-/* $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
+/* $Id: recgrs.c,v 1.98 2005-01-15 19:38:32 adam Exp $
+ Copyright (C) 1995-2005
+ Index Data ApS
This file is part of the Zebra server.
#include <yaz/log.h>
#include <yaz/oid.h>
-#include <recctrl.h>
-#include "grsread.h"
+#include <d1_absyn.h>
+#include <idzebra/recgrs.h>
#define GRS_MAX_WORD 512
-struct grs_handler {
- RecTypeGrs type;
- void *clientData;
- int initFlag;
- struct grs_handler *next;
-};
-
-struct grs_handlers {
- struct grs_handler *handlers;
-};
-
-static int read_grs_type (struct grs_handlers *h,
- struct grs_read_info *p, const char *type,
- data1_node **root)
-{
- struct grs_handler *gh = h->handlers;
- const char *cp = strchr (type, '.');
-
- if (cp == NULL || cp == type)
- {
- cp = strlen(type) + type;
- *p->type = 0;
- }
- else
- strcpy (p->type, cp+1);
- for (gh = h->handlers; gh; gh = gh->next)
- {
- if (!memcmp (type, gh->type->type, cp-type) &&
- gh->type->type[cp-type] == '\0')
- {
- if (!gh->initFlag)
- {
- gh->initFlag = 1;
- gh->clientData = (*gh->type->init)();
- }
- p->clientData = gh->clientData;
- *root = (gh->type->read)(p);
- gh->clientData = p->clientData;
- return 0;
- }
- }
- return 1;
-}
-
-static void grs_add_handler (struct grs_handlers *h, RecTypeGrs t)
-{
- struct grs_handler *gh = (struct grs_handler *) xmalloc (sizeof(*gh));
- gh->next = h->handlers;
- h->handlers = gh;
- gh->initFlag = 0;
- gh->clientData = 0;
- gh->type = t;
-}
-
-static void *grs_init(RecType recType)
-{
- struct grs_handlers *h = (struct grs_handlers *) xmalloc (sizeof(*h));
- h->handlers = 0;
-
- grs_add_handler (h, recTypeGrs_sgml);
- grs_add_handler (h, recTypeGrs_regx);
-#if HAVE_TCL_H
- grs_add_handler (h, recTypeGrs_tcl);
-#endif
- grs_add_handler (h, recTypeGrs_marc);
- grs_add_handler (h, recTypeGrs_marcxml);
-#if HAVE_EXPAT_H
- grs_add_handler (h, recTypeGrs_xml);
-#endif
-#if HAVE_PERL
- grs_add_handler (h, recTypeGrs_perl);
-#endif
- grs_add_handler (h, recTypeGrs_danbib);
- return h;
-}
-
-static void grs_destroy(void *clientData)
-{
- struct grs_handlers *h = (struct grs_handlers *) clientData;
- struct grs_handler *gh = h->handlers, *gh_next;
- while (gh)
- {
- gh_next = gh->next;
- if (gh->initFlag)
- (*gh->type->destroy)(gh->clientData);
- xfree (gh);
- gh = gh_next;
- }
- xfree (h);
-}
-
struct source_parser {
int len;
const char *tok;
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);
+ 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;
}
}
int ok = 0;
sprintf (pexpr, "%s\n", tagpath);
- yaz_log(LOG_DEBUG,"Checking tagpath %s",tagpath);
+ yaz_log(YLOG_DEBUG,"Checking tagpath %s",tagpath);
while (xpe)
{
struct DFA_state **dfaar = xpe->dfa->states;
} while (i >= 0);
}
if (ok)
- yaz_log(LOG_DEBUG," xpath match %s",xpe->xpath_expr);
+ yaz_log(YLOG_DEBUG," xpath match %s",xpe->xpath_expr);
else
- yaz_log(LOG_DEBUG," xpath no match %s",xpe->xpath_expr);
+ yaz_log(YLOG_DEBUG," xpath no match %s",xpe->xpath_expr);
pexpr--;
if (ok) {
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",
+ 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");
+ yaz_log(YLOG_DEBUG," Predicates didn't match");
ok = 0;
break;
}
xfree(pexpr);
if (ok) {
- yaz_log(LOG_DEBUG,"Got it");
+ yaz_log(YLOG_DEBUG,"Got it");
return xpe->termlists;
} else {
return NULL;
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 use=%d", level, use);
if ((!n->root->u.root.absyn) ||
(n->root->u.root.absyn->enable_xpath_indexing)) {
termlist_only = 0;
return dumpkeys(n, p, 0, &wrd);
}
-static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p,
- NMEM mem)
+static int grs_extract_sub(void *clientData, struct recExtractCtrl *p,
+ NMEM mem,
+ data1_node *(*grs_read)(struct grs_read_info *))
{
data1_node *n;
struct grs_read_info gri;
gri.offset = p->offset;
gri.mem = mem;
gri.dh = p->dh;
+ gri.clientData = clientData;
- if (read_grs_type (h, &gri, p->subType, &n))
- return RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER;
+ n = (*grs_read)(&gri);
if (!n)
return RECCTRL_EXTRACT_EOF;
oe.proto = PROTO_Z3950;
return RECCTRL_EXTRACT_OK;
}
-static int grs_extract(void *clientData, struct recExtractCtrl *p)
+int zebra_grs_extract(void *clientData, struct recExtractCtrl *p,
+ data1_node *(*grs_read)(struct grs_read_info *))
{
int ret;
NMEM mem = nmem_create ();
- struct grs_handlers *h = (struct grs_handlers *) clientData;
-
- ret = grs_extract_sub(h, p, mem);
+ ret = grs_extract_sub(clientData, p, mem, grs_read);
nmem_destroy(mem);
return ret;
}
/*
* 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;
}
}
data1_mk_text (p->dh, mem, i2, n);
}
-static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p)
+int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p,
+ data1_node *(*grs_read)(struct grs_read_info *))
{
data1_node *node = 0, *onode = 0, *top;
data1_node *dnew;
NMEM mem;
struct grs_read_info gri;
const char *tagname;
- struct grs_handlers *h = (struct grs_handlers *) clientData;
+
int requested_schema = VAL_NONE;
data1_marctab *marctab;
int dummy;
gri.offset = 0;
gri.mem = mem;
gri.dh = p->dh;
+ gri.clientData = clientData;
- yaz_log(LOG_DEBUG, "grs_retrieve");
- if (read_grs_type (h, &gri, p->subType, &node))
- {
- p->diagnostic = 14;
- nmem_destroy (mem);
- return 0;
- }
+ yaz_log(YLOG_DEBUG, "grs_retrieve");
+ node = (*grs_read)(&gri);
if (!node)
{
p->diagnostic = 14;
}
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;
sprintf(dnew->u.data.data, ZINT_FORMAT, p->localno);
dnew->u.data.len = strlen(dnew->u.data.data);
}
+
+ if (p->input_format == VAL_TEXT_XML)
+ zebra_xml_metadata (p, top, mem);
+
#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:
- zebra_xml_metadata (p, top, mem);
-
#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;
}
-static struct recType grs_type =
-{
- "grs",
- grs_init,
- grs_destroy,
- grs_extract,
- grs_retrieve
-};
-
-RecType recTypeGrs = &grs_type;