %s", i, xpath[i].part);
}
/ -> none
- a[@attr=value]/b[@other=othervalue]
+ 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)
+ /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)
*/
if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
{
- zh->errCode = 109; /* Database unavailable */
- zh->errString = basenames[base_no];
- return rset;
+ zebra_setError(zh, YAZ_BIB1_DATABASE_UNAVAILABLE,
+ basenames[base_no]);
+ *rset_out = rset;
+ return ZEBRA_FAIL;
}
while (--level >= 0)
{
}
}
wrbuf_puts(wbuf, "");
- rset_attr = xpath_trunc (
- zh, stream, '0', wrbuf_buf(wbuf), 3, curAttributeSet);
+ rset_attr = xpath_trunc(
+ zh, stream, '0', wrbuf_buf(wbuf), ZEBRA_XPATH_ATTR_NAME,
+ curAttributeSet, rset_nmem, kc);
wrbuf_free(wbuf, 1);
}
else
if (!first_path)
continue;
}
- yaz_log (LOG_LOG, "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_start_tag = xpath_trunc(zh, stream, '0',
+ xpath_rev,
+ ZEBRA_XPATH_ELM_BEGIN,
+ curAttributeSet,
+ rset_nmem, kc);
- 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);
- */
- rset=rsbetween_create( NULL, /* FIXME */
- sizeof(struct it_key), key_compare_it,
- rset_start_tag, rset, rset_end_tag, rset_attr,
- key_print_it );
+ rset_end_tag = xpath_trunc(zh, stream, '0',
+ xpath_rev,
+ ZEBRA_XPATH_ELM_END,
+ curAttributeSet,
+ rset_nmem, kc);
+
+ rset = rset_create_between(rset_nmem, kc, kc->scope,
+ rset_start_tag, rset,
+ rset_end_tag, rset_attr);
}
first_path = 0;
}
}
-
- return rset;
+ *rset_out = rset;
+ return ZEBRA_OK;
}
-
-
-static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
- oid_value attributeSet, NMEM stream,
- Z_SortKeySpecList *sort_sequence,
- int num_bases, char **basenames)
+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,
+ struct rset_key_control *kc)
{
+ ZEBRA_RES res = ZEBRA_OK;
unsigned reg_id;
char *search_type = NULL;
char rank_type[128];
int complete_flag;
int sort_flag;
char termz[IT_MAX_WORD+1];
- RSET rset = 0;
int xpath_len;
- int xpath_use = 0;
+ const char *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);
+ if (!log_level_set)
+ {
+ log_level_rpn = yaz_log_module_level("rpn");
+ log_level_set = 1;
+ }
+ zebra_maps_attr(zh->reg->zebra_maps, zapt, ®_id, &search_type,
+ rank_type, &complete_flag, &sort_flag);
- logf (LOG_DEBUG, "reg_id=%c", reg_id);
- logf (LOG_DEBUG, "complete_flag=%d", complete_flag);
- logf (LOG_DEBUG, "search_type=%s", search_type);
- logf (LOG_DEBUG, "rank_type=%s", rank_type);
+ yaz_log(YLOG_DEBUG, "reg_id=%c", reg_id);
+ yaz_log(YLOG_DEBUG, "complete_flag=%d", complete_flag);
+ 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))
- return 0;
+ if (zapt_term_to_utf8(zh, zapt, termz) == ZEBRA_FAIL)
+ return ZEBRA_FAIL;
if (sort_flag)
- return rpn_sort_spec (zh, zapt, attributeSet, stream, sort_sequence,
- rank_type);
+ return rpn_sort_spec(zh, zapt, attributeSet, stream, sort_sequence,
+ rank_type, rset_nmem, rset, kc);
+ /* consider if an X-Path query is used */
xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream);
if (xpath_len >= 0)
{
- xpath_use = 1016;
- if (xpath[xpath_len-1].part[0] == '@')
- xpath_use = 1015;
+ if (xpath[xpath_len-1].part[0] == '@')
+ xpath_use = ZEBRA_XPATH_ATTR_CDATA; /* last step is attribute */
+ else
+ xpath_use = ZEBRA_XPATH_CDATA; /* searching for cdata */
}
- if (!strcmp (search_type, "phrase"))
+ /* search using one of the various search type strategies
+ termz is our UTF-8 search term
+ attributeSet is top-level default attribute set
+ stream is ODR for search
+ reg_id is the register type
+ complete_flag is 1 for complete subfield, 0 for incomplete
+ xpath_use is use-attribute to be used for X-Path search, 0 for none
+ */
+ 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);
+ 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);
+ 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);
+ 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);
+ 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);
+ 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 if (!strcmp (search_type, "always"))
+ else if (!strcmp(search_type, "always"))
{
- rset = 0;
+ *termz = '\0';
+ 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
- zh->errCode = 118;
- return rpn_search_xpath (zh, attributeSet, num_bases, basenames,
- stream, rank_type, rset, xpath_len, xpath);
+ {
+ zebra_setError(zh, YAZ_BIB1_UNSUPP_STRUCTURE_ATTRIBUTE, 0);
+ res = ZEBRA_FAIL;
+ }
+ if (res != ZEBRA_OK)
+ return res;
+ if (!*rset)
+ return ZEBRA_FAIL;
+ return rpn_search_xpath(zh, attributeSet, num_bases, basenames,
+ stream, rank_type, *rset,
+ xpath_len, xpath, rset_nmem, rset, kc);
}
-static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs,
- oid_value attributeSet,
- NMEM stream, NMEM rset_nmem,
- Z_SortKeySpecList *sort_sequence,
- int num_bases, char **basenames)
+static ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs,
+ oid_value attributeSet,
+ NMEM stream, NMEM rset_nmem,
+ Z_SortKeySpecList *sort_sequence,
+ int num_bases, char **basenames,
+ RSET **result_sets, int *num_result_sets,
+ Z_Operator *parent_op,
+ struct rset_key_control *kc);
+
+ZEBRA_RES rpn_search_top(ZebraHandle zh, Z_RPNStructure *zs,
+ oid_value attributeSet,
+ NMEM stream, NMEM rset_nmem,
+ Z_SortKeySpecList *sort_sequence,
+ int num_bases, char **basenames,
+ RSET *result_set)
{
- RSET r = NULL;
+ RSET *result_sets = 0;
+ int num_result_sets = 0;
+ 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;
+ for (i = 0; i<num_result_sets; i++)
+ rset_delete(result_sets[i]);
+ *result_set = 0;
+ }
+ else
+ {
+ assert(num_result_sets == 1);
+ assert(result_sets);
+ assert(*result_sets);
+ *result_set = *result_sets;
+ }
+ (*kc->dec)(kc);
+ return res;
+}
+
+ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs,
+ oid_value attributeSet,
+ NMEM stream, NMEM rset_nmem,
+ Z_SortKeySpecList *sort_sequence,
+ int num_bases, char **basenames,
+ RSET **result_sets, int *num_result_sets,
+ Z_Operator *parent_op,
+ struct rset_key_control *kc)
+{
+ *num_result_sets = 0;
if (zs->which == Z_RPNStructure_complex)
{
+ ZEBRA_RES res;
Z_Operator *zop = zs->u.complex->roperator;
- RSET rset_l;
- RSET rset_r;
-
- rset_l = rpn_search_structure (zh, zs->u.complex->s1,
- attributeSet, stream, rset_nmem,
- sort_sequence,
- num_bases, basenames);
- if (rset_l == NULL)
- return NULL;
- rset_r = rpn_search_structure (zh, zs->u.complex->s2,
- attributeSet, stream, rset_nmem,
- sort_sequence,
- num_bases, basenames);
- if (rset_r == NULL)
- {
- rset_delete (rset_l);
- return NULL;
- }
-
- switch (zop->which)
- {
- case Z_Operator_and:
- r = rsbool_create_and( rset_nmem,
- sizeof(struct it_key), key_compare_it,
- rset_l, rset_r, key_logdump_txt );
- break;
- case Z_Operator_or:
- r = rsbool_create_or( rset_nmem,
- sizeof(struct it_key), key_compare_it,
- rset_l, rset_r, key_logdump_txt );
- break;
- case Z_Operator_and_not:
- r = rsbool_create_not( rset_nmem,
- sizeof(struct it_key), key_compare_it,
- rset_l, rset_r, key_logdump_txt );
- break;
- case Z_Operator_prox:
- if (zop->u.prox->which != Z_ProximityOperator_known)
- {
- zh->errCode = 132;
- return NULL;
- }
- if (*zop->u.prox->u.known != Z_ProxUnit_word)
- {
- char *val = (char *) nmem_malloc (stream, 16);
- zh->errCode = 132;
- zh->errString = val;
- sprintf (val, "%d", *zop->u.prox->u.known);
- return NULL;
- }
- else
- {
- /* new / old prox */
- RSET twosets[2];
-
- twosets[0] = rset_l;
- twosets[1] = rset_r;
- r=rsprox_create(rset_nmem,
- sizeof(struct it_key), key_compare_it,
- key_get_seq,
- 2, twosets,
- *zop->u.prox->ordered,
- (!zop->u.prox->exclusion ?
- 0 : *zop->u.prox->exclusion),
- *zop->u.prox->relationType,
- *zop->u.prox->distance );
- /*
- 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:
- zh->errCode = 110;
- return NULL;
- }
+ RSET *result_sets_l = 0;
+ int num_result_sets_l = 0;
+ RSET *result_sets_r = 0;
+ int num_result_sets_r = 0;
+
+ res = rpn_search_structure(zh, zs->u.complex->s1,
+ attributeSet, stream, rset_nmem,
+ sort_sequence,
+ num_bases, basenames,
+ &result_sets_l, &num_result_sets_l,
+ zop, kc);
+ if (res != ZEBRA_OK)
+ {
+ int i;
+ for (i = 0; i<num_result_sets_l; i++)
+ rset_delete(result_sets_l[i]);
+ return res;
+ }
+ res = rpn_search_structure(zh, zs->u.complex->s2,
+ attributeSet, stream, rset_nmem,
+ sort_sequence,
+ num_bases, basenames,
+ &result_sets_r, &num_result_sets_r,
+ zop, kc);
+ if (res != ZEBRA_OK)
+ {
+ int i;
+ for (i = 0; i<num_result_sets_l; i++)
+ rset_delete(result_sets_l[i]);
+ for (i = 0; i<num_result_sets_r; i++)
+ rset_delete(result_sets_r[i]);
+ return res;
+ }
+
+ /* make a new list of result for all children */
+ *num_result_sets = num_result_sets_l + num_result_sets_r;
+ *result_sets = nmem_malloc(stream, *num_result_sets *
+ sizeof(**result_sets));
+ memcpy(*result_sets, result_sets_l,
+ num_result_sets_l * sizeof(**result_sets));
+ memcpy(*result_sets + num_result_sets_l, result_sets_r,
+ num_result_sets_r * sizeof(**result_sets));
+
+ if (!parent_op || parent_op->which != zop->which
+ || (zop->which != Z_Operator_and &&
+ zop->which != Z_Operator_or))
+ {
+ /* parent node different from this one (or non-present) */
+ /* we must combine result sets now */
+ RSET rset;
+ switch (zop->which)
+ {
+ case Z_Operator_and:
+ rset = rset_create_and(rset_nmem, kc,
+ kc->scope,
+ *num_result_sets, *result_sets);
+ break;
+ case Z_Operator_or:
+ rset = rset_create_or(rset_nmem, kc,
+ kc->scope, 0, /* termid */
+ *num_result_sets, *result_sets);
+ break;
+ case Z_Operator_and_not:
+ rset = rset_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)
+ {
+ zebra_setError(zh,
+ YAZ_BIB1_UNSUPP_PROX_UNIT_CODE,
+ 0);
+ return ZEBRA_FAIL;
+ }
+ if (*zop->u.prox->u.known != Z_ProxUnit_word)
+ {
+ zebra_setError_zint(zh,
+ YAZ_BIB1_UNSUPP_PROX_UNIT_CODE,
+ *zop->u.prox->u.known);
+ return ZEBRA_FAIL;
+ }
+ else
+ {
+ rset = rset_create_prox(rset_nmem, kc,
+ kc->scope,
+ *num_result_sets, *result_sets,
+ *zop->u.prox->ordered,
+ (!zop->u.prox->exclusion ?
+ 0 : *zop->u.prox->exclusion),
+ *zop->u.prox->relationType,
+ *zop->u.prox->distance );
+ }
+ break;
+ default:
+ zebra_setError(zh, YAZ_BIB1_OPERATOR_UNSUPP, 0);
+ return ZEBRA_FAIL;
+ }
+ *num_result_sets = 1;
+ *result_sets = nmem_malloc(stream, *num_result_sets *
+ sizeof(**result_sets));
+ (*result_sets)[0] = rset;
+ }
}
else if (zs->which == Z_RPNStructure_simple)
{
+ RSET rset;
+ ZEBRA_RES res;
+
if (zs->u.simple->which == Z_Operand_APT)
{
- logf (LOG_DEBUG, "rpn_search_APT");
- r = rpn_search_APT (zh, zs->u.simple->u.attributesPlusTerm,
- attributeSet, stream, sort_sequence,
- num_bases, basenames);
+ 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,
+ kc);
+ if (res != ZEBRA_OK)
+ return res;
}
else if (zs->u.simple->which == Z_Operand_resultSetId)
{
- logf (LOG_DEBUG, "rpn_search_ref");
- r = resultSetRef (zh, zs->u.simple->u.resultSetId);
- if (!r)
+ yaz_log(YLOG_DEBUG, "rpn_search_ref");
+ rset = resultSetRef(zh, zs->u.simple->u.resultSetId);
+ if (!rset)
{
- r = rsnull_create (rset_nmem);
- zh->errCode = 30;
- zh->errString =
- nmem_strdup (stream, zs->u.simple->u.resultSetId);
- return 0;
+ zebra_setError(zh,
+ YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST,
+ zs->u.simple->u.resultSetId);
+ return ZEBRA_FAIL;
}
- else
- rset_dup(r);
+ rset_dup(rset);
}
else
{
- zh->errCode = 3;
- return 0;
+ zebra_setError(zh, YAZ_BIB1_UNSUPP_SEARCH, 0);
+ return ZEBRA_FAIL;
}
+ *num_result_sets = 1;
+ *result_sets = nmem_malloc(stream, *num_result_sets *
+ sizeof(**result_sets));
+ (*result_sets)[0] = rset;
}
else
{
- zh->errCode = 3;
- return 0;
- }
- return r;
-}
-
-
-RSET rpn_search (ZebraHandle zh, NMEM nmem, NMEM rset_nmem,
- Z_RPNQuery *rpn, int num_bases, char **basenames,
- const char *setname,
- ZebraSet sset)
-{
- RSET rset;
- oident *attrset;
- oid_value attributeSet;
- Z_SortKeySpecList *sort_sequence;
- int sort_status, i;
-
- zh->errCode = 0;
- zh->errString = NULL;
- zh->hits = 0;
-
- sort_sequence = (Z_SortKeySpecList *)
- nmem_malloc (nmem, sizeof(*sort_sequence));
- sort_sequence->num_specs = 10;
- sort_sequence->specs = (Z_SortKeySpec **)
- nmem_malloc (nmem, sort_sequence->num_specs *
- sizeof(*sort_sequence->specs));
- for (i = 0; i<sort_sequence->num_specs; i++)
- sort_sequence->specs[i] = 0;
-
- attrset = oid_getentbyoid (rpn->attributeSetId);
- attributeSet = attrset->value;
- rset = rpn_search_structure (zh, rpn->RPNStructure, attributeSet,
- nmem, rset_nmem,
- sort_sequence, num_bases, basenames);
- if (!rset)
- return 0;
-
- if (zh->errCode)
- logf (LOG_DEBUG, "search error: %d", zh->errCode);
-
- for (i = 0; sort_sequence->specs[i]; i++)
- ;
- sort_sequence->num_specs = i;
- if (!i)
- resultSetRank (zh, sset, rset);
- else
- {
- logf (LOG_DEBUG, "resultSetSortSingle in rpn_search");
- resultSetSortSingle (zh, nmem, sset, rset,
- sort_sequence, &sort_status);
- if (zh->errCode)
- {
- logf (LOG_DEBUG, "resultSetSortSingle status = %d", zh->errCode);
- }
+ zebra_setError(zh, YAZ_BIB1_UNSUPP_SEARCH, 0);
+ return ZEBRA_FAIL;
}
- return rset;
+ return ZEBRA_OK;
}
struct scan_info_entry {
char *term;
- ISAMC_P isam_p;
+ ISAM_P isam_p;
};
struct scan_info {
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);
- assert (*info == sizeof(ISAMC_P));
- memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAMC_P));
+ odr_malloc(scan_info->odr, strlen(name + len_prefix)+1);
+ strcpy(scan_info->list[idx].term, name + len_prefix);
+ assert (*info == sizeof(ISAM_P));
+ memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAM_P));
return 0;
}
-static void scan_term_untrans (ZebraHandle zh, NMEM stream, int reg_type,
- char **dst, const char *src)
+void zebra_term_untrans_iconv(ZebraHandle zh, NMEM stream, int reg_type,
+ char **dst, const char *src)
{
char term_src[IT_MAX_WORD];
char term_dst[IT_MAX_WORD];
- term_untrans (zh, reg_type, term_src, src);
+ zebra_term_untrans (zh, reg_type, term_src, src);
if (zh->iconv_from_utf8 != 0)
{
len = 0;
else
len = outbuf - term_dst;
- *dst = nmem_malloc (stream, len + 1);
+ *dst = nmem_malloc(stream, len + 1);
if (len > 0)
memcpy (*dst, term_dst, len);
(*dst)[len] = '\0';
}
else
- *dst = nmem_strdup (stream, term_src);
+ *dst = nmem_strdup(stream, term_src);
}
-static void count_set (RSET r, int *count)
+static void count_set(ZebraHandle zh, RSET rset, zint *count)
{
zint psysno = 0;
- int kno = 0;
struct it_key key;
RSFD rfd;
- logf (LOG_DEBUG, "count_set");
+ yaz_log(YLOG_DEBUG, "count_set");
+
+ rset->hits_limit = zh->approx_limit;
*count = 0;
- rfd = rset_open (r, RSETF_READ);
- while (rset_read (r, rfd, &key))
+ rfd = rset_open(rset, RSETF_READ);
+ while (rset_read(rfd, &key,0 /* never mind terms */))
{
-#if IT_KEY_NEW
if (key.mem[0] != psysno)
{
psysno = key.mem[0];
- (*count)++;
+ if (rfd->counted_items >= rset->hits_limit)
+ break;
}
-#else
- if (key.sysno != psysno)
- {
- psysno = key.sysno;
- (*count)++;
- }
-#endif
- kno++;
}
- rset_close (r, rfd);
- logf (LOG_DEBUG, "%d keys, %d records", kno, *count);
+ rset_close (rfd);
+ *count = rset->hits_count;
}
-void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
- oid_value attributeset,
- int num_bases, char **basenames,
- int *position, int *num_entries, ZebraScanEntry **list,
- int *is_partial, RSET limit_set, int return_zero)
+ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
+ oid_value attributeset,
+ int num_bases, char **basenames,
+ int *position, int *num_entries, ZebraScanEntry **list,
+ int *is_partial, RSET limit_set, int return_zero)
{
int i;
int pos = *position;
int after;
int base_no;
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 errCode = 0; /* err code (if any is not OK) */
char *errString = 0; /* addinfo */
- unsigned reg_id;
+ unsigned index_type;
char *search_type = NULL;
char rank_type[128];
int complete_flag;
int sort_flag;
+ NMEM rset_nmem = NULL;
+ struct rset_key_control *kc = 0;
*list = 0;
+ *is_partial = 0;
if (attributeset == VAL_NONE)
attributeset = VAL_BIB1;
AttrType termset;
int termset_value_numeric;
const char *termset_value_string;
- attr_init (&termset, zapt, 8);
+ attr_init_APT(&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 (LOG_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);
-
- 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, &index_type, &search_type,
+ rank_type, &complete_flag, &sort_flag))
{
*num_entries = 0;
- zh->errCode = 113;
- return ;
+ zebra_setError(zh, YAZ_BIB1_UNSUPP_ATTRIBUTE_TYPE, 0);
+ return ZEBRA_FAIL;
}
- yaz_log (LOG_DEBUG, "use_value = %d", use_value);
-
- if (use_value == -1)
- use_value = 1016;
for (base_no = 0; base_no < num_bases && ord_no < 32; base_no++)
{
- int r;
- attent attp;
- data1_local_attribute *local_attr;
+ int ord;
- 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)
- {
- char val_str[32];
- sprintf (val_str, "%d", use_value);
- errCode = 114;
- errString = odr_strdup (stream, val_str);
- }
- else
- errCode = 121;
- continue;
- }
- if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
- {
- zh->errString = basenames[base_no];
- zh->errCode = 109; /* Database unavailable */
- *num_entries = 0;
- return;
- }
- bases_ok++;
- for (local_attr = attp.local_attributes; local_attr && ord_no < 32;
- local_attr = local_attr->next)
- {
- int ord;
+ if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
+ {
+ zebra_setError(zh, YAZ_BIB1_DATABASE_UNAVAILABLE,
+ basenames[base_no]);
+ *num_entries = 0;
+ return ZEBRA_FAIL;
+ }
- ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- local_attr->local);
- if (ord > 0)
- ords[ord_no++] = ord;
+ if (zebra_apt_get_ord(zh, zapt, index_type, 0, attributeset, &ord)
+ != ZEBRA_OK)
+ {
+ break;
}
+ ords[ord_no++] = ord;
}
if (!bases_ok && errCode)
{
- zh->errCode = errCode;
- zh->errString = errString;
+ zebra_setError(zh, errCode, errString);
*num_entries = 0;
+ return ZEBRA_FAIL;
}
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;
- return;
+ 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));
+ odr_malloc(stream, ord_no * sizeof(*scan_info_array));
for (i = 0; i < ord_no; i++)
{
int j, prefix_len = 0;
struct scan_info *scan_info = scan_info_array + i;
struct rpn_char_map_info rcmi;
- rpn_char_map_prepare (zh->reg, reg_id, &rcmi);
+ rpn_char_map_prepare (zh->reg, index_type, &rcmi);
scan_info->before = before;
scan_info->after = after;
scan_info->odr = stream;
scan_info->list = (struct scan_info_entry *)
- odr_malloc (stream, (before+after) * sizeof(*scan_info->list));
+ odr_malloc(stream, (before+after) * sizeof(*scan_info->list));
for (j = 0; j<before+after; j++)
scan_info->list[j].term = NULL;
prefix_len += key_SU_encode (ords[i], termz + prefix_len);
- termz[prefix_len++] = reg_id;
termz[prefix_len] = 0;
- strcpy (scan_info->prefix, termz);
-
- if (trans_scan_term (zh, zapt, termz+prefix_len, reg_id))
- return ;
-
- dict_scan (zh->reg->dict, termz, &before_tmp, &after_tmp,
- scan_info, scan_handle);
+ strcpy(scan_info->prefix, termz);
+
+ if (trans_scan_term(zh, zapt, termz+prefix_len, index_type) ==
+ ZEBRA_FAIL)
+ return ZEBRA_FAIL;
+
+ dict_scan(zh->reg->dict, termz, &before_tmp, &after_tmp,
+ scan_info, scan_handle);
}
glist = (ZebraScanEntry *)
- odr_malloc (stream, (before+after)*sizeof(*glist));
+ 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 &&
- (tst=scan_info_array[j].list[ptr[j]].term) &&
+ if (ptr[j] < before+after && ptr[j] >= 0 &&
+ (tst = scan_info_array[j].list[ptr[j]].term) &&
(!mterm || strcmp (tst, mterm) < 0))
{
j0 = j;
}
}
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);
-
- 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 */
+ zebra_term_untrans_iconv(zh, stream->mem, index_type,
+ &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, 0, index_type, 0 /* hits_limit */,
+ 0 /* term_ref_id_str */);
+ }
+ 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 &&
- (tst=scan_info_array[j].list[ptr[j]].term) &&
+ if (ptr[j] < before+after && ptr[j] >= 0 &&
+ (tst = scan_info_array[j].list[ptr[j]].term) &&
!strcmp (tst, mterm))
{
- RSET rset2;
-
- rset2 =
- rset_trunc (zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
- glist[i+before].term,
- strlen(glist[i+before].term), NULL, 0,
- zapt->term->which);
-/*
- 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 = rset2;
- rset = rset_create (rset_kind_or, &bool_parms);
-*/
- rset = rsbool_create_or(NULL, /* FIXME -use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- rset, rset2, key_logdump_txt);
-
+ 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, 0, index_type, 0 /* hits_limit */,
+ 0 /* term_ref_id_str */ );
+ rset = rset_create_or(rset_nmem, kc,
+ kc->scope, 0 /* termid */,
+ 2, rsets);
+ }
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);
- */
- rset = rsbool_create_and(NULL, /* FIXME -use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- rset, rset_dup(limit_set),
- key_logdump_txt);
- }
- count_set (rset, &glist[i+before].occurrences);
- rset_delete (rset);
+ if (lo >= 0)
+ {
+ zint count;
+ /* merge with limit_set if given */
+ if (limit_set)
+ {
+ RSET rsets[2];
+ rsets[0] = rset;
+ rsets[1] = rset_dup(limit_set);
+
+ rset = rset_create_and(rset_nmem, kc, kc->scope, 2, rsets);
+ }
+ /* count it */
+ count_set(zh, rset, &count);
+ glist[lo].occurrences = count;
+ rset_delete(rset);
+ }
}
if (i < after)
{
- *num_entries -= (after-i);
- *is_partial = 1;
+ *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;
-
+ ptr[i] = 0;
+
for (i = 0; i<before; i++)
{
- int j, j0 = -1;
- const char *mterm = NULL;
- const char *tst;
- RSET rset;
-
- for (j = 0; j <ord_no; j++)
- {
- if (ptr[j] < before &&
- (tst=scan_info_array[j].list[before-1-ptr[j]].term) &&
- (!mterm || strcmp (tst, mterm) > 0))
- {
- j0 = j;
- mterm = tst;
- }
- }
- if (j0 == -1)
- break;
-
- scan_term_untrans (zh, stream->mem, reg_id,
- &glist[before-1-i].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),
- NULL, 0, zapt->term->which);
-
- ptr[j0]++;
-
- for (j = j0+1; j<ord_no; j++)
- {
- if (ptr[j] < before &&
- (tst=scan_info_array[j].list[before-1-ptr[j]].term) &&
- !strcmp (tst, mterm))
- {
- RSET rset2;
-
- rset2 = rset_trunc (zh,
- &scan_info_array[j].list[before-1-ptr[j]].isam_p, 1,
- glist[before-1-i].term,
- strlen(glist[before-1-i].term), NULL, 0,
- zapt->term->which);
-/*
- 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 = rset2;
- rset = rset_create (rset_kind_or, &bool_parms);
-*/
- rset = rsbool_create_and(NULL, /* FIXME -use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- rset, rset2,
- key_logdump_txt);
- ptr[j]++;
- }
- }
+ int j, j0 = -1;
+ const char *mterm = NULL;
+ const char *tst;
+ RSET rset;
+ int lo = before-1-i; /* offset in result list */
+ zint count;
+
+ for (j = 0; j <ord_no; j++)
+ {
+ if (ptr[j] < before && ptr[j] >= 0 &&
+ (tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
+ (!mterm || strcmp (tst, mterm) > 0))
+ {
+ j0 = j;
+ mterm = tst;
+ }
+ }
+ if (j0 == -1)
+ break;
+
+ zebra_term_untrans_iconv(zh, stream->mem, index_type,
+ &glist[lo].term, mterm);
+
+ rset = rset_trunc
+ (zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1,
+ glist[lo].term, strlen(glist[lo].term),
+ NULL, 0, zapt->term->which, rset_nmem,
+ kc, kc->scope, 0, index_type, 0 /* hits_limit */,
+ 0 /* term_ref_id_str */);
+
+ ptr[j0]++;
+
+ for (j = j0+1; j<ord_no; j++)
+ {
+ if (ptr[j] < before && ptr[j] >= 0 &&
+ (tst = scan_info_array[j].list[before-1-ptr[j]].term) &&
+ !strcmp (tst, mterm))
+ {
+ RSET rsets[2];
+
+ rsets[0] = rset;
+ rsets[1] = rset_trunc(
+ zh,
+ &scan_info_array[j].list[before-1-ptr[j]].isam_p, 1,
+ glist[lo].term,
+ strlen(glist[lo].term), NULL, 0,
+ zapt->term->which, rset_nmem,
+ kc, kc->scope, 0, index_type, 0 /* hits_limit */,
+ 0 /* term_ref_id_str */);
+ rset = rset_create_or(rset_nmem, kc,
+ kc->scope, 0 /* termid */, 2, rsets);
+
+ 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);
- */
- rset = rsbool_create_and(NULL, /* FIXME -use a proper nmem */
- sizeof(struct it_key), key_compare_it,
- rset, rset_dup(limit_set),
- key_logdump_txt);
- }
- count_set (rset, &glist[before-1-i].occurrences);
- rset_delete (rset);
+ {
+ RSET rsets[2];
+ rsets[0] = rset;
+ rsets[1] = rset_dup(limit_set);
+
+ rset = rset_create_and(rset_nmem, kc, kc->scope, 2, rsets);
+ }
+ count_set(zh, rset, &count);
+ glist[lo].occurrences = count;
+ 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;
+ }
}
+
*list = glist + i; /* list is set to first 'real' entry */
- logf (LOG_DEBUG, "position = %d, num_entries = %d",
- *position, *num_entries);
- if (zh->errCode)
- logf (LOG_DEBUG, "scan error: %d", zh->errCode);
+ yaz_log(YLOG_DEBUG, "position = %d, num_entries = %d",
+ *position, *num_entries);
+ return ZEBRA_OK;
}
-
+
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+