-/* $Id: recgrs.c,v 1.82 2003-08-21 10:29:00 adam Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003
+/* $Id: recgrs.c,v 1.89 2004-08-24 14:29:09 adam Exp $
+ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004
Index Data Aps
This file is part of the Zebra server.
#if HAVE_PERL
grs_add_handler (h, recTypeGrs_perl);
#endif
+ grs_add_handler (h, recTypeGrs_danbib);
return h;
}
xfree (h);
}
+struct source_parser {
+ int len;
+ const char *tok;
+ const char *src;
+ int lookahead;
+};
+
+static int sp_lex(struct source_parser *sp)
+{
+ while (*sp->src == ' ')
+ (sp->src)++;
+ sp->tok = sp->src;
+ sp->len = 0;
+ while (*sp->src && !strchr("<>();,-: ", *sp->src))
+ {
+ sp->src++;
+ sp->len++;
+ }
+ if (sp->len)
+ sp->lookahead = 't';
+ else
+ {
+ sp->lookahead = *sp->src;
+ if (*sp->src)
+ sp->src++;
+ }
+ return sp->lookahead;
+}
+
+
+static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd)
+{
+ if (sp->lookahead != 't')
+ return 0;
+ if (sp->len == 4 && !memcmp(sp->tok, "data", sp->len))
+ {
+ if (n->which == DATA1N_data)
+ {
+ wrd->string = n->u.data.data;
+ wrd->length = n->u.data.len;
+ }
+ sp_lex(sp);
+ }
+ else if (sp->len == 3 && !memcmp(sp->tok, "tag", sp->len))
+ {
+ if (n->which == DATA1N_tag)
+ {
+ wrd->string = n->u.tag.tag;
+ wrd->length = strlen(n->u.tag.tag);
+ }
+ sp_lex(sp);
+ }
+ else if (sp->len == 4 && !memcmp(sp->tok, "attr", sp->len))
+ {
+ sp_lex(sp);
+ if (sp->lookahead != '(')
+ return 0;
+ sp_lex(sp);
+ if (sp->lookahead != 't')
+ return 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))
+ p = p->next;
+ if (p)
+ {
+ wrd->string = p->value;
+ wrd->length = strlen(p->value);
+ }
+ }
+ sp_lex(sp);
+ if (sp->lookahead != ')')
+ return 0;
+ sp_lex(sp);
+ }
+ 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;
+
+ sp_lex(sp);
+ if (sp->lookahead != 't')
+ return 0;
+ len = atoi_n(sp->tok, sp->len);
+
+ sp_lex(sp);
+ if (sp->lookahead != ')')
+ return 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)
+{
+ 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);
+}
+
int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p)
{
int res = 1;
if (p->which == XPATH_PREDICATE_RELATION) {
if (p->u.relation.name[0]) {
if (*p->u.relation.name != '@') {
- logf(LOG_WARN,
+ yaz_log(LOG_WARN,
" Only attributes (@) are supported in xelm xpath predicates");
- logf(LOG_WARN, "predicate %s ignored", p->u.relation.name);
+ yaz_log(LOG_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) {
- logf(LOG_DEBUG," - attribute %s <-> %s", attname, attr->name );
+ yaz_log(LOG_DEBUG," - attribute %s <-> %s", attname, attr->name );
if (!strcmp(attr->name, attname)) {
if (p->u.relation.op[0]) {
if (*p->u.relation.op != '=') {
- logf(LOG_WARN,
+ yaz_log(LOG_WARN,
"Only '=' relation is supported (%s)",p->u.relation.op);
- logf(LOG_WARN, "predicate %s ignored", p->u.relation.name);
+ yaz_log(LOG_WARN, "predicate %s ignored", p->u.relation.name);
res = 1; break;
} else {
- logf(LOG_DEBUG," - value %s <-> %s",
+ yaz_log(LOG_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);
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 {
- logf(LOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op);
+ yaz_log(LOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op);
return 1;
}
}
int ok = 0;
sprintf (pexpr, "%s\n", tagpath);
- logf(LOG_DEBUG,"Checking tagpath %s",tagpath);
+ yaz_log(LOG_DEBUG,"Checking tagpath %s",tagpath);
while (xpe)
{
struct DFA_state **dfaar = xpe->dfa->states;
unsigned char c;
int start_line = 1;
- logf(LOG_DEBUG," - xpath %s",xpe->xpath_expr);
-
c = *pexpr++; t = s->trans; i = s->tran_no;
if ((c >= t->ch[0] && c <= t->ch[1]) || (!t->ch[0])) {
p = pexpr;
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);
+
pexpr--;
if (ok) {
#ifdef ENHANCED_XELM
backwards trough xpath location steps ... */
for (i=xpe->xpath_len - 1; i>0; i--) {
- logf(LOG_DEBUG,"Checking step %d: %s on tag %s",
+ yaz_log(LOG_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)) {
- logf(LOG_DEBUG," Predicates didn't match");
+ yaz_log(LOG_DEBUG," Predicates didn't match");
ok = 0;
break;
}
xfree(pexpr);
if (ok) {
- logf(LOG_DEBUG,"Got it");
+ yaz_log(LOG_DEBUG,"Got it");
return xpe->termlists;
} else {
return NULL;
size_t flen = 0;
data1_node *nn;
int termlist_only = 1;
+ data1_termlist *tl;
+ int xpdone = 0;
+ yaz_log(LOG_DEBUG, "index_xpath level=%d use=%d", level, use);
if ((!n->root->u.root.absyn) ||
(n->root->u.root.absyn->enable_xpath_indexing)) {
termlist_only = 0;
case DATA1N_data:
wrd->string = n->u.data.data;
wrd->length = n->u.data.len;
- if (p->flagShowRecords)
- {
- printf("%*s XData:\"", (level + 1) * 4, "");
- for (i = 0; i<wrd->length && i < 8; i++)
- fputc (wrd->string[i], stdout);
- printf("\"\n");
- }
- else {
- data1_termlist *tl;
- int xpdone = 0;
- flen = 0;
-
- // logf (LOG_LOG,"Index value, %s",wrd->string);
-
- /* 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;
+ xpdone = 0;
+ flen = 0;
- /* If we have a matching termlist... */
- if ((tl = xpath_termlist_by_tagpath(tag_path_full, n))) {
- for (; tl; tl = tl->next) {
- wrd->reg_type = *tl->structure;
- /* this is the ! case, so structure is for the xpath index */
- if (!tl->att) {
- wrd->attrSet = VAL_IDXPATH;
- wrd->attrUse = use;
- (*p->tokenAdd)(wrd);
- xpdone = 1;
- /* this is just the old fashioned attribute based index */
- } else {
- wrd->attrSet = (int) (tl->att->parent->reference);
- wrd->attrUse = tl->att->locals->local;
- (*p->tokenAdd)(wrd);
- }
- }
- }
- /* xpath indexing is done, if there was no termlist given,
- or no ! in the termlist, and default indexing is enabled... */
- if ((!xpdone) && (!termlist_only)) {
- wrd->attrSet = VAL_IDXPATH;
- wrd->attrUse = use;
- wrd->reg_type = 'w';
- (*p->tokenAdd)(wrd);
- }
+ /* 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;
+
+ /* If we have a matching termlist... */
+ if (n->root->u.root.absyn &&
+ (tl = xpath_termlist_by_tagpath(tag_path_full, n)))
+ {
+ for (; tl; tl = tl->next)
+ {
+ /* 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 */
+ 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;
+ 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);
+ fputc ('"', stdout);
+ if (wrd_tl.length > 40)
+ printf (" ...");
+ fputc ('\n', stdout);
+ }
+ else
+ (*p->tokenAdd)(&wrd_tl);
+ 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;
+ 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 (" XData:\"");
+ for (i = 0; i<wrd_tl.length && i < 40; i++)
+ fputc (wrd_tl.string[i], stdout);
+ fputc ('"', stdout);
+ if (wrd_tl.length > 40)
+ printf (" ...");
+ fputc ('\n', stdout);
+ }
+ else
+ (*p->tokenAdd)(&wrd_tl);
+ }
+ }
+ }
+ /* xpath indexing is done, if there was no termlist given,
+ 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';
+ (*p->tokenAdd)(wrd);
}
break;
case DATA1N_tag:
if (use == 1) /* only for the starting tag... */
{
-
#define MAX_ATTR_COUNT 50
-
data1_termlist *tll[MAX_ATTR_COUNT];
int i = 0;
for (; tlist; tlist = tlist->next)
{
-
- char xattr[512];
/* consider source */
wrd->string = 0;
+ assert(tlist->source);
+ sp_parse(n, wrd, tlist->source);
- if (!strcmp (tlist->source, "data") && n->which == DATA1N_data)
- {
- wrd->string = n->u.data.data;
- wrd->length = n->u.data.len;
- }
- else if (!strcmp (tlist->source, "tag") && n->which == DATA1N_tag)
- {
- wrd->string = n->u.tag.tag;
- wrd->length = strlen(n->u.tag.tag);
- }
- else if (sscanf (tlist->source, "attr(%511[^)])", xattr) == 1 &&
- n->which == DATA1N_tag)
- {
- data1_xattr *p = n->u.tag.attributes;
- while (p && strcmp (p->name, xattr))
- p = p->next;
- if (p)
- {
- wrd->string = p->value;
- wrd->length = strlen(p->value);
- }
- }
if (wrd->string)
{
if (p->flagShowRecords)
if (!(eset = data1_getesetbyname(dh, n->u.root.absyn,
c->u.simple->u.generic)))
{
- logf(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic);
+ yaz_log(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic);
return 25; /* invalid esetname */
}
- logf(LOG_DEBUG, "Esetname '%s' in simple compspec",
+ yaz_log(LOG_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)))
{
- logf(LOG_LOG, "Unknown esetname '%s'",
+ yaz_log(LOG_LOG, "Unknown esetname '%s'",
p->u.elementSetName);
return 25; /* invalid esetname */
}
- logf(LOG_DEBUG, "Esetname '%s' in complex compspec",
+ yaz_log(LOG_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)
{
- logf(LOG_DEBUG, "Got Espec-1");
+ yaz_log(LOG_DEBUG, "Got Espec-1");
espec = p->u.externalSpec-> u.espec1;
}
else
{
- logf(LOG_LOG, "Unknown external espec.");
+ yaz_log(LOG_LOG, "Unknown external espec.");
return 25; /* bad. what is proper diagnostic? */
}
break;
}
if (espec)
{
- logf (LOG_DEBUG, "Element: Espec-1 match");
+ yaz_log(LOG_DEBUG, "Element: Espec-1 match");
return data1_doespec1(dh, n, espec);
}
else
{
- logf (LOG_DEBUG, "Element: all match");
+ yaz_log(LOG_DEBUG, "Element: all match");
return -1;
}
}
data1_mk_tag_data_int (p->dh, n, "score", p->score, mem);
}
data1_mk_text (p->dh, mem, i4, n);
- data1_mk_tag_data_int (p->dh, n, "localnumber", p->localno, mem);
+ data1_mk_tag_data_zint (p->dh, n, "localnumber", p->localno, mem);
if (p->fname)
{
data1_mk_text (p->dh, mem, i4, n);
gri.mem = mem;
gri.dh = p->dh;
- logf (LOG_DEBUG, "grs_retrieve");
+ yaz_log(LOG_DEBUG, "grs_retrieve");
if (read_grs_type (h, &gri, p->subType, &node))
{
p->diagnostic = 14;
#endif
top = data1_get_root_tag (p->dh, node);
- logf (LOG_DEBUG, "grs_retrieve: size");
+ yaz_log(LOG_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)))
{
- logf (LOG_DEBUG, "grs_retrieve: %s", tagname);
+ yaz_log(LOG_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)))
{
- logf (LOG_DEBUG, "grs_retrieve: %s", tagname);
+ yaz_log(LOG_DEBUG, "grs_retrieve: %s", tagname);
dnew->u.data.what = DATA1I_text;
dnew->u.data.data = dnew->lbuf;
- sprintf(dnew->u.data.data, "%d", p->localno);
+ sprintf(dnew->u.data.data, ZINT_FORMAT, p->localno);
dnew->u.data.len = strlen(dnew->u.data.data);
}
#if 0
*/
if (requested_schema != VAL_NONE)
{
- logf (LOG_DEBUG, "grs_retrieve: schema mapping");
+ yaz_log(LOG_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(LOG_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(LOG_DEBUG, "grs_retrieve: schemaIdentifier");
if (node->u.root.absyn &&
node->u.root.absyn->reference != VAL_NONE &&
p->input_format == VAL_GRS1)
}
}
- logf (LOG_DEBUG, "grs_retrieve: element spec");
+ yaz_log(LOG_DEBUG, "grs_retrieve: element spec");
if (p->comp && (res = process_comp(p->dh, node, p->comp)) > 0)
{
p->diagnostic = res;
#if 0
data1_pr_tree (p->dh, node, stdout);
#endif
- logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping");
+ yaz_log(LOG_DEBUG, "grs_retrieve: transfer syntax mapping");
switch (p->output_format = (p->input_format != VAL_NONE ?
p->input_format : VAL_SUTRS))
{