-/* $Id: recgrs.c,v 1.92 2004-10-12 18:21:35 quinn Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
- Index Data Aps
+/* $Id: recgrs.c,v 1.110 2006-05-19 13:49:35 adam Exp $
+ Copyright (C) 1995-2006
+ Index Data ApS
This file is part of the Zebra server.
#include <stdio.h>
#include <assert.h>
#include <sys/types.h>
-#ifndef WIN32
-#include <unistd.h>
-#include <dlfcn.h>
-#endif
+#include <ctype.h>
#include <yaz/log.h>
#include <yaz/oid.h>
const char *tok;
const char *src;
int lookahead;
+ NMEM nmem;
};
static int sp_lex(struct source_parser *sp)
return sp->lookahead;
}
+static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd);
+
+static int sp_range(struct source_parser *sp, data1_node *n, RecWord *wrd)
+{
+ int start, len;
+ RecWord tmp_w;
+
+ /* ( */
+ sp_lex(sp);
+ if (sp->lookahead != '(')
+ return 0;
+ sp_lex(sp); /* skip ( */
+
+ /* 1st arg: string */
+ if (!sp_expr(sp, n, wrd))
+ return 0;
+
+ if (sp->lookahead != ',')
+ return 0;
+ sp_lex(sp); /* skip , */
+
+ /* 2nd arg: start */
+ if (!sp_expr(sp, n, &tmp_w))
+ return 0;
+ start = atoi_n(tmp_w.term_buf, tmp_w.term_len);
+
+ if (sp->lookahead == ',')
+ {
+ sp_lex(sp); /* skip , */
+
+ /* 3rd arg: length */
+ if (!sp_expr(sp, n, &tmp_w))
+ return 0;
+ len = atoi_n(tmp_w.term_buf, tmp_w.term_len);
+ }
+ else
+ len = wrd->term_len;
+
+ /* ) */
+ if (sp->lookahead != ')')
+ return 0;
+ sp_lex(sp);
+
+ if (wrd->term_buf && wrd->term_len)
+ {
+ wrd->term_buf += start;
+ wrd->term_len -= start;
+ if (wrd->term_len > len)
+ wrd->term_len = len;
+ }
+ return 1;
+}
+
+static int sp_first(struct source_parser *sp, data1_node *n, RecWord *wrd)
+{
+ char num_str[20];
+ int min_pos = -1;
+ sp_lex(sp);
+ if (sp->lookahead != '(')
+ return 0;
+ sp_lex(sp); /* skip ( */
+ if (!sp_expr(sp, n, wrd))
+ return 0;
+ while (sp->lookahead == ',')
+ {
+ RecWord search_w;
+ int i;
+ sp_lex(sp); /* skip , */
+
+ if (!sp_expr(sp, n, &search_w))
+ return 0;
+ for (i = 0; i<wrd->term_len; i++)
+ {
+ int j;
+ for (j = 0; j<search_w.term_len && i+j < wrd->term_len; j++)
+ if (wrd->term_buf[i+j] != search_w.term_buf[j])
+ break;
+ if (j == search_w.term_len) /* match ? */
+ {
+ if (min_pos == -1 || i < min_pos)
+ min_pos = i;
+ break;
+ }
+ }
+ }
+ if (sp->lookahead != ')')
+ return 0;
+ sp_lex(sp);
+ if (min_pos == -1)
+ min_pos = 0; /* the default if not found */
+ sprintf(num_str, "%d", min_pos);
+ wrd->term_buf = nmem_strdup(sp->nmem, num_str);
+ wrd->term_len = strlen(wrd->term_buf);
+ return 1;
+}
static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd)
{
{
if (n->which == DATA1N_data)
{
- wrd->string = n->u.data.data;
- wrd->length = n->u.data.len;
+ wrd->term_buf = n->u.data.data;
+ wrd->term_len = n->u.data.len;
}
sp_lex(sp);
}
{
if (n->which == DATA1N_tag)
{
- wrd->string = n->u.tag.tag;
- wrd->length = strlen(n->u.tag.tag);
+ wrd->term_buf = n->u.tag.tag;
+ wrd->term_len = strlen(n->u.tag.tag);
}
sp_lex(sp);
}
else if (sp->len == 4 && !memcmp(sp->tok, "attr", sp->len))
{
+ RecWord tmp_w;
sp_lex(sp);
if (sp->lookahead != '(')
return 0;
sp_lex(sp);
- if (sp->lookahead != 't')
+
+ if (!sp_expr(sp, n, &tmp_w))
return 0;
+ wrd->term_buf = "";
+ wrd->term_len = 0;
if (n->which == DATA1N_tag)
{
data1_xattr *p = n->u.tag.attributes;
- while (p && strlen(p->name) != sp->len &&
- memcmp (p->name, sp->tok, sp->len))
+ while (p && strlen(p->name) != tmp_w.term_len &&
+ memcmp (p->name, tmp_w.term_buf, tmp_w.term_len))
p = p->next;
if (p)
{
- wrd->string = p->value;
- wrd->length = strlen(p->value);
+ wrd->term_buf = p->value;
+ wrd->term_len = strlen(p->value);
}
}
- sp_lex(sp);
if (sp->lookahead != ')')
return 0;
sp_lex(sp);
}
+ else if (sp->len == 5 && !memcmp(sp->tok, "first", sp->len))
+ {
+ return sp_first(sp, n, wrd);
+ }
else if (sp->len == 5 && !memcmp(sp->tok, "range", sp->len))
{
- int start, len;
- sp_lex(sp);
- if (sp->lookahead != '(')
- return 0;
-
- sp_lex(sp);
- sp_expr(sp, n, wrd);
- if (sp->lookahead != ',')
- return 0;
-
- sp_lex(sp);
- if (sp->lookahead != 't')
- return 0;
- start = atoi_n(sp->tok, sp->len);
-
- sp_lex(sp);
- if (sp->lookahead != ',')
- return 0;
-
+ return sp_range(sp, n, wrd);
+ }
+ else if (sp->len > 0 && isdigit(*(unsigned char *)sp->tok))
+ {
+ char *b;
+ wrd->term_len = sp->len;
+ b = nmem_malloc(sp->nmem, sp->len);
+ memcpy(b, sp->tok, sp->len);
+ wrd->term_buf = b;
sp_lex(sp);
- if (sp->lookahead != 't')
- return 0;
- len = atoi_n(sp->tok, sp->len);
-
+ }
+ else if (sp->len > 2 && sp->tok[0] == '\'' && sp->tok[sp->len-1] == '\'')
+ {
+ char *b;
+ wrd->term_len = sp->len - 2;
+ b = nmem_malloc(sp->nmem, wrd->term_len);
+ memcpy(b, sp->tok+1, wrd->term_len);
+ wrd->term_buf = b;
sp_lex(sp);
- if (sp->lookahead != ')')
- return 0;
-
+ }
+ else
+ {
+ wrd->term_buf = "";
+ wrd->term_len = 0;
sp_lex(sp);
- if (wrd->string && wrd->length)
- {
- wrd->string += start;
- wrd->length -= start;
- if (wrd->length > len)
- wrd->length = len;
- }
}
return 1;
}
-static int sp_parse(data1_node *n, RecWord *wrd, const char *src)
+static struct source_parser *source_parser_create()
+{
+ struct source_parser *sp = xmalloc(sizeof(*sp));
+
+ sp->nmem = nmem_create();
+ return sp;
+}
+
+static void source_parser_destroy(struct source_parser *sp)
+{
+ if (!sp)
+ return;
+ nmem_destroy(sp->nmem);
+ xfree(sp);
+}
+
+static int sp_parse(struct source_parser *sp,
+ data1_node *n, RecWord *wrd, const char *src)
{
- struct source_parser sp;
- sp.len = 0;
- sp.tok = 0;
- sp.src = src;
- sp.lookahead = 0;
- sp_lex(&sp);
-
- return sp_expr(&sp, n, wrd);
+ sp->len = 0;
+ sp->tok = 0;
+ sp->src = src;
+ sp->lookahead = 0;
+ nmem_reset(sp->nmem);
+
+ sp_lex(sp);
+ return sp_expr(sp, n, wrd);
}
int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p)
if (p->which == XPATH_PREDICATE_RELATION) {
if (p->u.relation.name[0]) {
if (*p->u.relation.name != '@') {
- yaz_log(LOG_WARN,
+ yaz_log(YLOG_WARN,
" Only attributes (@) are supported in xelm xpath predicates");
- yaz_log(LOG_WARN, "predicate %s ignored", p->u.relation.name);
- return (1);
+ yaz_log(YLOG_WARN, "predicate %s ignored", p->u.relation.name);
+ return 1;
}
attname = p->u.relation.name + 1;
res = 0;
/* looking for the attribute with a specified name */
for (attr = n->u.tag.attributes; attr; attr = attr->next) {
- yaz_log(LOG_DEBUG," - attribute %s <-> %s", attname, attr->name );
+ yaz_log(YLOG_DEBUG," - attribute %s <-> %s", attname, attr->name );
if (!strcmp(attr->name, attname)) {
if (p->u.relation.op[0]) {
if (*p->u.relation.op != '=') {
- yaz_log(LOG_WARN,
+ yaz_log(YLOG_WARN,
"Only '=' relation is supported (%s)",p->u.relation.op);
- yaz_log(LOG_WARN, "predicate %s ignored", p->u.relation.name);
+ yaz_log(YLOG_WARN, "predicate %s ignored", p->u.relation.name);
res = 1; break;
} else {
- yaz_log(LOG_DEBUG," - value %s <-> %s",
+ yaz_log(YLOG_DEBUG," - value %s <-> %s",
p->u.relation.value, attr->value );
if (!strcmp(attr->value, p->u.relation.value)) {
res = 1; break;
}
}
}
- yaz_log(LOG_DEBUG, "return %d", res);
+ yaz_log(YLOG_DEBUG, "return %d", res);
return res;
} else {
return 1;
return (d1_check_xpath_predicate(n, p->u.boolean.left)
|| d1_check_xpath_predicate(n, p->u.boolean.right));
} else {
- yaz_log(LOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op);
+ yaz_log(YLOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op);
return 1;
}
}
}
+static int dfa_match_first(struct DFA_state **dfaar, const char *text)
+{
+ struct DFA_state *s = dfaar[0]; /* start state */
+ struct DFA_tran *t;
+ int i;
+ const char *p = text;
+ unsigned char c;
+
+ for (c = *p++, t = s->trans, i = s->tran_no; --i >= 0; t++)
+ {
+ if (c >= t->ch[0] && c <= t->ch[1])
+ {
+ while (i >= 0)
+ {
+ /* move to next state and return if we get a match */
+ s = dfaar[t->to];
+ if (s->rule_no)
+ return 1;
+ /* next char */
+ if (!c)
+ return 0;
+ c = *p++;
+ for (t = s->trans, i = s->tran_no; --i >= 0; t++)
+ if (c >= t->ch[0] && c <= t->ch[1])
+ break;
+ }
+ }
+ }
+ return 0;
+}
+
/* *ostrich*
New function, looking for xpath "element" definitions in abs, by
data1_node *nn;
#ifdef ENHANCED_XELM
struct xpath_location_step *xp;
-
#endif
- char *pexpr = xmalloc(strlen(tagpath)+2);
+ char *pexpr = xmalloc(strlen(tagpath)+5);
int ok = 0;
- sprintf (pexpr, "%s\n", tagpath);
- yaz_log(LOG_DEBUG,"Checking tagpath %s",tagpath);
- while (xpe)
+ sprintf (pexpr, "/%s\n", tagpath);
+ for (; xpe; xpe = xpe->next)
{
- struct DFA_state **dfaar = xpe->dfa->states;
- struct DFA_state *s=dfaar[0];
- struct DFA_tran *t;
- const char *p;
- int i;
- unsigned char c;
- int start_line = 1;
-
- c = *pexpr++; t = s->trans; i = s->tran_no;
- if ((c >= t->ch[0] && c <= t->ch[1]) || (!t->ch[0])) {
- p = pexpr;
- do {
- if ((s = dfaar[t->to])->rule_no &&
- (start_line || s->rule_nno)) {
- ok = 1;
- break;
- }
- for (t=s->trans, i=s->tran_no; --i >= 0; t++) {
- if ((unsigned) *p >= t->ch[0] && (unsigned) *p <= t->ch[1])
- break;
- }
- p++;
- } while (i >= 0);
- }
- if (ok)
- yaz_log(LOG_DEBUG," xpath match %s",xpe->xpath_expr);
- else
- yaz_log(LOG_DEBUG," xpath no match %s",xpe->xpath_expr);
+ int i;
+ ok = dfa_match_first(xpe->dfa->states, pexpr);
- pexpr--;
if (ok) {
#ifdef ENHANCED_XELM
/* we have to check the perdicates up to the root node */
xp = xpe->xpath;
/* find the first tag up in the node structure */
- nn = n; while (nn && nn->which != DATA1N_tag) {
- nn = nn->parent;
- }
+ for (nn = n; nn && nn->which != DATA1N_tag; nn = nn->parent)
+ ;
/* go from inside out in the node structure, while going
backwards trough xpath location steps ... */
- for (i=xpe->xpath_len - 1; i>0; i--) {
-
- yaz_log(LOG_DEBUG,"Checking step %d: %s on tag %s",
- i,xp[i].part,nn->u.tag.tag);
+ for (i = xpe->xpath_len - 1; i>0; i--)
+ {
+ yaz_log(YLOG_DEBUG, "Checking step %d: %s on tag %s",
+ i, xp[i].part, nn->u.tag.tag);
- if (!d1_check_xpath_predicate(nn, xp[i].predicate)) {
- yaz_log(LOG_DEBUG," Predicates didn't match");
+ if (!d1_check_xpath_predicate(nn, xp[i].predicate))
+ {
+ yaz_log(YLOG_DEBUG, " Predicates didn't match");
ok = 0;
break;
}
- if (nn->which == DATA1N_tag) {
+ if (nn->which == DATA1N_tag)
nn = nn->parent;
- }
}
#endif
- if (ok) {
+ if (ok)
break;
- }
}
- xpe = xpe->next;
}
xfree(pexpr);
if (ok) {
- yaz_log(LOG_DEBUG,"Got it");
+ yaz_log(YLOG_DEBUG, "Got it");
return xpe->termlists;
} else {
return NULL;
char *structure, struct recExtractCtrl *p,
RecWord *wrd)
{
- wrd->attrSet = VAL_IDXPATH;
- wrd->attrUse = 1;
- wrd->reg_type = '0';
- wrd->string = tag_path;
- wrd->length = strlen(tag_path);
+ wrd->index_name = ZEBRA_XPATH_ELM_BEGIN;
+ wrd->index_type = '0';
+ wrd->term_buf = tag_path;
+ wrd->term_len = strlen(tag_path);
(*p->tokenAdd)(wrd);
if (value) {
- wrd->attrUse = 1015;
- wrd->reg_type = 'w';
- wrd->string = value;
- wrd->length = strlen(value);
+ wrd->index_name = ZEBRA_XPATH_ATTR_CDATA;
+ wrd->index_type = 'w';
+ wrd->term_buf = value;
+ wrd->term_len = strlen(value);
(*p->tokenAdd)(wrd);
}
-
- wrd->attrUse = 2;
- wrd->reg_type = '0';
- wrd->string = tag_path;
- wrd->length = strlen(tag_path);
+ wrd->index_name = ZEBRA_XPATH_ELM_END;
+ wrd->index_type = '0';
+ wrd->term_buf = tag_path;
+ wrd->term_len = strlen(tag_path);
(*p->tokenAdd)(wrd);
}
-static void index_xpath (data1_node *n, struct recExtractCtrl *p,
- int level, RecWord *wrd, int use)
+static void mk_tag_path_full(char *tag_path_full, size_t max, data1_node *n)
{
- int i;
- char tag_path_full[1024];
size_t flen = 0;
data1_node *nn;
+
+ /* we have to fetch the whole path to the data tag */
+ for (nn = n; nn; nn = nn->parent)
+ {
+ if (nn->which == DATA1N_tag)
+ {
+ size_t tlen = strlen(nn->u.tag.tag);
+ if (tlen + flen > (max - 2))
+ break;
+ memcpy (tag_path_full + flen, nn->u.tag.tag, tlen);
+ flen += tlen;
+ tag_path_full[flen++] = '/';
+ }
+ else
+ if (nn->which == DATA1N_root)
+ break;
+ }
+ tag_path_full[flen] = 0;
+}
+
+
+static void index_xpath(struct source_parser *sp, data1_node *n,
+ struct recExtractCtrl *p,
+ int level, RecWord *wrd,
+ char *xpath_index,
+ int xpath_is_start
+ )
+{
+ int i;
+ char tag_path_full[1024];
int termlist_only = 1;
data1_termlist *tl;
int xpdone = 0;
-
- yaz_log(LOG_DEBUG, "index_xpath level=%d use=%d", level, use);
+ yaz_log(YLOG_DEBUG, "index_xpath level=%d xpath_index=%s",
+ level, xpath_index);
if ((!n->root->u.root.absyn) ||
(n->root->u.root.absyn->enable_xpath_indexing)) {
- termlist_only = 0;
+ termlist_only = 0;
}
switch (n->which)
{
case DATA1N_data:
- wrd->string = n->u.data.data;
- wrd->length = n->u.data.len;
+ wrd->term_buf = n->u.data.data;
+ wrd->term_len = n->u.data.len;
xpdone = 0;
- flen = 0;
-
- /* we have to fetch the whole path to the data tag */
- for (nn = n; nn; nn = nn->parent) {
- if (nn->which == DATA1N_tag) {
- size_t tlen = strlen(nn->u.tag.tag);
- if (tlen + flen > (sizeof(tag_path_full)-2)) return;
- memcpy (tag_path_full + flen, nn->u.tag.tag, tlen);
- flen += tlen;
- tag_path_full[flen++] = '/';
- }
- else if (nn->which == DATA1N_root) break;
- }
-
- tag_path_full[flen] = 0;
+
+ mk_tag_path_full(tag_path_full, sizeof(tag_path_full), n);
/* If we have a matching termlist... */
if (n->root->u.root.absyn &&
{
/* need to copy recword because it may be changed */
RecWord wrd_tl;
- wrd->reg_type = *tl->structure;
- /* this is the ! case, so structure is for the xpath index */
+ wrd->index_type = *tl->structure;
memcpy (&wrd_tl, wrd, sizeof(*wrd));
if (tl->source)
- sp_parse(n, &wrd_tl, tl->source);
- if (!tl->att) {
- wrd_tl.attrSet = VAL_IDXPATH;
- wrd_tl.attrUse = use;
+ sp_parse(sp, n, &wrd_tl, tl->source);
+ if (!tl->index_name)
+ {
+ /* this is the ! case, so structure is for the xpath index */
+ wrd_tl.index_name = xpath_index;
if (p->flagShowRecords)
{
int i;
printf("%*sXPath index", (level + 1) * 4, "");
printf (" XData:\"");
- for (i = 0; i<wrd_tl.length && i < 40; i++)
- fputc (wrd_tl.string[i], stdout);
+ for (i = 0; i<wrd_tl.term_len && i < 40; i++)
+ fputc (wrd_tl.term_buf[i], stdout);
fputc ('"', stdout);
- if (wrd_tl.length > 40)
+ if (wrd_tl.term_len > 40)
printf (" ...");
fputc ('\n', stdout);
}
xpdone = 1;
} else {
/* this is just the old fashioned attribute based index */
- wrd_tl.attrSet = (int) (tl->att->parent->reference);
- wrd_tl.attrUse = tl->att->locals->local;
+ wrd_tl.index_name = tl->index_name;
if (p->flagShowRecords)
{
int i;
printf("%*sIdx: [%s]", (level + 1) * 4, "",
tl->structure);
- printf("%s:%s [%d] %s",
- tl->att->parent->name,
- tl->att->name, tl->att->value,
- tl->source);
+ printf("%s %s", tl->index_name, tl->source);
printf (" XData:\"");
- for (i = 0; i<wrd_tl.length && i < 40; i++)
- fputc (wrd_tl.string[i], stdout);
+ for (i = 0; i<wrd_tl.term_len && i < 40; i++)
+ fputc (wrd_tl.term_buf[i], stdout);
fputc ('"', stdout);
- if (wrd_tl.length > 40)
+ if (wrd_tl.term_len > 40)
printf (" ...");
fputc ('\n', stdout);
}
or no ! in the termlist, and default indexing is enabled... */
if (!p->flagShowRecords && !xpdone && !termlist_only)
{
- wrd->attrSet = VAL_IDXPATH;
- wrd->attrUse = use;
- wrd->reg_type = 'w';
+ wrd->index_name = xpath_index;
+ wrd->index_type = 'w';
(*p->tokenAdd)(wrd);
}
break;
case DATA1N_tag:
- flen = 0;
- for (nn = n; nn; nn = nn->parent)
- {
- if (nn->which == DATA1N_tag)
- {
- size_t tlen = strlen(nn->u.tag.tag);
- if (tlen + flen > (sizeof(tag_path_full)-2))
- return;
- memcpy (tag_path_full + flen, nn->u.tag.tag, tlen);
- flen += tlen;
- tag_path_full[flen++] = '/';
- }
- else if (nn->which == DATA1N_root)
- break;
- }
-
+ mk_tag_path_full(tag_path_full, sizeof(tag_path_full), n);
- wrd->reg_type = '0';
- wrd->string = tag_path_full;
- wrd->length = flen;
- wrd->attrSet = VAL_IDXPATH;
- wrd->attrUse = use;
+ wrd->index_type = '0';
+ wrd->term_buf = tag_path_full;
+ wrd->term_len = strlen(tag_path_full);
+ wrd->index_name = xpath_index;
if (p->flagShowRecords)
{
printf("%*s tag=", (level + 1) * 4, "");
- for (i = 0; i<wrd->length && i < 40; i++)
- fputc (wrd->string[i], stdout);
+ for (i = 0; i<wrd->term_len && i < 40; i++)
+ fputc (wrd->term_buf[i], stdout);
if (i == 40)
printf (" ..");
printf("\n");
data1_termlist *tl;
int do_xpindex;
- tag_path_full[flen] = 0;
-
- /* Add tag start/end xpath index, only when there is a ! in the apropriate xelm
- directive, or default xpath indexing is enabled */
- if (!(do_xpindex = 1 - termlist_only)) {
- if ((tl = xpath_termlist_by_tagpath(tag_path_full, n))) {
- for (; tl; tl = tl->next) { if (!tl->att) {do_xpindex = 1;} }
+ /* Add tag start/end xpath index, only when there is a ! in
+ the apropriate xelm directive, or default xpath indexing
+ is enabled
+ */
+ if (!(do_xpindex = 1 - termlist_only))
+ {
+ if ((tl = xpath_termlist_by_tagpath(tag_path_full, n)))
+ {
+ for (; tl; tl = tl->next)
+ {
+ if (!tl->index_name)
+ do_xpindex = 1;
+ }
}
}
if (do_xpindex) {
(*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */
}
- if (use == 1) /* only for the starting tag... */
+ if (xpath_is_start == 1) /* only for the starting tag... */
{
#define MAX_ATTR_COUNT 50
data1_termlist *tll[MAX_ATTR_COUNT];
int do_xpindex = 1 - termlist_only;
data1_termlist *tl;
char attr_tag_path_full[1024];
- int int_len = flen;
/* this could be cached as well */
- sprintf (attr_tag_path_full, "@%s/%.*s",
- xp->name, int_len, tag_path_full);
+ sprintf (attr_tag_path_full, "@%s/%s",
+ xp->name, tag_path_full);
tll[i] = xpath_termlist_by_tagpath(attr_tag_path_full,n);
if ((tl = tll[i]))
{
for (; tl; tl = tl->next)
- {
- if (!tl->att)
- do_xpindex = 1;
+ {
+ if (!tl->index_name)
+ do_xpindex = 1;
}
}
if (do_xpindex) {
/* attribute (no value) */
- wrd->reg_type = '0';
- wrd->attrUse = 3;
- wrd->string = xp->name;
- wrd->length = strlen(xp->name);
+ wrd->index_type = '0';
+ wrd->index_name = ZEBRA_XPATH_ATTR_NAME;
+ wrd->term_buf = xp->name;
+ wrd->term_len = strlen(xp->name);
wrd->seqno--;
(*p->tokenAdd)(wrd);
strcpy (comb, xp->name);
strcat (comb, "=");
strcat (comb, xp->value);
-
- wrd->attrUse = 3;
- wrd->reg_type = '0';
- wrd->string = comb;
- wrd->length = strlen(comb);
+
+ wrd->index_name = ZEBRA_XPATH_ATTR_NAME;
+ wrd->index_type = '0';
+ wrd->term_buf = comb;
+ wrd->term_len = strlen(comb);
wrd->seqno--;
(*p->tokenAdd)(wrd);
for (xp = n->u.tag.attributes; xp; xp = xp->next) {
data1_termlist *tl;
char attr_tag_path_full[1024];
- int int_len = flen;
int xpdone = 0;
- sprintf (attr_tag_path_full, "@%s/%.*s",
- xp->name, int_len, tag_path_full);
+ sprintf (attr_tag_path_full, "@%s/%s",
+ xp->name, tag_path_full);
if ((tl = tll[i]))
{
/* If there is a termlist given (=xelm directive) */
for (; tl; tl = tl->next)
{
- if (!tl->att) {
+ if (!tl->index_name)
+ {
/* add xpath index for the attribute */
index_xpath_attr (attr_tag_path_full, xp->name,
xp->value, tl->structure,
p, wrd);
xpdone = 1;
} else {
- /* add attribute based index for the attribute */
- if (xp->value) {
- wrd->attrSet = (int)
- (tl->att->parent->reference);
- wrd->attrUse = tl->att->locals->local;
- wrd->reg_type = *tl->structure;
- wrd->string = xp->value;
- wrd->length = strlen(xp->value);
+ /* index attribute value (only path/@attr) */
+ if (xp->value)
+ {
+ wrd->index_name = tl->index_name;
+ wrd->index_type = *tl->structure;
+ wrd->term_buf = xp->value;
+ wrd->term_len = strlen(xp->value);
(*p->tokenAdd)(wrd);
}
}
}
}
-static void index_termlist (data1_node *par, data1_node *n,
+static void index_termlist (struct source_parser *sp, data1_node *par,
+ data1_node *n,
struct recExtractCtrl *p, int level, RecWord *wrd)
{
data1_termlist *tlist = 0;
return;
if (par->u.tag.element->tag)
dtype = par->u.tag.element->tag->kind;
-
+
for (; tlist; tlist = tlist->next)
{
/* consider source */
- wrd->string = 0;
+ wrd->term_buf = 0;
assert(tlist->source);
- sp_parse(n, wrd, tlist->source);
+ sp_parse(sp, n, wrd, tlist->source);
- if (wrd->string)
+ if (wrd->term_buf && wrd->term_len)
{
if (p->flagShowRecords)
{
int i;
printf("%*sIdx: [%s]", (level + 1) * 4, "",
tlist->structure);
- printf("%s:%s [%d] %s",
- tlist->att->parent->name,
- tlist->att->name, tlist->att->value,
- tlist->source);
+ printf("%s %s", tlist->index_name, tlist->source);
printf (" XData:\"");
- for (i = 0; i<wrd->length && i < 40; i++)
- fputc (wrd->string[i], stdout);
+ for (i = 0; i<wrd->term_len && i < 40; i++)
+ fputc (wrd->term_buf[i], stdout);
fputc ('"', stdout);
- if (wrd->length > 40)
+ if (wrd->term_len > 40)
printf (" ...");
fputc ('\n', stdout);
}
else
{
- wrd->reg_type = *tlist->structure;
- wrd->attrSet = (int) (tlist->att->parent->reference);
- wrd->attrUse = tlist->att->locals->local;
+ wrd->index_type = *tlist->structure;
+ wrd->index_name = tlist->index_name;
(*p->tokenAdd)(wrd);
}
}
}
}
-static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level,
- RecWord *wrd)
+static int dumpkeys_r(struct source_parser *sp,
+ data1_node *n, struct recExtractCtrl *p, int level,
+ RecWord *wrd)
{
for (; n; n = n->next)
{
if (n->which == DATA1N_tag)
{
- index_termlist (n, n, p, level, wrd);
+ index_termlist(sp, n, n, p, level, wrd);
/* index start tag */
if (n->root->u.root.absyn)
- index_xpath (n, p, level, wrd, 1);
+ index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_BEGIN,
+ 1 /* is start */);
}
if (n->child)
- if (dumpkeys(n->child, p, level + 1, wrd) < 0)
+ if (dumpkeys_r(sp, n->child, p, level + 1, wrd) < 0)
return -1;
}
if (par)
- index_termlist (par, n, p, level, wrd);
+ index_termlist(sp, par, n, p, level, wrd);
- index_xpath (n, p, level, wrd, 1016);
+ index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_CDATA,
+ 0 /* is start */);
}
if (n->which == DATA1N_tag)
{
/* index end tag */
- index_xpath (n, p, level, wrd, 2);
+ index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_END,
+ 0 /* is start */);
}
if (p->flagShowRecords && n->which == DATA1N_root)
return 0;
}
+static int dumpkeys(data1_node *n, struct recExtractCtrl *p, RecWord *wrd)
+{
+ struct source_parser *sp = source_parser_create();
+ int r = dumpkeys_r(sp, n, p, 0, wrd);
+ source_parser_destroy(sp);
+ return r;
+}
+
int grs_extract_tree(struct recExtractCtrl *p, data1_node *n)
{
oident oe;
}
(*p->init)(p, &wrd);
- return dumpkeys(n, p, 0, &wrd);
+ return dumpkeys(n, p, &wrd);
}
static int grs_extract_sub(void *clientData, struct recExtractCtrl *p,
#endif
(*p->init)(p, &wrd);
- if (dumpkeys(n, p, 0, &wrd) < 0)
+ if (dumpkeys(n, p, &wrd) < 0)
{
data1_free_tree(p->dh, n);
return RECCTRL_EXTRACT_ERROR_GENERIC;
/*
* Return: -1: Nothing done. 0: Ok. >0: Bib-1 diagnostic.
*/
-static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c)
+static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c,
+ char **addinfo, ODR o)
{
data1_esetname *eset;
Z_Espec1 *espec = 0;
if (!(eset = data1_getesetbyname(dh, n->u.root.absyn,
c->u.simple->u.generic)))
{
- yaz_log(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic);
+ yaz_log(YLOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic);
+ *addinfo = odr_strdup(o, c->u.simple->u.generic);
return 25; /* invalid esetname */
}
- yaz_log(LOG_DEBUG, "Esetname '%s' in simple compspec",
+ yaz_log(YLOG_DEBUG, "Esetname '%s' in simple compspec",
c->u.simple->u.generic);
espec = eset->spec;
break;
data1_getesetbyname(dh, n->u.root.absyn,
p->u.elementSetName)))
{
- yaz_log(LOG_LOG, "Unknown esetname '%s'",
+ yaz_log(YLOG_DEBUG, "Unknown esetname '%s'",
p->u.elementSetName);
+ *addinfo = odr_strdup(o, p->u.elementSetName);
return 25; /* invalid esetname */
}
- yaz_log(LOG_DEBUG, "Esetname '%s' in complex compspec",
+ yaz_log(YLOG_DEBUG, "Esetname '%s' in complex compspec",
p->u.elementSetName);
espec = eset->spec;
break;
case Z_ElementSpec_externalSpec:
if (p->u.externalSpec->which == Z_External_espec1)
{
- yaz_log(LOG_DEBUG, "Got Espec-1");
+ yaz_log(YLOG_DEBUG, "Got Espec-1");
espec = p->u.externalSpec-> u.espec1;
}
else
{
- yaz_log(LOG_LOG, "Unknown external espec.");
+ yaz_log(YLOG_LOG, "Unknown external espec.");
return 25; /* bad. what is proper diagnostic? */
}
break;
}
if (espec)
{
- yaz_log(LOG_DEBUG, "Element: Espec-1 match");
+ yaz_log(YLOG_DEBUG, "Element: Espec-1 match");
return data1_doespec1(dh, n, espec);
}
else
{
- yaz_log(LOG_DEBUG, "Element: all match");
+ yaz_log(YLOG_DEBUG, "Element: all match");
return -1;
}
}
gri.dh = p->dh;
gri.clientData = clientData;
- yaz_log(LOG_DEBUG, "grs_retrieve");
+ yaz_log(YLOG_DEBUG, "grs_retrieve");
node = (*grs_read)(&gri);
if (!node)
{
}
data1_concat_text(p->dh, mem, node);
- /* ensure our data1 tree is UTF-8 */
- data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node));
-
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
top = data1_get_root_tag (p->dh, node);
- yaz_log(LOG_DEBUG, "grs_retrieve: size");
+ yaz_log(YLOG_DEBUG, "grs_retrieve: size");
tagname = data1_systag_lookup(node->u.root.absyn, "size", "size");
if (tagname &&
(dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem)))
if (tagname && p->score >= 0 &&
(dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem)))
{
- yaz_log(LOG_DEBUG, "grs_retrieve: %s", tagname);
+ yaz_log(YLOG_DEBUG, "grs_retrieve: %s", tagname);
dnew->u.data.what = DATA1I_num;
dnew->u.data.data = dnew->lbuf;
sprintf(dnew->u.data.data, "%d", p->score);
if (tagname && p->localno > 0 &&
(dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem)))
{
- yaz_log(LOG_DEBUG, "grs_retrieve: %s", tagname);
+ yaz_log(YLOG_DEBUG, "grs_retrieve: %s", tagname);
dnew->u.data.what = DATA1I_text;
dnew->u.data.data = dnew->lbuf;
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
-#if YAZ_VERSIONL >= 0x010903L
if (p->comp && p->comp->which == Z_RecordComp_complex &&
p->comp->u.complex->generic &&
p->comp->u.complex->generic->which == Z_Schema_oid &&
if (oe)
requested_schema = oe->value;
}
-#else
- if (p->comp && p->comp->which == Z_RecordComp_complex &&
- p->comp->u.complex->generic && p->comp->u.complex->generic->schema)
- {
- oident *oe = oid_getentbyoid (p->comp->u.complex->generic->schema);
- if (oe)
- requested_schema = oe->value;
- }
-#endif
-
/* If schema has been specified, map if possible, then check that
* we got the right one
*/
if (requested_schema != VAL_NONE)
{
- yaz_log(LOG_DEBUG, "grs_retrieve: schema mapping");
+ yaz_log(YLOG_DEBUG, "grs_retrieve: schema mapping");
for (map = node->u.root.absyn->maptabs; map; map = map->next)
{
if (map->target_absyn_ref == requested_schema)
* the overlap of schema and formatting which is inherent in the MARC
* family)
*/
- yaz_log(LOG_DEBUG, "grs_retrieve: syntax mapping");
+ yaz_log(YLOG_DEBUG, "grs_retrieve: syntax mapping");
if (node->u.root.absyn)
for (map = node->u.root.absyn->maptabs; map; map = map->next)
{
break;
}
}
- yaz_log(LOG_DEBUG, "grs_retrieve: schemaIdentifier");
+ yaz_log(YLOG_DEBUG, "grs_retrieve: schemaIdentifier");
if (node->u.root.absyn &&
node->u.root.absyn->reference != VAL_NONE &&
p->input_format == VAL_GRS1)
}
}
- yaz_log(LOG_DEBUG, "grs_retrieve: element spec");
- if (p->comp && (res = process_comp(p->dh, node, p->comp)) > 0)
+ yaz_log(YLOG_DEBUG, "grs_retrieve: element spec");
+ if (p->comp && (res = process_comp(p->dh, node, p->comp, &p->addinfo,
+ p->odr)) > 0)
{
p->diagnostic = res;
if (onode)
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
- yaz_log(LOG_DEBUG, "grs_retrieve: transfer syntax mapping");
+ yaz_log(YLOG_DEBUG, "grs_retrieve: transfer syntax mapping");
switch (p->output_format = (p->input_format != VAL_NONE ?
p->input_format : VAL_SUTRS))
{
case VAL_TEXT_XML:
-
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
-
- if (p->encoding)
- data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
+ /* default output encoding for XML is UTF-8 */
+ data1_iconv (p->dh, mem, node,
+ p->encoding ? p->encoding : "UTF-8",
+ data1_get_encoding(p->dh, node));
if (!(p->rec_buf = data1_nodetoidsgml(p->dh, node, selected,
&p->rec_len)))
}
break;
case VAL_GRS1:
+ data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node));
dummy = 0;
if (!(p->rec_buf = data1_nodetogr(p->dh, node, selected,
p->odr, &dummy)))
p->rec_len = (size_t) (-1);
break;
case VAL_EXPLAIN:
+ /* ensure our data1 tree is UTF-8 */
+ data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node));
+
if (!(p->rec_buf = data1_nodetoexplain(p->dh, node, selected,
p->odr)))
p->diagnostic = 238;
p->rec_len = (size_t) (-1);
break;
case VAL_SUMMARY:
+ /* ensure our data1 tree is UTF-8 */
+ data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node));
if (!(p->rec_buf = data1_nodetosummary(p->dh, node, selected,
p->odr)))
p->diagnostic = 238;
p->rec_len = (size_t) (-1);
break;
case VAL_SUTRS:
- if (p->encoding)
- data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
+ if (p->encoding)
+ data1_iconv (p->dh, mem, node, p->encoding,
+ data1_get_encoding(p->dh, node));
if (!(p->rec_buf = data1_nodetobuf(p->dh, node, selected,
&p->rec_len)))
p->diagnostic = 238;
}
break;
case VAL_SOIF:
+ if (p->encoding)
+ data1_iconv (p->dh, mem, node, p->encoding,
+ data1_get_encoding(p->dh, node));
if (!(p->rec_buf = data1_nodetosoif(p->dh, node, selected,
&p->rec_len)))
p->diagnostic = 238;
p->diagnostic = 238;
break;
}
- if (p->encoding)
- data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
+ if (p->encoding)
+ data1_iconv (p->dh, mem, node, p->encoding,
+ data1_get_encoding(p->dh, node));
if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node,
selected, &p->rec_len)))
p->diagnostic = 238;
return 0;
}
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+
projects / idzebra-moved-to-github.git / blobdiff