-/* $Id: recgrs.c,v 1.2 2006-07-06 11:40:25 adam Exp $
+/* $Id: recgrs.c,v 1.12 2006-12-22 13:57:27 adam Exp $
Copyright (C) 1995-2006
Index Data ApS
for more details.
You should have received a copy of the GNU General Public License
-along with Zebra; see the file LICENSE.zebra. If not, write to the
-Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
-02111-1307, USA.
+along with this program; if not, write to the Free Software
+Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+
*/
#include <stdio.h>
return 1;
}
-static struct source_parser *source_parser_create()
+static struct source_parser *source_parser_create(void)
{
struct source_parser *sp = xmalloc(sizeof(*sp));
res = 0;
/* looking for the attribute with a specified name */
for (attr = n->u.tag.attributes; attr; attr = attr->next) {
- 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(YLOG_WARN, "predicate %s ignored", p->u.relation.name);
res = 1; break;
} else {
- 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(YLOG_DEBUG, "return %d", res);
return res;
} else {
return 1;
data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n)
{
data1_absyn *abs = n->root->u.root.absyn;
- data1_xpelement *xpe = abs->xp_elements;
+
+ data1_xpelement *xpe = 0;
data1_node *nn;
#ifdef ENHANCED_XELM
struct xpath_location_step *xp;
#endif
char *pexpr = xmalloc(strlen(tagpath)+5);
- int ok = 0;
sprintf (pexpr, "/%s\n", tagpath);
- for (; xpe; xpe = xpe->next)
+
+ for (xpe = abs->xp_elements; xpe; xpe = xpe->next)
+ xpe->match_state = -1; /* don't know if it matches yet */
+
+ for (xpe = abs->xp_elements; xpe; xpe = xpe->next)
{
int i;
- ok = dfa_match_first(xpe->dfa->states, pexpr);
-
+ int ok = xpe->match_state;
+ if (ok == -1)
+ { /* don't know whether there is a match yet */
+ data1_xpelement *xpe1;
+
+ assert(xpe->dfa);
+ ok = dfa_match_first(xpe->dfa->states, pexpr);
+
+#if OPTIMIZE_MELM
+ /* mark this and following ones with same regexp */
+ for (xpe1 = xpe; xpe1; xpe1 = xpe1->match_next)
+ xpe1->match_state = ok;
+#endif
+ }
+ assert (ok == 0 || ok == 1);
if (ok) {
#ifdef ENHANCED_XELM
/* we have to check the perdicates up to the root node */
backwards trough xpath location steps ... */
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(YLOG_DEBUG, " Predicates didn't match");
ok = 0;
break;
}
xfree(pexpr);
- if (ok) {
+ if (xpe) {
yaz_log(YLOG_DEBUG, "Got it");
return xpe->termlists;
} else {
}
+static void index_staticrank(struct recExtractCtrl *p,
+ RecWord *wrd,
+ data1_absyn *absyn)
+{
+ const char *staticrank_index = data1_absyn_get_staticrank(absyn);
+
+ if (staticrank_index && !strcmp(wrd->index_name, staticrank_index))
+ {
+ char valz[20];
+ size_t len = wrd->term_len;
+
+ if (len > sizeof(valz)-1)
+ len = sizeof(valz)-1;
+ memcpy(valz, wrd->term_buf, len);
+ valz[len] = '\0';
+ p->staticrank = atozint(valz);
+ }
+}
+
static void index_xpath(struct source_parser *sp, data1_node *n,
struct recExtractCtrl *p,
int level, RecWord *wrd,
int termlist_only = 1;
data1_termlist *tl;
int xpdone = 0;
- if ((!n->root->u.root.absyn) ||
- (n->root->u.root.absyn->xpath_indexing == DATA1_XPATH_INDEXING_ENABLE)) {
+
+ if (!n->root->u.root.absyn
+ ||
+ n->root->u.root.absyn->xpath_indexing == DATA1_XPATH_INDEXING_ENABLE)
+ {
termlist_only = 0;
}
+
switch (n->which)
{
case DATA1N_data:
if (n->root->u.root.absyn &&
(tl = xpath_termlist_by_tagpath(tag_path_full, n)))
{
+ zint max_seqno = 0;
for (; tl; tl = tl->next)
{
/* need to copy recword because it may be changed */
memcpy (&wrd_tl, wrd, sizeof(*wrd));
if (tl->source)
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.term_len && i < 40; i++)
- fputc (wrd_tl.term_buf[i], stdout);
- fputc ('"', stdout);
- if (wrd_tl.term_len > 40)
- printf (" ...");
- fputc ('\n', stdout);
- }
- else
- (*p->tokenAdd)(&wrd_tl);
- xpdone = 1;
- } else {
- /* this is just the old fashioned attribute based index */
- wrd_tl.index_name = tl->index_name;
- if (p->flagShowRecords)
- {
- int i;
- printf("%*sIdx: [%s]", (level + 1) * 4, "",
- tl->structure);
- printf("%s %s", tl->index_name, tl->source);
- printf (" XData:\"");
- for (i = 0; i<wrd_tl.term_len && i < 40; i++)
- fputc (wrd_tl.term_buf[i], stdout);
- fputc ('"', stdout);
- if (wrd_tl.term_len > 40)
- printf (" ...");
- fputc ('\n', stdout);
- }
- else
- (*p->tokenAdd)(&wrd_tl);
- }
+
+ /* this is just the old fashioned attribute based index */
+ wrd_tl.index_name = tl->index_name;
+ if (p->flagShowRecords)
+ {
+ int i;
+ printf("%*sIdx: [%s]", (level + 1) * 4, "",
+ tl->structure);
+ printf("%s %s", tl->index_name, tl->source);
+ printf (" XData:\"");
+ for (i = 0; i<wrd_tl.term_len && i < 40; i++)
+ fputc (wrd_tl.term_buf[i], stdout);
+ fputc ('"', stdout);
+ if (wrd_tl.term_len > 40)
+ printf (" ...");
+ fputc ('\n', stdout);
+ }
+ else
+ {
+ (*p->tokenAdd)(&wrd_tl);
+ index_staticrank(p, &wrd_tl, n->root->u.root.absyn);
+ }
+ if (wrd_tl.seqno > max_seqno)
+ max_seqno = wrd_tl.seqno;
}
+ if (max_seqno)
+ wrd->seqno = max_seqno;
+
}
/* xpath indexing is done, if there was no termlist given,
or no ! in the termlist, and default indexing is enabled... */
}
break;
case DATA1N_tag:
+ if (termlist_only)
+ return;
mk_tag_path_full(tag_path_full, sizeof(tag_path_full), n);
wrd->index_type = '0';
else
{
data1_xattr *xp;
- data1_termlist *tl;
- int do_xpindex;
-
- /* 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) */
- }
+
+ (*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */
if (xpath_is_start == 1) /* only for the starting tag... */
{
data1_termlist *tll[MAX_ATTR_COUNT];
int i = 0;
-
- /* get termlists for attributes, and find out, if we have to do xpath indexing */
- for (xp = n->u.tag.attributes; xp; xp = xp->next) {
- i++;
- }
-
- i = 0;
for (xp = n->u.tag.attributes; xp; xp = xp->next) {
char comb[512];
- int do_xpindex = 1 - termlist_only;
- data1_termlist *tl;
char attr_tag_path_full[1024];
/* this could be cached as well */
tll[i] = xpath_termlist_by_tagpath(attr_tag_path_full,n);
- /* if there is a ! in the xelm termlist, or default indexing is on,
- proceed with xpath idx */
- if ((tl = tll[i]))
- {
- for (; tl; tl = tl->next)
- {
- if (!tl->index_name)
- do_xpindex = 1;
- }
- }
+ /* attribute (no value) */
+ wrd->index_type = '0';
+ wrd->index_name = ZEBRA_XPATH_ATTR_NAME;
+ wrd->term_buf = xp->name;
+ wrd->term_len = strlen(xp->name);
- if (do_xpindex) {
+ wrd->seqno--;
+ (*p->tokenAdd)(wrd);
+
+ if (xp->value
+ &&
+ strlen(xp->name) + strlen(xp->value) < sizeof(comb)-2)
+ {
+ /* attribute value exact */
+ strcpy (comb, xp->name);
+ strcat (comb, "=");
+ strcat (comb, xp->value);
- /* attribute (no value) */
+ wrd->index_name = ZEBRA_XPATH_ATTR_NAME;
wrd->index_type = '0';
- wrd->index_name = ZEBRA_XPATH_ATTR_NAME;
- wrd->term_buf = xp->name;
- wrd->term_len = strlen(xp->name);
-
+ wrd->term_buf = comb;
+ wrd->term_len = strlen(comb);
wrd->seqno--;
- (*p->tokenAdd)(wrd);
- if (xp->value &&
- strlen(xp->name) + strlen(xp->value) < sizeof(comb)-2) {
-
- /* attribute value exact */
- strcpy (comb, xp->name);
- strcat (comb, "=");
- strcat (comb, xp->value);
-
- 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);
- }
+ (*p->tokenAdd)(wrd);
}
i++;
}
wrd->term_buf = xp->value;
wrd->term_len = strlen(xp->value);
(*p->tokenAdd)(wrd);
+ index_staticrank(p, wrd,
+ n->root->u.root.absyn);
}
}
}
{
wrd->index_type = *tlist->structure;
wrd->index_name = tlist->index_name;
+ index_staticrank(p, wrd, n->root->u.root.absyn);
(*p->tokenAdd)(wrd);
}
}
}
(*p->init)(p, &wrd);
+ /* data1_pr_tree(p->dh, n, stdout); */
+
return dumpkeys(n, p, &wrd);
}
int oidtmp[OID_SIZE];
RecWord wrd;
- gri.readf = p->readf;
- gri.seekf = p->seekf;
- gri.tellf = p->tellf;
- gri.endf = p->endf;
- gri.fh = p->fh;
- gri.offset = p->offset;
+ gri.stream = p->stream;
gri.mem = mem;
gri.dh = p->dh;
gri.clientData = clientData;
/* ensure our data1 tree is UTF-8 */
data1_iconv (p->dh, mem, n, "UTF-8", data1_get_encoding(p->dh, n));
+
+ data1_remove_idzebra_subtree (p->dh, n);
+
#if 0
data1_pr_tree (p->dh, n, stdout);
#endif
int dummy;
mem = nmem_create();
- gri.readf = p->readf;
- gri.seekf = p->seekf;
- gri.tellf = p->tellf;
- gri.endf = NULL;
- gri.fh = p->fh;
- gri.offset = 0;
+ gri.stream = p->stream;
gri.mem = mem;
gri.dh = p->dh;
gri.clientData = clientData;
}
data1_concat_text(p->dh, mem, node);
+ data1_remove_idzebra_subtree (p->dh, node);
+
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
projects / idzebra-moved-to-github.git / blobdiff