-/* $Id: mod_dom.c,v 1.12 2007-02-15 13:01:00 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;
;
else
yaz_log(YLOG_WARN, "%s: dom filter: "
- "bad attribute %s"
- " for <xslt>",
- tinfo->fname, attr->name);
+ "%s bad attribute @%s, "
+ "expected @stylesheet",
+ tinfo->fname,
+ xmlGetNodePath(ptr), attr->name);
if (p->stylesheet)
{
char tmp_xslt_full_name[1024];
NULL,
tmp_xslt_full_name))
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: "
+ yaz_log(YLOG_WARN, "%s: dom filter: "
"stylesheet %s not found in "
"path %s",
tinfo->fname,
tmp_xslt_full_name);
if (!p->stylesheet_xsp)
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: "
+ yaz_log(YLOG_WARN, "%s: dom filter: "
"could not parse xslt "
"stylesheet %s",
tinfo->fname, tmp_xslt_full_name);
}
else
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: "
- "missing attribute 'stylesheet' "
- "for element 'xslt'", tinfo->fname);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s missing attribute 'stylesheet' ",
+ tinfo->fname, xmlGetNodePath(ptr));
return ZEBRA_FAIL;
}
*l = p;
else
{
yaz_log(YLOG_LOG,
- "%s: dom filter: bad node '%s' for <conv>",
- tinfo->fname, ptr->name);
+ "%s: dom filter: "
+ "%s bad node '%s'",
+ tinfo->fname, xmlGetNodePath(ptr), ptr->name);
return ZEBRA_FAIL;
}
if (attr_content(attr, "charset", &input_charset))
;
else
- yaz_log(YLOG_WARN,
- "%s: dom filter: bad attribute %s"
- " for <marc>",
- tinfo->fname, attr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad attribute @%s,"
+ " expected @charset",
+ tinfo->fname,
+ xmlGetNodePath(ptr), attr->name);
}
iconv = yaz_iconv_open("utf-8", input_charset);
if (!iconv)
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: unsupported charset "
- "'%s' for <marc>",
- tinfo->fname, input_charset);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s unsupported @charset '%s'",
+ tinfo->fname, xmlGetNodePath(ptr),
+ input_charset);
return ZEBRA_FAIL;
}
else
if (attr_content(attr, "level", &level_str))
;
else
- yaz_log(YLOG_WARN,
- "%s: dom filter: bad attribute %s"
- " for <xmlreader>",
- tinfo->fname, attr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad attribute @%s,"
+ " expected @level",
+ tinfo->fname, xmlGetNodePath(ptr),
+ attr->name);
}
if (level_str)
p->u.xmlreader.split_level = atoi(level_str);
}
else
{
- yaz_log(YLOG_WARN, "%s: dom filter: bad input type %s",
- tinfo->fname, ptr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad element <%s>,"
+ " expected <marc>|<xmlreader>",
+ tinfo->fname, xmlGetNodePath(ptr), ptr->name);
return ZEBRA_FAIL;
}
}
else
tinfo->full_name = odr_strdup(tinfo->odr_config, tinfo->fname);
- yaz_log(YLOG_LOG, "dom filter: loading config file %s", tinfo->full_name);
+ yaz_log(YLOG_LOG, "%s dom filter: "
+ "loading config file %s", tinfo->fname, tinfo->full_name);
doc = xmlParseFile(tinfo->full_name);
if (!doc)
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: failed to parse config file %s",
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "failed to parse config file %s",
tinfo->fname, tinfo->full_name);
return ZEBRA_FAIL;
}
if (!ptr || ptr->type != XML_ELEMENT_NODE
|| XML_STRCMP(ptr->name, "dom"))
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: expected root element <dom>",
- tinfo->fname);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad root element <%s>,"
+ " expected root element <dom>",
+ tinfo->fname, xmlGetNodePath(ptr), ptr->name);
return ZEBRA_FAIL;
}
if (attr_content(attr, "name", &f->name))
;
else
- yaz_log(YLOG_WARN,
- "%s: dom filter: bad attribute %s"
- " for <extract>",
- tinfo->fname, attr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad attribute @%s"
+ " expected @name",
+ tinfo->fname,
+ xmlGetNodePath(ptr),attr->name);
}
parse_convert(tinfo, ptr->children, &f->convert);
else if (attr_content(attr, "name", &f->name))
;
else
- yaz_log(YLOG_WARN,
- "%s: dom filter: bad attribute %s"
- " for <retrieve>",
- tinfo->fname, attr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad attribute @%s"
+ " expected @identifier|@name",
+ tinfo->fname,
+ xmlGetNodePath(ptr),attr->name);
}
parse_convert(tinfo, ptr->children, &f->convert);
}
else if (!XML_STRCMP(ptr->name, "store"))
{
/*
- <retrieve name="F">
+ <store name="F">
<xslt stylesheet="some.xsl"/>
<xslt stylesheet="some.xsl"/>
</retrieve>
else if (attr_content(attr, "name", &name))
;
else
- yaz_log(YLOG_WARN,
- "%s: dom filter: bad attribute %s"
- " for <input>",
- tinfo->fname, attr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad attribute @%s"
+ " expected @syntax|@name",
+ tinfo->fname,
+ xmlGetNodePath(ptr),attr->name);
}
parse_input(tinfo, ptr->children, syntax, name);
}
else
{
- yaz_log(YLOG_WARN, "%s: dom filter: bad element %s",
- tinfo->fname, ptr->name);
+ yaz_log(YLOG_WARN, "%s: dom filter: "
+ "%s bad element <%s>,"
+ " expected <extract>|<input>|<retrieve>|<store>",
+ tinfo->fname, xmlGetNodePath(ptr), ptr->name);
return ZEBRA_FAIL;
}
}
}
-
-/* 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, "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, "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 ':' */
}
/* actually indexing the text given */
- /* printf("INDEX '%s:%s' '%s'\n", index, type, text); */
+ yaz_log(YLOG_DEBUG, "%s dom filter: "
+ "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)){
xmlChar * rank_p,
xmlChar * type_p)
{
- printf("RECORD id=%s rank=%s type=%s\n", id_p, rank_p, type_p);
+ yaz_log(YLOG_DEBUG, "%s dom filter: "
+ "RECORD id=%s rank=%s type=%s",
+ tinfo->fname, id_p, rank_p, type_p);
if (id_p)
sscanf((const char *)id_p, "%255s", extctr->match_criteria);
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
- // printf("%s: dom filter: s% bad attribute %s",
- // tinfo->fname, xmlGetNodePath(node)), nodeattr->name);
- printf("dom filter: %s bad attribute @%s, "
- "expected @name\n",
- xmlGetNodePath(node), attr->name);
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s bad attribute @%s, expected @name",
+ tinfo->fname, xmlGetNodePath(node), attr->name);
}
}
else if (0 == XML_STRCMP(node->name, "record")){
else if (attr_content_xml(attr, "rank", &rank_p))
;
else if (attr_content_xml(attr, "type", &type_p))
- ;
+ ;
else
- // printf("%s: dom filter: s% bad attribute %s",
- // tinfo->fname, xmlGetNodePath(node)), nodeattr->name);
- printf("dom filter: %s bad attribute @%s,"
- " expected @id|@rank|@type\n",
- xmlGetNodePath(node), attr->name);
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s bad attribute @%s,"
+ " expected @id|@rank|@type",
+ tinfo->fname, xmlGetNodePath(node), attr->name);
if (type_p && 0 != strcmp("update", (const char *)type_p))
- printf("dom filter: %s attribute @%s,"
- " only implemented '@type=\"update\"\n",
- xmlGetNodePath(node), attr->name);
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s attribute @%s,"
+ " only implemented '@type='update'",
+ tinfo->fname, xmlGetNodePath(node), attr->name);
}
set_record_info(tinfo, extctr, id_p, rank_p, type_p);
} else {
- // printf("%s: dom filter: s% bad attribute %s",
- // tinfo->fname, xmlGetNodePath(node)), nodeattr->name);
- printf("dom filter: %s bad element <%s>,"
- " expected <record>|<index> in namespace '%s'\n",
- xmlGetNodePath(node), node->name, zebra_dom_ns);
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s bad element <%s>,"
+ " expected <record>|<index> in namespace '%s'",
+ tinfo->fname, xmlGetNodePath(node),
+ node->name, zebra_dom_ns);
}
}
xmlChar **index_pp)
{
- /* printf("PI %s\n", xmlGetNodePath(node)); */
+ /* yaz_log(YLOG_DEBUG,"PI %s\n", xmlGetNodePath(node)); */
/* if right PI name, continue parsing PI */
if (0 == strcmp(zebra_pi_name, (const char *)node->name)){
while (*look && ' ' == *look && *(look+1))
look++;
- if (look && '\0' != *look){
- printf ("ERROR %s: content '%s'; can not parse '%s'\n",
- xmlGetNodePath(node), pi_p, look);
- } else {
- /* set_record_info(id, rank, type); */
+ if (look && '\0' != *look)
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s content '%s', can not parse '%s'",
+ tinfo->fname, xmlGetNodePath(node), pi_p, look);
+ else
set_record_info(tinfo, extctr, id, rank, 0);
- }
}
/* export index instructions to outside */
*index_pp = look;
-
- /* nor record, neither index */
- } else {
-
- printf ("ERROR %s: content '%s'; can not parse '%s'\n",
- xmlGetNodePath(node), pi_p, look);
- }
+ }
+ else
+ yaz_log(YLOG_WARN,"%s dom filter: "
+ "%s content '%s', can not parse '%s'",
+ tinfo->fname, xmlGetNodePath(node), pi_p, look);
}
}
/* 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;
- /* printf("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)
{
- /* printf("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);
}
{
int i;
- yaz_log(YLOG_WARN, "MARC: Skipping bad byte %d (0x%02X)",
- *buf & 0xff, *buf & 0xff);
+ yaz_log(YLOG_WARN, "%s dom filter: "
+ "MARC: Skipping bad byte %d (0x%02X)",
+ tinfo->fname, *buf & 0xff, *buf & 0xff);
for (i = 0; i<4; i++)
buf[i] = buf[i+1];
record_length = atoi_n (buf, 5);
if (record_length < 25)
{
- yaz_log (YLOG_WARN, "MARC record length < 25, is %d",
- record_length);
+ yaz_log (YLOG_WARN, "%s dom filter: "
+ "MARC record length < 25, is %d",
+ tinfo->fname, record_length);
return RECCTRL_EXTRACT_ERROR_GENERIC;
}
read_bytes = p->stream->readf(p->stream, buf+5, record_length-5);
if (read_bytes < record_length-5)
{
- yaz_log (YLOG_WARN, "Couldn't read whole MARC record");
+ yaz_log (YLOG_WARN, "%s dom filter: "
+ "Couldn't read whole MARC record",
+ tinfo->fname);
return RECCTRL_EXTRACT_ERROR_GENERIC;
}
r = yaz_marc_read_iso2709(input->u.marc.handle, buf, record_length);
if (r < record_length)
{
- yaz_log (YLOG_WARN, "Parsing of MARC record failed r=%d length=%d",
- r, record_length);
+ yaz_log (YLOG_WARN, "%s dom filter: "
+ "Parsing of MARC record failed r=%d length=%d",
+ tinfo->fname, r, record_length);
return RECCTRL_EXTRACT_ERROR_GENERIC;
}
else
projects / idzebra-moved-to-github.git / blobdiff