-/* $Id: zrpn.c,v 1.152 2004-09-14 14:38:07 quinn Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
- Index Data Aps
+/* $Id: zrpn.c,v 1.175 2005-04-15 10:47:48 adam Exp $
+ Copyright (C) 1995-2005
+ Index Data ApS
This file is part of the Zebra server.
#include <charmap.h>
#include <rset.h>
+/* maximum number of terms in an and/or/phrase item */
+#define TERM_LIST_LENGTH_MAX 256
-static const struct key_control it_ctrl = {
+static const struct key_control it_ctrl =
+{
sizeof(struct it_key),
2, /* we have sysnos and seqnos in this key, nothing more */
key_compare_it,
const struct key_control *key_it_ctrl = &it_ctrl;
-struct rpn_char_map_info {
+struct rpn_char_map_info
+{
ZebraMaps zm;
int reg_type;
};
-typedef struct {
+typedef struct
+{
int type;
int major;
int minor;
} AttrType;
-static const char **rpn_char_map_handler (void *vp, const char **from, int len)
+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)
{
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, 0);
+ 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);
+ dict_grep_cmap(reg->dict, map_info, rpn_char_map_handler);
}
-static int attr_find_ex (AttrType *src, oid_value *attributeSetP,
+static int attr_find_ex(AttrType *src, oid_value *attributeSetP,
const char **string_value)
{
int num_attributes;
{
oident *attrset;
- attrset = oid_getentbyoid (element->attributeSet);
+ attrset = oid_getentbyoid(element->attributeSet);
*attributeSetP = attrset->value;
}
return *element->value.numeric;
{
oident *attrset;
- attrset = oid_getentbyoid (element->attributeSet);
+ attrset = oid_getentbyoid(element->attributeSet);
*attributeSetP = attrset->value;
}
if (element->value.complex->list[src->minor]->which ==
else
break;
default:
- assert (0);
+ assert(0);
}
}
++(src->major);
return -1;
}
-static int attr_find (AttrType *src, oid_value *attributeSetP)
+static int attr_find(AttrType *src, oid_value *attributeSetP)
{
- return attr_find_ex (src, attributeSetP, 0);
+ return attr_find_ex(src, attributeSetP, 0);
}
-static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt,
+static void attr_init(AttrType *src, Z_AttributesPlusTerm *zapt,
int type)
{
src->zapt = zapt;
#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,
+static void 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);
+ const char *cp = zebra_maps_output(zh->reg->zebra_maps,
+ reg_type, &src);
if (!cp && len < IT_MAX_WORD-1)
dst[len++] = *src++;
else
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)
int len = key_SU_decode (&su_code, 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);
+ yaz_log(log_level_rpn, "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);
+ yaz_log(log_level_rpn, "grep: set=%d use=%d db=%s", set, use, db);
- resultSetAddTerm (p->zh, p->termset, name[len], db,
- set, use, term_tmp);
+ resultSetAddTerm(p->zh, p->termset, name[len], db,
+ set, use, 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,
+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;
/* 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), first);
+ 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, int 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,
+/* 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, !space_split))
+ 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), 0);
+ const char *s1 = s0;
+ int q_map_match = 0;
+ map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0),
+ &q_map_match);
if (space_split)
{
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,
+/* 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, "#", "#", !space_split))
+ 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), 0);
+ const char *s1 = s0;
+ int q_map_match = 0;
+ map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0),
+ &q_map_match);
if (space_split && **map == *CHR_SPACE)
break;
- while (s1 < s0)
- {
- if (strchr(REGEX_CHARS, *s1))
- dst[i++] = '\\';
- dst_term[j++] = *s1;
- dst[i++] = *s1++;
- }
+
+ /* add non-space char */
+ memcpy(dst_term+j, s1, s0 - s1);
+ j += (s0 - s1);
+ if (!q_map_match)
+ {
+ while (s1 < s0)
+ {
+ if (strchr(REGEX_CHARS, *s1))
+ dst[i++] = '\\';
+ dst[i++] = *s1++;
+ }
+ }
+ else
+ {
+ char tmpbuf[80];
+ esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0]));
+
+ strcpy(dst + i, map[0]);
+ i += strlen(map[0]);
+ }
}
}
dst[i] = '\0';
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,
+/* 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, "^\\()[].*+?|", "(", !space_split))
+ 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), 0);
- 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) */
+/* 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,
+ 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,
+/* 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)
{
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "?*#", "?*#", !space_split))
+ 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), 0);
+ map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0);
if (space_split && **map == *CHR_SPACE)
break;
while (s1 < s0)
return i;
}
-/* term_105/106: handle term, where trunc=Process * and ! and right trunc */
+/* 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)
int i = 0;
int j = 0;
- if (!term_pre (zebra_maps, reg_type, src, "*!", "*!", !space_split))
+ 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), 0);
+ map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0);
if (space_split && **map == *CHR_SPACE)
break;
while (s1 < s0)
* 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(&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:
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]; )
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 (!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 (!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];)
*term_tmp = '\0';
break;
case 3:
- default:
- logf (LOG_DEBUG, "Relation =");
+ case 102:
+ case -1:
+ yaz_log(log_level_rpn, "Relation =");
if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub,
term_component, space_split, term_dst))
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 = 117;
+ 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,
- NMEM rset_nmem)
+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, int xpath_use);
+
+static RSET term_trunc(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ oid_value attributeSet, NMEM stream,
+ struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst,
+ const char *rank_type, int xpath_use,
+ NMEM rset_nmem)
{
- int r;
+ ZEBRA_RES res;
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)
+ res = string_term(zh, zapt, term_sub, attributeSet, stream, grep_info,
+ reg_type, complete_flag, num_bases, basenames,
+ term_dst, xpath_use);
+ if (res != ZEBRA_OK)
return 0;
- 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, rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
+ if (!*term_sub) /* no more terms ? */
+ return 0;
+ yaz_log(log_level_rpn, "term: %s", term_dst);
+ return rset_trunc(zh, grep_info->isam_p_buf,
+ grep_info->isam_p_indx, term_dst,
+ strlen(term_dst), rank_type, 1 /* preserve pos */,
+ zapt->term->which, rset_nmem,
+ key_it_ctrl, key_it_ctrl->scope);
+}
+static char *nmem_strdup_i(NMEM nmem, int v)
+{
+ char val_str[64];
+ sprintf (val_str, "%d", v);
+ return nmem_strdup(nmem, val_str);
}
-
-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, int xpath_use)
{
char term_dict[2*IT_MAX_WORD+4000];
int j, r, base_no;
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);
+ yaz_log(log_level_rpn, "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);
+ yaz_log(log_level_rpn, "truncation value %d", truncation_value);
if (use_value == -1) /* no attribute - assumy "any" */
use_value = 1016;
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;
attent attp;
data1_local_attribute id_xpath_attr;
data1_local_attribute *local_attr;
int max_pos, prefix_len = 0;
+ int relation_error;
termp = *term_sub;
{
zh->errCode = 109; /* Database unavailable */
zh->errString = basenames[base_no];
- return -1;
+ return ZEBRA_FAIL;
}
if (xpath_use > 0 && use_value == -2)
{
- use_value = xpath_use;
+ /* xpath mode and we have a string attribute */
attp.local_attributes = &id_xpath_attr;
attp.attset_ordinal = VAL_IDXPATH;
id_xpath_attr.next = 0;
+
+ use_value = xpath_use; /* xpath_use as use-attribute now */
id_xpath_attr.local = use_value;
}
- else if (curAttributeSet == VAL_IDXPATH)
+ else if (curAttributeSet == VAL_IDXPATH && use_value >= 0)
{
+ /* X-Path attribute, use numeric value directly */
attp.local_attributes = &id_xpath_attr;
attp.attset_ordinal = VAL_IDXPATH;
id_xpath_attr.next = 0;
id_xpath_attr.local = use_value;
}
- else
+ else if (use_string &&
+ (ord = zebraExplain_lookup_attr_str(zh->reg->zei,
+ use_string)) >= 0)
+ {
+ /* we have a match for a raw string attribute */
+ char ord_buf[32];
+ int i, ord_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++)
+ {
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = ord_buf[i];
+ }
+ attp.local_attributes = 0; /* no more attributes */
+ }
+ else
{
- if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value,
- use_string)))
+ /* lookup in the .att files . Allow string as well */
+ 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);
+ yaz_log(YLOG_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 */
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);
- }
+ errString = nmem_strdup_i (stream, use_value);
}
else
{
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;
- }
- bases_ok++; /* this has OK attributes */
+ for (local_attr = attp.local_attributes; local_attr;
+ local_attr = local_attr->next)
+ {
+ char ord_buf[32];
+ int i, ord_len;
+
+ ord = zebraExplain_lookup_attr_su(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];
+ }
+ }
+ 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]);
+ yaz_log(log_level_rpn, "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_DEBUG, "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)
+ {
+ zh->errCode = relation_error;
+ 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;
term_dict[j++] = '(';
+ init_pos = 2;
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;
+ &termp, term_dict + j, ®ex_range,
+ space_split, term_dst))
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
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);
+ {
+ *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);
+ {
+ *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);
+ {
+ *term_sub = 0;
+ return ZEBRA_OK;
+ }
+ strcat(term_dict, ")");
break;
- }
+ default:
+ zh->errCode = 120;
+ zh->errString = nmem_strdup_i(stream, 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 SCAN term to internal cmap */
-static int trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+static int trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
char *termz, int reg_type)
{
char termz0[IT_MAX_WORD];
return nmem_strdup(stream, termz);
else
{
- char *buf = (char*) nmem_malloc (stream, wrbuf_len(wrbuf)+1);
+ 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;
}
else
termset_name = termset_value_string;
- logf (LOG_LOG, "creating termset set %s", termset_name);
+ yaz_log(log_level_rpn, "creating termset set %s", termset_name);
grep_info->termset = resultSetAdd (zh, termset_name, 1);
if (!grep_info->termset)
{
}
-static RSET rpn_search_APT_phrase (ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type, int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static RSET rpn_search_APT_phrase(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type, int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem)
{
char term_dst[IT_MAX_WORD+1];
- RSET rset[60], result;
- int rset_no = 0;
+ RSET rset[TERM_LIST_LENGTH_MAX], result;
+ size_t rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
*term_dst = 0;
if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
return 0;
- while (1)
+ for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
{
- 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,rset_nmem);
- if (!rset[rset_no])
- break;
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
+ yaz_log(log_level_rpn, "APT_phrase termp=%s", termp);
+ rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use, rset_nmem);
+ if (!rset[rset_no]) /* error or no more terms */
+ break;
}
grep_info_delete (&grep_info);
+ if (termp)
+ {
+ int i;
+ for (i = 0; i<rset_no; i++)
+ rset_delete(rset[i]);
+ return 0;
+ }
if (rset_no == 0)
return rsnull_create (rset_nmem,key_it_ctrl);
else if (rset_no == 1)
return result;
}
-static RSET rpn_search_APT_or_list (ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type,
- int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static RSET rpn_search_APT_or_list(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type,
+ int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem)
{
char term_dst[IT_MAX_WORD+1];
- RSET rset[60];
- int rset_no = 0;
+ RSET rset[TERM_LIST_LENGTH_MAX];
+ size_t rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
return 0;
- while (1)
+ for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
{
- 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,rset_nmem);
- if (!rset[rset_no])
- break;
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
+ yaz_log(log_level_rpn, "APT_or_list termp=%s", termp);
+ rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use, rset_nmem);
+ if (!rset[rset_no]) /* error or no more terms */
+ break;
}
grep_info_delete (&grep_info);
+ if (termp)
+ {
+ int i;
+ for (i = 0; i<rset_no; i++)
+ rset_delete(rset[i]);
+ return 0;
+ }
if (rset_no == 0)
return rsnull_create (rset_nmem,key_it_ctrl);
- return rsmultior_create(rset_nmem, key_it_ctrl,key_it_ctrl->scope,
- rset_no, rset);
+ return rsmulti_or_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
+ rset_no, rset);
}
-static RSET rpn_search_APT_and_list (ZebraHandle zh,
- Z_AttributesPlusTerm *zapt,
- const char *termz_org,
- oid_value attributeSet,
- NMEM stream,
- int reg_type, int complete_flag,
- const char *rank_type,
- int xpath_use,
- int num_bases, char **basenames,
- NMEM rset_nmem)
+static RSET rpn_search_APT_and_list(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz_org,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type,
+ int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem)
{
char term_dst[IT_MAX_WORD+1];
- RSET rset[60], result;
- int i, rset_no = 0;
+ RSET rset[TERM_LIST_LENGTH_MAX];
+ size_t rset_no = 0;
struct grep_info grep_info;
char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
const char *termp = termz;
- if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
+ if (grep_info_prepare(zh, zapt, &grep_info, reg_type, stream))
return 0;
- while (1)
+ for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
{
- 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, rset_nmem);
- if (!rset[rset_no])
- break;
- assert (rset[rset_no]);
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
+ yaz_log(log_level_rpn, "APT_and_list termp=%s", termp);
+ rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet,
+ stream, &grep_info,
+ reg_type, complete_flag,
+ num_bases, basenames,
+ term_dst, rank_type,
+ xpath_use, rset_nmem);
+ if (!rset[rset_no]) /* error or no more terms */
+ break;
}
grep_info_delete (&grep_info);
+ if (termp)
+ {
+ int i;
+ for (i = 0; i<rset_no; i++)
+ rset_delete(rset[i]);
+ return 0;
+ }
if (rset_no == 0)
- return rsnull_create (rset_nmem,key_it_ctrl);
- result = rset[0];
- /* FIXME - Use a proper rsmultiand */
- for (i = 1; i<rset_no; i++)
- result= rsbool_create_and(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- result, rset[i] );
- return result;
+ return rsnull_create(rset_nmem, key_it_ctrl);
+
+ return rsmulti_and_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
+ rset_no, rset);
}
-static 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);
+ *error_code = 0;
attr_init (&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 = 117;
+ 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, int xpath_use, NMEM stream)
{
char term_dict[2*IT_MAX_WORD+2];
int r, base_no;
data1_local_attribute id_xpath_attr;
data1_local_attribute *local_attr;
int max_pos, prefix_len = 0;
+ int relation_error = 0;
termp = *term_sub;
if (use_value == -2) /* string attribute (assume IDXPATH/any) */
}
else
{
- if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value,
+ if ((r = att_getentbyatt (zh, &attp, curAttributeSet, use_value,
use_string)))
{
- logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d",
+ yaz_log(YLOG_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;
+ if (use_string)
+ errString = nmem_strdup(stream, use_string);
+ else
+ errString = nmem_strdup_i (stream, use_value);
}
else
errCode = 121;
char ord_buf[32];
int i, ord_len;
- ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- local_attr->local);
+ ord = zebraExplain_lookup_attr_su(zh->reg->zei,
+ attp.attset_ordinal,
+ local_attr->local);
if (ord < 0)
continue;
if (prefix_len)
}
if (!prefix_len)
{
- char val_str[32];
- sprintf (val_str, "%d", use_value);
errCode = 114;
- errString = nmem_strdup (stream, val_str);
+ errString = nmem_strdup_i(stream, use_value);
continue;
}
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]);
+ yaz_log(YLOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]);
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)
+ {
+ zh->errCode = relation_error;
+ zh->errString = 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,
- NMEM rset_nmem)
+static RSET rpn_search_APT_numeric(ZebraHandle zh,
+ Z_AttributesPlusTerm *zapt,
+ const char *termz,
+ oid_value attributeSet,
+ NMEM stream,
+ int reg_type, int complete_flag,
+ const char *rank_type, int xpath_use,
+ int num_bases, char **basenames,
+ NMEM rset_nmem)
{
char term_dst[IT_MAX_WORD+1];
const char *termp = termz;
- RSET rset[60], result;
- int i, r, rset_no = 0;
+ RSET rset[TERM_LIST_LENGTH_MAX];
+ ZEBRA_RES r;
+ size_t rset_no = 0;
struct grep_info grep_info;
+ yaz_log(log_level_rpn, "APT_numeric t='%s'",termz);
if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
return 0;
- while (1)
+ for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++)
{
- logf (LOG_DEBUG, "APT_numeric termp=%s", termp);
+ 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,
+ r = numeric_term(zh, zapt, &termp, attributeSet, &grep_info,
+ reg_type, complete_flag, num_bases, basenames,
+ term_dst, xpath_use,
+ stream);
+ if (r == ZEBRA_FAIL || termp == 0)
+ break;
+ yaz_log(YLOG_DEBUG, "term: %s", term_dst);
+ rset[rset_no] = rset_trunc(zh, grep_info.isam_p_buf,
grep_info.isam_p_indx, term_dst,
strlen(term_dst), rank_type,
0 /* preserve position */,
zapt->term->which, rset_nmem,
key_it_ctrl,key_it_ctrl->scope);
- assert (rset[rset_no]);
- if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset)))
- break;
+ if (!rset[rset_no])
+ break;
}
grep_info_delete (&grep_info);
- if (rset_no == 0)
- return rsnull_create (rset_nmem,key_it_ctrl);
- result = rset[0];
- for (i = 1; i<rset_no; i++)
+ if (termp)
{
- /* FIXME - Use a proper multi-and */
- result= rsbool_create_and(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- result, rset[i] );
+ int i;
+ for (i = 0; i<rset_no; i++)
+ rset_delete(rset[i]);
+ return 0;
}
- return result;
+ if (rset_no == 0)
+ return rsnull_create(rset_nmem,key_it_ctrl);
+ if (rset_no == 1)
+ return rset[0];
+ return rsmulti_and_create(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
+ rset_no, rset);
}
static RSET rpn_search_APT_local (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
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( rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- res_get (zh->res, "setTmpDir") );
+ res_get (zh->res, "setTmpDir"),0 );
rsfd = rset_open (result, 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 (rsfd, &key);
rset_close (rsfd);
return result;
{
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;
if (!oid_ent_to_oid (&oe, oid))
return 0;
- 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));
+ 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));
+ 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));
+ nmem_malloc(stream, sizeof(**sk->u.sortAttributes->list->attributes));
ae->attributeSet = 0;
ae->attributeType = (int *)
- nmem_malloc (stream, sizeof(*ae->attributeType));
+ 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));
+ nmem_malloc(stream, sizeof(*ae->value.numeric));
*ae->value.numeric = use_value;
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;
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_su(zh->reg->zei, curAttributeSet, use);
int ord_len, i, r, max_pos;
int term_type = Z_Term_characterString;
const char *flags = "void";
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,
+ rset = rset_trunc(zh, grep_info.isam_p_buf,
grep_info.isam_p_indx, term, strlen(term),
flags, 1, term_type,rset_nmem,
key_it_ctrl, key_it_ctrl->scope);
if (xpath_len < 0)
return rset;
- 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)
*/
}
}
wrbuf_puts(wbuf, "");
- rset_attr = xpath_trunc (
+ rset_attr = xpath_trunc(
zh, stream, '0', wrbuf_buf(wbuf), 3,
curAttributeSet,rset_nmem);
wrbuf_free(wbuf, 1);
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',
rset_end_tag = xpath_trunc(zh, stream, '0',
xpath_rev, 2, curAttributeSet, rset_nmem);
- /*
- 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( rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- rset_start_tag, rset, rset_end_tag, rset_attr);
+ rset = rsbetween_create(rset_nmem, key_it_ctrl,
+ key_it_ctrl->scope,
+ rset_start_tag, rset,
+ rset_end_tag, rset_attr);
}
first_path = 0;
}
-static RSET 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)
+static RSET 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)
{
unsigned reg_id;
char *search_type = NULL;
int xpath_use = 0;
struct xpath_location_step xpath[10];
- zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type,
- rank_type, &complete_flag, &sort_flag);
+ 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 (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);
xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream);
if (xpath_len >= 0)
{
if (!strcmp (search_type, "phrase"))
{
- rset = rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream,
- reg_id, complete_flag, rank_type,
- xpath_use,
- num_bases, basenames, rset_nmem);
+ rset = rpn_search_APT_phrase(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem);
}
else if (!strcmp (search_type, "and-list"))
{
- rset = rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream,
- reg_id, complete_flag, rank_type,
- xpath_use,
- num_bases, basenames, rset_nmem);
+ rset = rpn_search_APT_and_list(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem);
}
else if (!strcmp (search_type, "or-list"))
{
- rset = rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream,
- reg_id, complete_flag, rank_type,
- xpath_use,
- num_bases, basenames, rset_nmem);
+ rset = rpn_search_APT_or_list(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem);
}
else if (!strcmp (search_type, "local"))
{
- rset = rpn_search_APT_local (zh, zapt, termz, attributeSet, stream,
- rank_type, rset_nmem);
+ rset = rpn_search_APT_local(zh, zapt, termz, attributeSet, stream,
+ rank_type, rset_nmem);
}
else if (!strcmp (search_type, "numeric"))
{
- rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream,
- reg_id, complete_flag, rank_type,
- xpath_use,
- num_bases, basenames, rset_nmem);
+ rset = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, stream,
+ reg_id, complete_flag, rank_type,
+ xpath_use,
+ num_bases, basenames, rset_nmem);
}
else if (!strcmp (search_type, "always"))
{
rset = 0;
}
else
+ {
zh->errCode = 118;
+ return 0;
+ }
return rpn_search_xpath (zh, attributeSet, num_bases, basenames,
stream, rank_type, rset,
xpath_len, xpath, rset_nmem);
}
-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)
+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)
{
RSET r = NULL;
if (zs->which == Z_RPNStructure_complex)
{
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)
+ RSET rsets[2]; /* l and r argument */
+
+ rsets[0] = rpn_search_structure(zh, zs->u.complex->s1,
+ attributeSet, stream, rset_nmem,
+ sort_sequence,
+ num_bases, basenames);
+ if (rsets[0] == 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)
+ rsets[1] = rpn_search_structure(zh, zs->u.complex->s2,
+ attributeSet, stream, rset_nmem,
+ sort_sequence,
+ num_bases, basenames);
+ if (rsets[1] == NULL)
{
- rset_delete (rset_l);
+ rset_delete(rsets[0]);
return NULL;
}
switch (zop->which)
{
case Z_Operator_and:
- r = rsbool_create_and(rset_nmem,key_it_ctrl, key_it_ctrl->scope,
- rset_l,rset_r );
+ r = rsmulti_and_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
+ 2, rsets);
break;
case Z_Operator_or:
- r = rsbool_create_or(rset_nmem,key_it_ctrl, key_it_ctrl->scope,
- rset_l,rset_r );
+ r = rsmulti_or_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope,
+ 2, rsets);
break;
case Z_Operator_and_not:
r = rsbool_create_not(rset_nmem,key_it_ctrl, key_it_ctrl->scope,
- rset_l,rset_r );
+ rsets[0], rsets[1]);
break;
case Z_Operator_prox:
if (zop->u.prox->which != Z_ProximityOperator_known)
}
if (*zop->u.prox->u.known != Z_ProxUnit_word)
{
- char *val = (char *) nmem_malloc (stream, 16);
+ char *val = (char *) nmem_malloc(stream, 16);
zh->errCode = 132;
zh->errString = val;
sprintf (val, "%d", *zop->u.prox->u.known);
else
{
/* new / old prox */
- RSET twosets[2];
-
- twosets[0] = rset_l;
- twosets[1] = rset_r;
- r=rsprox_create(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- 2, twosets,
+ r = rsprox_create(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
+ 2, rsets,
*zop->u.prox->ordered,
(!zop->u.prox->exclusion ?
0 : *zop->u.prox->exclusion),
{
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,rset_nmem);
+ yaz_log(YLOG_DEBUG, "rpn_search_APT");
+ r = rpn_search_APT(zh, zs->u.simple->u.attributesPlusTerm,
+ attributeSet, stream, sort_sequence,
+ num_bases, basenames,rset_nmem);
}
else if (zs->u.simple->which == Z_Operand_resultSetId)
{
- logf (LOG_DEBUG, "rpn_search_ref");
+ yaz_log(YLOG_DEBUG, "rpn_search_ref");
r = resultSetRef (zh, zs->u.simple->u.resultSetId);
if (!r)
{
- r = rsnull_create (rset_nmem,key_it_ctrl);
zh->errCode = 30;
zh->errString =
nmem_strdup (stream, zs->u.simple->u.resultSetId);
return 0;
}
- else
- rset_dup(r);
+ rset_dup(r);
}
else
{
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);
- }
- }
- return rset;
-}
-
struct scan_info_entry {
char *term;
- ISAMC_P isam_p;
+ ISAM_P isam_p;
};
struct scan_info {
else
idx = - pos - 1;
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;
}
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)
struct it_key key;
RSFD rfd;
- logf (LOG_DEBUG, "count_set");
+ yaz_log(YLOG_DEBUG, "count_set");
*count = 0;
rfd = rset_open (r, RSETF_READ);
- while (rset_read (rfd, &key))
+ while (rset_read (rfd, &key,0 /* never mind terms */))
{
-#if IT_KEY_NEW
if (key.mem[0] != psysno)
{
psysno = key.mem[0];
(*count)++;
}
-#else
- if (key.sysno != psysno)
- {
- psysno = key.sysno;
- (*count)++;
- }
-#endif
kno++;
}
rset_close (rfd);
- logf (LOG_DEBUG, "%d keys, %d records", kno, *count);
+ yaz_log(YLOG_DEBUG, "%d keys, %d records", kno, *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;
char rank_type[128];
int complete_flag;
int sort_flag;
- NMEM rset_nmem=NULL;
+ NMEM rset_nmem = NULL;
*list = 0;
+ *is_partial = 0;
if (attributeset == VAL_NONE)
attributeset = VAL_BIB1;
}
}
- yaz_log (LOG_DEBUG, "position = %d, num = %d set=%d",
+ yaz_log (YLOG_DEBUG, "position = %d, num = %d set=%d",
pos, num, attributeset);
attr_init (&use, zapt, 1);
{
*num_entries = 0;
zh->errCode = 113;
- return ;
+ return ZEBRA_FAIL;
}
- yaz_log (LOG_DEBUG, "use_value = %d", use_value);
+ yaz_log (YLOG_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;
-
- 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;
-
- ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal,
- local_attr->local);
- if (ord > 0)
- ords[ord_no++] = ord;
- }
+ data1_local_attribute *local_attr;
+ attent attp;
+ int ord;
+
+ if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no]))
+ {
+ zh->errString = basenames[base_no];
+ zh->errCode = 109; /* Database unavailable */
+ *num_entries = 0;
+ return ZEBRA_FAIL;
+ }
+
+ if (use_string &&
+ (ord = zebraExplain_lookup_attr_str(zh->reg->zei,
+ use_string)) >= 0)
+ {
+ /* we have a match for a raw string attribute */
+ if (ord > 0)
+ ords[ord_no++] = ord;
+ attp.local_attributes = 0; /* no more attributes */
+ }
+ else
+ {
+ int r;
+
+ if ((r = att_getentbyatt (zh, &attp, attributeset, use_value,
+ use_string)))
+ {
+ yaz_log(YLOG_DEBUG, "att_getentbyatt fail. set=%d use=%d",
+ attributeset, use_value);
+ if (r == -1)
+ {
+ errCode = 114;
+ if (use_string)
+ errString = odr_strdup(stream, use_string);
+ else
+ {
+ char val_str[32];
+ sprintf (val_str, "%d", use_value);
+ errString = odr_strdup(stream, val_str);
+ }
+ }
+ else
+ errCode = 121;
+ continue;
+ }
+ }
+ bases_ok++;
+ for (local_attr = attp.local_attributes; local_attr && ord_no < 32;
+ local_attr = local_attr->next)
+ {
+ ord = zebraExplain_lookup_attr_su(zh->reg->zei,
+ attp.attset_ordinal,
+ local_attr->local);
+ if (ord > 0)
+ ords[ord_no++] = ord;
+ }
}
if (!bases_ok && errCode)
{
zh->errCode = errCode;
zh->errString = 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 */
before = pos-1;
after = 1+num-pos;
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;
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);
+ 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);
+ if (trans_scan_term(zh, zapt, termz+prefix_len, reg_id))
+ 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();
+ rset_nmem = nmem_create();
/* consider terms after main term */
for (i = 0; i < ord_no; i++)
for (j = 0; j < ord_no; j++)
{
if (ptr[j] < before+after &&
- (tst=scan_info_array[j].list[ptr[j]].term) &&
+ (tst = scan_info_array[j].list[ptr[j]].term) &&
(!mterm || strcmp (tst, mterm) < 0))
{
j0 = j;
break;
scan_term_untrans (zh, stream->mem, reg_id,
&glist[i+before].term, mterm);
- rset = rset_trunc (zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
- glist[i+before].term, strlen(glist[i+before].term),
- NULL, 0, zapt->term->which, rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
+ rset = rset_trunc(zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
+ glist[i+before].term, strlen(glist[i+before].term),
+ NULL, 0, zapt->term->which, rset_nmem,
+ key_it_ctrl,key_it_ctrl->scope);
ptr[j0]++;
for (j = j0+1; j<ord_no; j++)
{
if (ptr[j] < before+after &&
- (tst=scan_info_array[j].list[ptr[j]].term) &&
+ (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,rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
- rset = rsbool_create_or(rset_nmem,key_it_ctrl,
- key_it_ctrl->scope, rset, rset2);
- /* FIXME - Use a proper multi-or */
-
+ RSET rsets[2];
+
+ rsets[0] = rset;
+ rsets[1] =
+ rset_trunc(zh, &scan_info_array[j].list[ptr[j]].isam_p, 1,
+ glist[i+before].term,
+ strlen(glist[i+before].term), NULL, 0,
+ zapt->term->which,rset_nmem,
+ key_it_ctrl, key_it_ctrl->scope);
+ rset = rsmulti_or_create(rset_nmem, key_it_ctrl,
+ 2, key_it_ctrl->scope, rsets);
ptr[j]++;
}
}
if (limit_set)
- rset = rsbool_create_and(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- rset, rset_dup(limit_set));
- count_set (rset, &glist[i+before].occurrences);
- rset_delete (rset);
+ {
+ RSET rsets[2];
+ rsets[0] = rset;
+ rsets[1] = rset_dup(limit_set);
+
+ rset = rsmulti_and_create(rset_nmem, key_it_ctrl,
+ key_it_ctrl->scope, 2, rsets);
+ }
+ count_set(rset, &glist[i+before].occurrences);
+ rset_delete(rset);
}
if (i < after)
{
- *num_entries -= (after-i);
- *is_partial = 1;
+ *num_entries -= (after-i);
+ *is_partial = 1;
}
-
+
/* 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,rset_nmem,
- key_it_ctrl,key_it_ctrl->scope);
-
- 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, rset_nmem,
- key_it_ctrl, key_it_ctrl->scope);
- rset = rsbool_create_and(rset_nmem,key_it_ctrl,
- key_it_ctrl->scope, rset, rset2);
- /* FIXME - multi-and ?? */
- ptr[j]++;
- }
- }
+ 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,rset_nmem,
+ key_it_ctrl,key_it_ctrl->scope);
+
+ 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 rsets[2];
+
+ rsets[0] = rset;
+ rsets[1] = rset_trunc(
+ zh,
+ &scan_info_array[j].list[before-1-ptr[j]].isam_p, 1,
+ glist[before-1-i].term,
+ strlen(glist[before-1-i].term), NULL, 0,
+ zapt->term->which, rset_nmem,
+ key_it_ctrl, key_it_ctrl->scope);
+ rset = rsmulti_or_create(rset_nmem, key_it_ctrl,
+ 2, key_it_ctrl->scope, rsets);
+
+ ptr[j]++;
+ }
+ }
if (limit_set)
- rset = rsbool_create_and(rset_nmem,key_it_ctrl,key_it_ctrl->scope,
- rset, rset_dup(limit_set));
- count_set (rset, &glist[before-1-i].occurrences);
- rset_delete (rset);
+ {
+ RSET rsets[2];
+ rsets[0] = rset;
+ rsets[1] = rset_dup(limit_set);
+
+ rset = rsmulti_and_create(rset_nmem, key_it_ctrl,
+ key_it_ctrl->scope, 2, rsets);
+ }
+ count_set (rset, &glist[before-1-i].occurrences);
+ rset_delete (rset);
}
i = before-i;
if (i)
*position -= i;
*num_entries -= i;
}
-
+
nmem_destroy(rset_nmem);
*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;
}
-
+
projects / idzebra-moved-to-github.git / blobdiff