-/* $Id: mod_dom.c,v 1.13 2007-02-15 14:33:41 marc Exp $
+/* $Id: mod_dom.c,v 1.15 2007-02-15 15:08:41 marc Exp $
Copyright (C) 1995-2007
Index Data ApS
#include <idzebra/util.h>
#include <idzebra/recctrl.h>
-
-
-/* Alvis style indexing */
-#define ZEBRA_SCHEMA_XSLT_NS "http://indexdata.dk/zebra/xslt/1"
-static const char *zebra_xslt_ns = ZEBRA_SCHEMA_XSLT_NS;
-
/* DOM filter style indexing */
#define ZEBRA_DOM_NS "http://indexdata.com/zebra-2.0"
static const char *zebra_dom_ns = ZEBRA_DOM_NS;
}
-
-/* Alvis style indexing */
-static void index_cdata(struct filter_info *tinfo, struct recExtractCtrl *ctrl,
- xmlNodePtr ptr, RecWord *recWord)
-{
- for(; ptr; ptr = ptr->next)
- {
- index_cdata(tinfo, ctrl, ptr->children, recWord);
- if (ptr->type != XML_TEXT_NODE)
- continue;
- recWord->term_buf = (const char *)ptr->content;
- recWord->term_len = XML_STRLEN(ptr->content);
- (*ctrl->tokenAdd)(recWord);
- }
-}
-
-/* Alvis style indexing */
-static void index_node(struct filter_info *tinfo, struct recExtractCtrl *ctrl,
- xmlNodePtr ptr, RecWord *recWord)
-{
- for(; ptr; ptr = ptr->next)
- {
- index_node(tinfo, ctrl, ptr->children, recWord);
- if (ptr->type != XML_ELEMENT_NODE || !ptr->ns ||
- XML_STRCMP(ptr->ns->href, zebra_xslt_ns))
- continue;
- if (!XML_STRCMP(ptr->name, "index"))
- {
- const char *name_str = 0;
- const char *type_str = 0;
- const char *xpath_str = 0;
- struct _xmlAttr *attr;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "name", &name_str))
- ;
- else if (attr_content(attr, "xpath", &xpath_str))
- ;
- else if (attr_content(attr, "type", &type_str))
- ;
- else
- yaz_log(YLOG_WARN, "%s: dom filter: "
- "bad attribute %s for <index>",
- tinfo->fname, attr->name);
- }
- if (name_str)
- {
- /* save default type */
- int prev_type = recWord->index_type;
-
- /* type was given */
- if (type_str && *type_str)
- recWord->index_type = *type_str;
-
- recWord->index_name = name_str;
- index_cdata(tinfo, ctrl, ptr->children, recWord);
-
- /* restore it again */
- recWord->index_type = prev_type;
- }
- }
- }
-}
-
-/* Alvis style indexing */
-static void index_record(struct filter_info *tinfo,struct recExtractCtrl *ctrl,
- xmlNodePtr ptr, RecWord *recWord)
-{
- const char *type_str = "update";
-
- if (ptr && ptr->type == XML_ELEMENT_NODE && ptr->ns &&
- !XML_STRCMP(ptr->ns->href, zebra_xslt_ns)
- && !XML_STRCMP(ptr->name, "record"))
- {
- const char *id_str = 0;
- const char *rank_str = 0;
- struct _xmlAttr *attr;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "type", &type_str))
- ;
- else if (attr_content(attr, "id", &id_str))
- ;
- else if (attr_content(attr, "rank", &rank_str))
- ;
- else
- yaz_log(YLOG_WARN, "%s: dom filter: "
- "bad attribute %s for <record>",
- tinfo->fname, attr->name);
- }
- if (id_str)
- sscanf(id_str, "%255s", ctrl->match_criteria);
-
- if (rank_str)
- ctrl->staticrank = atozint(rank_str);
- ptr = ptr->children;
- }
-
- if (!strcmp("update", type_str))
- index_node(tinfo, ctrl, ptr, recWord);
- else if (!strcmp("delete", type_str))
- yaz_log(YLOG_WARN, "%s dom filter: "
- "delete: to be implemented");
- else
- yaz_log(YLOG_WARN, "dom filter: "
- "unknown record type '%s'",
- type_str);
-}
-
-
-/* Alvis style indexing */
-static void extract_doc_alvis(struct filter_info *tinfo,
- struct recExtractCtrl *extctr,
- xmlDocPtr doc)
-{
- if (doc){
- RecWord recWord;
- xmlChar *buf_out;
- int len_out;
- xmlNodePtr root_ptr;
-
- (*extctr->init)(extctr, &recWord);
-
- if (extctr->flagShowRecords){
- xmlDocDumpMemory(doc, &buf_out, &len_out);
- fwrite(buf_out, len_out, 1, stdout);
- xmlFree(buf_out);
- }
- root_ptr = xmlDocGetRootElement(doc);
- if (root_ptr)
- index_record(tinfo, extctr, root_ptr, &recWord);
- else
- yaz_log(YLOG_WARN, "%s dom filter: "
- "No root for index XML record");
- }
-}
-
-
/* DOM filter style indexing */
static int attr_content_xml(struct _xmlAttr *attr, const char *name,
xmlChar **dst_content)
/* DOM filter style indexing */
static void index_value_of(struct filter_info *tinfo,
- struct recExtractCtrl *extctr,
+ struct recExtractCtrl *extctr,
+ RecWord* recword,
xmlNodePtr node,
xmlChar * index_p)
{
xmlChar type[256];
/* assingning text to be indexed */
- RecWord recWord;
- (*extctr->init)(extctr, &recWord);
- recWord.term_buf = (const char *)text;
- recWord.term_len = text_len;
+ recword->term_buf = (const char *)text;
+ recword->term_len = text_len;
/* parsing all index name/type pairs */
/* may not start with ' ' or ':' */
"INDEX '%s:%s' '%s'",
tinfo->fname, index, type, text);
- recWord.index_name = (const char *)index;
+ recword->index_name = (const char *)index;
if (type && *type)
- recWord.index_type = *type;
- (extctr->tokenAdd)(&recWord);
+ recword->index_type = *type;
+ (extctr->tokenAdd)(recword);
/* eat whitespaces */
if (*look && ' ' == *look && *(look+1)){
extctr->staticrank = atozint((const char *)rank_p);
/* if (!strcmp("update", type_str)) */
- /* index_node(tinfo, ctrl, ptr, recWord); */
+ /* index_node(tinfo, ctrl, ptr, recword); */
/* else if (!strcmp("delete", type_str)) */
/* yaz_log(YLOG_WARN, "dom filter delete: to be implemented"); */
/* else */
/* DOM filter style indexing */
static void process_xml_element_zebra_node(struct filter_info *tinfo,
struct recExtractCtrl *extctr,
+ RecWord* recword,
xmlNodePtr node)
{
if (node->type == XML_ELEMENT_NODE
struct _xmlAttr *attr;
for (attr = node->properties; attr; attr = attr->next){
if (attr_content_xml(attr, "name", &index_p)){
- index_value_of(tinfo, extctr, node, index_p);
+ index_value_of(tinfo, extctr, recword,node, index_p);
}
else
yaz_log(YLOG_WARN,"%s dom filter: "
/* DOM filter style indexing */
static void process_xml_element_node(struct filter_info *tinfo,
struct recExtractCtrl *extctr,
+ RecWord* recword,
xmlNodePtr node)
{
/* remember indexing instruction from PI to next element node */
xmlChar *index_p = 0;
- /* yaz_log(YLOG_DEBUG,"ELEM %s\n", xmlGetNodePath(node)); */
-
/* check if we are an element node in the special zebra namespace
and either set record data or index value-of node content*/
- process_xml_element_zebra_node(tinfo, extctr, node);
+ process_xml_element_zebra_node(tinfo, extctr, recword, node);
/* loop through kid nodes */
for (node = node->children; node; node = node->next)
else if (node->type == XML_ELEMENT_NODE){
/* if there was a PI index instruction before this element */
if (index_p){
- index_value_of(tinfo, extctr, node, index_p);
+ index_value_of(tinfo, extctr, recword, node, index_p);
index_p = 0;
}
- process_xml_element_node(tinfo, extctr, node);
+ process_xml_element_node(tinfo, extctr, recword,node);
}
else
continue;
struct recExtractCtrl *extctr,
xmlDocPtr doc)
{
- /* yaz_log(YLOG_DEBUG,"DOC %s\n", xmlGetNodePath((xmlNodePtr)doc)); */
-
xmlChar *buf_out;
int len_out;
+
+ /* only need to do the initialization once, reuse recword for all terms */
+ RecWord recword;
+ (*extctr->init)(extctr, &recword);
+
if (extctr->flagShowRecords){
xmlDocDumpMemory(doc, &buf_out, &len_out);
fwrite(buf_out, len_out, 1, stdout);
xmlFree(buf_out);
}
- process_xml_element_node(tinfo, extctr, (xmlNodePtr)doc);
+ process_xml_element_node(tinfo, extctr, &recword, (xmlNodePtr)doc);
}
projects / idzebra-moved-to-github.git / blobdiff