-/* $Id: mod_dom.c,v 1.4 2007-02-13 11:37:02 marc Exp $
+/* $Id: mod_dom.c,v 1.25 2007-03-01 10:35:46 adam Exp $
Copyright (C) 1995-2007
Index Data ApS
-This file is part of the Zebra server.
+ This file is part of the Zebra server.
-Zebra is free software; you can redistribute it and/or modify it under
-the terms of the GNU General Public License as published by the Free
-Software Foundation; either version 2, or (at your option) any later
-version.
+ Zebra is free software; you can redistribute it and/or modify it under
+ the terms of the GNU General Public License as published by the Free
+ Software Foundation; either version 2, or (at your option) any later
+ version.
-Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
-WARRANTY; without even the implied warranty of MERCHANTABILITY or
-FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
-for more details.
+ Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
+ WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
-You should have received a copy of the GNU General Public License
-along with this program; if not, write to the Free Software
-Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+ You should have received a copy of the GNU General Public License
+ 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>
#include <assert.h>
#include <ctype.h>
+#include <stdarg.h>
#include <yaz/diagbib1.h>
#include <yaz/tpath.h>
+#include <yaz/snprintf.h>
#include <libxml/xmlversion.h>
#include <libxml/parser.h>
#include <idzebra/util.h>
#include <idzebra/recctrl.h>
+/* DOM filter style indexing */
+#define ZEBRA_DOM_NS "http://indexdata.com/zebra-2.0"
+static const char *zebra_dom_ns = ZEBRA_DOM_NS;
+
+/* DOM filter style indexing */
+#define ZEBRA_PI_NAME "zebra-2.0"
+static const char *zebra_pi_name = ZEBRA_PI_NAME;
+
+
+
struct convert_s {
const char *stylesheet;
xsltStylesheetPtr stylesheet_xsp;
struct filter_retrieve *retrieve_list;
struct filter_input *input_list;
struct filter_store *store;
+ int record_info_invoked;
};
-#define XML_STRCMP(a,b) strcmp((char*)a, b)
-#define XML_STRLEN(a) strlen((char*)a)
-
-static void format_pi_zebra_err(char *err_str, const char *pi_str, const char *look)
-{
- strncpy(err_str, pi_str, look - pi_str);
- strncpy(err_str + (look - pi_str), "->", 2);
- strcpy(err_str + (look - pi_str + 2) , look);
-}
+#define XML_STRCMP(a,b) strcmp((char*)a, b)
+#define XML_STRLEN(a) strlen((char*)a)
-/*
-use PI parsing like this
+#define FOR_EACH_ELEMENT(ptr) for (; ptr; ptr = ptr->next) if (ptr->type == XML_ELEMENT_NODE)
- if (!parse_pi_zebra_20(pi_str, err_str))
- printf("ERROR '%s'\n", err_str);
-
-*/
+static void dom_log(int level, struct filter_info *tinfo, xmlNodePtr ptr,
+ const char *fmt, ...)
+#ifdef __GNUC__
+ __attribute__ ((format (printf, 4, 5)))
+#endif
+ ;
-static int parse_pi_zebra_20(const char *pi_str, char *err_str)
+static void dom_log(int level, struct filter_info *tinfo, xmlNodePtr ptr,
+ const char *fmt, ...)
{
- const char *look = pi_str;
- const char *bval;
- const char *eval;
-
- char value[256];
- char index[256];
- char type[256];
-
- *value = '\0';
- *index = '\0';
- *type = '\0';
-
- // parsing record instruction
- if (0 == strncmp(look, "record", 6)){
- look += 6;
- printf("record\n");
-
- if (*look && 0 == strncmp(look, " id=", 4)){
- look += 4;
- bval = look;
- printf(" id=");
- while (*look && ' ' != *look)
- look++;
- eval = look;
- strncpy(value, bval, eval - bval);
- value[eval - bval] = '\0';
-
- printf("%s\n", value);
- }
-
- if (*look && 0 == strncmp(look, " rank=", 6)){
- look += 6;
- bval = look;
- printf(" rank=");
- while (*look && ' ' != *look)
- look++;
- eval = look;
- strncpy(value, bval, eval - bval);
- value[eval - bval] = '\0';
-
- printf("%s\n", value);
- }
-
- if (!*look){
- return 1;
- }
- format_pi_zebra_err(err_str, pi_str, look);
- }
-
- // parsing index instruction
- else if (0 == strncmp(look, "index", 5)){
- look += 5;
- printf("index\n");
-
- // parsing all index name/type pairs
- while (*look && ' ' == *look && *(look+1)){
- look++;
-
- // index name must not start with ';' or ' '
- if (!*look || ':' == *look || ' ' == *look){
- format_pi_zebra_err(err_str, pi_str, look);
- return 0;
- }
-
- // setting name and type to zero
- *index = '\0';
- *type = '\0';
-
- // parsing one index name
- bval = look;
- while (*look && ':' != *look && ' ' != *look){
- look++;
- }
- eval = look;
- strncpy(index, bval, eval - bval);
- index[eval - bval] = '\0';
-
-
- // parsing one index type, if existing
- if (':' == *look){
- look++;
+ va_list ap;
+ char buf[4096];
- bval = look;
- while (*look && ' ' != *look){
- look++;
- }
- eval = look;
- strncpy(type, bval, eval - bval);
- type[eval - bval] = '\0';
- }
-
- printf(" %s:%s\n", index, type);
- }
-
- if (!*look){
- return 1;
- }
- format_pi_zebra_err(err_str, pi_str, look);
- }
-
-
- // remaining unparsed rest of PI
- else {
- format_pi_zebra_err(err_str, pi_str, look);
- }
-
- return 0;
+ va_start(ap, fmt);
+ yaz_vsnprintf(buf, sizeof(buf)-1, fmt, ap);
+ if (ptr)
+ {
+ yaz_log(level, "%s:%ld: %s", tinfo->fname ? tinfo->fname : "none",
+ xmlGetLineNo(ptr), buf);
+ }
+ else
+ {
+ yaz_log(level, "%s: %s", tinfo->fname ? tinfo->fname : "none", buf);
+ }
+ va_end(ap);
}
tinfo->input_list = 0;
tinfo->store = 0;
tinfo->doc_config = 0;
+ tinfo->record_info_invoked = 0;
#if YAZ_HAVE_EXSLT
exsltRegisterAll();
if (!XML_STRCMP(attr->name, name) && attr->children
&& attr->children->type == XML_TEXT_NODE)
{
- *dst_content = (const char *)(attr->children->content);
- return 1;
+ *dst_content = (const char *)(attr->children->content);
+ return 1;
}
return 0;
}
struct convert_s **l)
{
*l = 0;
- for(; ptr; ptr = ptr->next)
- {
- if (ptr->type != XML_ELEMENT_NODE)
- continue;
+ FOR_EACH_ELEMENT(ptr) {
if (!XML_STRCMP(ptr->name, "xslt"))
{
struct _xmlAttr *attr;
- struct convert_s *p = odr_malloc(tinfo->odr_config, sizeof(*p));
-
+ struct convert_s *p
+ = odr_malloc(tinfo->odr_config, sizeof(*p));
+
p->next = 0;
p->stylesheet = 0;
p->stylesheet_xsp = 0;
-
+
for (attr = ptr->properties; attr; attr = attr->next)
if (attr_content(attr, "stylesheet", &p->stylesheet))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <xslt>",
- tinfo->fname, attr->name);
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s", attr->name);
+ }
if (p->stylesheet)
{
char tmp_xslt_full_name[1024];
- if (!yaz_filepath_resolve(p->stylesheet, tinfo->profile_path,
- NULL, tmp_xslt_full_name))
+ if (!yaz_filepath_resolve(p->stylesheet,
+ tinfo->profile_path,
+ NULL,
+ tmp_xslt_full_name))
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: stylesheet %s not found in "
+ dom_log(YLOG_WARN, tinfo, 0,
+ "stylesheet %s not found in "
"path %s",
- tinfo->fname,
- p->stylesheet, tinfo->profile_path);
+ p->stylesheet,
+ tinfo->profile_path);
return ZEBRA_FAIL;
}
p->stylesheet_xsp
- = xsltParseStylesheetFile((const xmlChar*) tmp_xslt_full_name);
+ = xsltParseStylesheetFile((const xmlChar*)
+ tmp_xslt_full_name);
if (!p->stylesheet_xsp)
{
- yaz_log(YLOG_WARN,
- "%s: dom filter: could not parse xslt "
- "stylesheet %s",
- tinfo->fname, tmp_xslt_full_name);
+ dom_log(YLOG_WARN, tinfo, 0,
+ "could not parse xslt stylesheet %s",
+ tmp_xslt_full_name);
return ZEBRA_FAIL;
}
- }
- else
- {
- yaz_log(YLOG_WARN,
- "%s: dom filter: missing attribute 'stylesheet' "
- "for element 'xslt'", tinfo->fname);
- return ZEBRA_FAIL;
- }
- *l = p;
- l = &p->next;
+ }
+ else
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "missing attribute 'stylesheet' ");
+ return ZEBRA_FAIL;
+ }
+ *l = p;
+ l = &p->next;
}
else
{
- yaz_log(YLOG_LOG, "%s: dom filter: bad node '%s' for <conv>",
- tinfo->fname, ptr->name);
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad element '%s', expected <xslt>", ptr->name);
return ZEBRA_FAIL;
}
-
}
return ZEBRA_OK;
}
static ZEBRA_RES perform_convert(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
struct convert_s *convert,
const char **params,
xmlDocPtr *doc,
{
for (; convert; convert = convert->next)
{
+ xmlChar *buf_out = 0;
+ int len_out = 0;
xmlDocPtr res_doc = xsltApplyStylesheet(convert->stylesheet_xsp,
- *doc, params);
+ *doc, params);
if (last_xsp)
*last_xsp = convert->stylesheet_xsp;
+
xmlFreeDoc(*doc);
- *doc = res_doc;
+
+ /* now saving into buffer and re-reading into DOM to avoid annoing
+ XSLT problem with thrown-out indentation text nodes */
+ if (res_doc){
+ xsltSaveResultToString(&buf_out, &len_out, res_doc,
+ convert->stylesheet_xsp);
+ xmlFreeDoc(res_doc);
+ }
+
+
+ *doc = xmlParseDoc(buf_out);
+
+ /* writing debug info out */
+ if (extctr->flagShowRecords)
+ yaz_log(YLOG_LOG, "%s: XSLT %s \n %s",
+ tinfo->fname ? tinfo->fname : "(none)",
+ convert->stylesheet,
+ buf_out);
+
+ xmlFree(buf_out);
}
return ZEBRA_OK;
}
}
static ZEBRA_RES parse_input(struct filter_info *tinfo, xmlNodePtr ptr,
- const char *syntax,
- const char *name)
+ const char *syntax, const char *name)
{
- for (; ptr; ptr = ptr->next)
- {
- if (ptr->type != XML_ELEMENT_NODE)
- continue;
+ FOR_EACH_ELEMENT(ptr) {
if (!XML_STRCMP(ptr->name, "marc"))
{
yaz_iconv_t iconv = 0;
const char *input_charset = "marc-8";
- struct _xmlAttr *attr;
+ struct _xmlAttr *attr;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "charset", &input_charset))
+ for (attr = ptr->properties; attr; attr = attr->next)
+ {
+ if (attr_content(attr, "inputcharset", &input_charset))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <marc>",
- tinfo->fname, attr->name);
- }
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s, expected @inputcharset",
+ 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);
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "unsupported @charset '%s'", input_charset);
return ZEBRA_FAIL;
}
else
{
- struct filter_input *p = new_input(tinfo, DOM_INPUT_MARC);
+ struct filter_input *p
+ = new_input(tinfo, DOM_INPUT_MARC);
p->u.marc.handle = yaz_marc_create();
p->u.marc.iconv = iconv;
}
else if (!XML_STRCMP(ptr->name, "xmlreader"))
{
- struct filter_input *p = new_input(tinfo, DOM_INPUT_XMLREADER);
- struct _xmlAttr *attr;
+ struct filter_input *p
+ = new_input(tinfo, DOM_INPUT_XMLREADER);
+ struct _xmlAttr *attr;
const char *level_str = 0;
p->u.xmlreader.split_level = 0;
p->u.xmlreader.reader = 0;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "level", &level_str))
+ for (attr = ptr->properties; attr; attr = attr->next)
+ {
+ if (attr_content(attr, "level", &level_str))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <xmlreader>",
- tinfo->fname, attr->name);
- }
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s, expected @level",
+ 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);
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad element <%s>, expected <marc>|<xmlreader>",
+ 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;
+ return ZEBRA_FAIL;
}
/* save because we store ptrs to the content */
tinfo->doc_config = doc;
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);
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad root element <%s>, expected root element <dom>",
+ ptr->name);
return ZEBRA_FAIL;
}
- for (ptr = ptr->children; ptr; ptr = ptr->next)
- {
- if (ptr->type != XML_ELEMENT_NODE)
- continue;
+ ptr = ptr->children;
+ FOR_EACH_ELEMENT(ptr) {
if (!XML_STRCMP(ptr->name, "extract"))
{
/*
<xslt stylesheet="second.xsl"/>
</extract>
*/
- struct _xmlAttr *attr;
+ struct _xmlAttr *attr;
struct filter_extract *f =
odr_malloc(tinfo->odr_config, sizeof(*f));
tinfo->extract = f;
f->name = 0;
f->convert = 0;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "name", &f->name))
+ for (attr = ptr->properties; attr; attr = attr->next)
+ {
+ if (attr_content(attr, "name", &f->name))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <extract>",
- tinfo->fname, attr->name);
-
- }
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s, expected @name",
+ attr->name);
+ }
+ }
parse_convert(tinfo, ptr->children, &f->convert);
}
- else if (!XML_STRCMP(ptr->name, "retrieve"))
- {
+ else if (!XML_STRCMP(ptr->name, "retrieve"))
+ {
/*
<retrieve name="F">
<xslt stylesheet="some.xsl"/>
<xslt stylesheet="some.xsl"/>
</retrieve>
*/
- struct _xmlAttr *attr;
+ struct _xmlAttr *attr;
struct filter_retrieve **fp = &tinfo->retrieve_list;
struct filter_retrieve *f =
odr_malloc(tinfo->odr_config, sizeof(*f));
f->convert = 0;
f->next = 0;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "identifier", &f->identifier))
+ for (attr = ptr->properties; attr; attr = attr->next)
+ {
+ if (attr_content(attr, "identifier",
+ &f->identifier))
;
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);
- }
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s, expected @identifier|@name",
+ attr->name);
+ }
+ }
parse_convert(tinfo, ptr->children, &f->convert);
- }
- else if (!XML_STRCMP(ptr->name, "store"))
- {
+ }
+ else if (!XML_STRCMP(ptr->name, "store"))
+ {
/*
- <retrieve name="F">
- <xslt stylesheet="some.xsl"/>
- <xslt stylesheet="some.xsl"/>
- </retrieve>
+ <store name="F">
+ <xslt stylesheet="some.xsl"/>
+ <xslt stylesheet="some.xsl"/>
+ </retrieve>
*/
struct filter_store *f =
odr_malloc(tinfo->odr_config, sizeof(*f));
f->convert = 0;
parse_convert(tinfo, ptr->children, &f->convert);
}
- else if (!XML_STRCMP(ptr->name, "input"))
- {
+ else if (!XML_STRCMP(ptr->name, "input"))
+ {
/*
<input syntax="xml">
<xmlreader level="1"/>
<marc inputcharset="marc-8"/>
</input>
*/
- struct _xmlAttr *attr;
+ struct _xmlAttr *attr;
const char *syntax = 0;
const char *name = 0;
- for (attr = ptr->properties; attr; attr = attr->next)
- {
- if (attr_content(attr, "syntax", &syntax))
+ for (attr = ptr->properties; attr; attr = attr->next)
+ {
+ if (attr_content(attr, "syntax", &syntax))
;
else if (attr_content(attr, "name", &name))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <input>",
- tinfo->fname, attr->name);
- }
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad attribute @%s, expected @syntax|@name",
+ attr->name);
+ }
+ }
parse_input(tinfo, ptr->children, syntax, name);
- }
- else
- {
- yaz_log(YLOG_WARN, "%s: dom filter: bad element %s",
- tinfo->fname, ptr->name);
- return ZEBRA_FAIL;
- }
+ }
+ else
+ {
+ dom_log(YLOG_WARN, tinfo, ptr,
+ "bad element <%s>, "
+ "expected <extract>|<input>|<retrieve>|<store>",
+ ptr->name);
+ return ZEBRA_FAIL;
+ }
}
return ZEBRA_OK;
}
for (; f; f = f->next)
{
/* find requested schema */
- if (est)
- {
- if (f->identifier && !strcmp(f->identifier, est))
+ if (est)
+ {
+ if (f->identifier && !strcmp(f->identifier, est))
return f;
- if (f->name && !strcmp(f->name, est))
+ if (f->name && !strcmp(f->name, est))
return f;
- }
+ }
}
return 0;
}
return 0;
}
-static void index_cdata(struct filter_info *tinfo, struct recExtractCtrl *ctrl,
- xmlNodePtr ptr, RecWord *recWord)
+
+/* DOM filter style indexing */
+static int attr_content_xml(struct _xmlAttr *attr, const char *name,
+ xmlChar **dst_content)
{
- for(; ptr; ptr = ptr->next)
+ if (0 == XML_STRCMP(attr->name, name) && attr->children
+ && attr->children->type == XML_TEXT_NODE)
{
- 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);
+ *dst_content = (attr->children->content);
+ return 1;
}
+ return 0;
}
-#define ZEBRA_SCHEMA_XSLT_NS "http://indexdata.dk/zebra/xslt/1"
+/* DOM filter style indexing */
+static void index_value_of(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
+ RecWord* recword,
+ xmlNodePtr node,
+ xmlChar * index_p)
+{
+ if (tinfo->record_info_invoked == 1)
+ {
+ xmlChar *text = xmlNodeGetContent(node);
+ size_t text_len = strlen((const char *)text);
+
+ yaz_log(YLOG_LOG, "Indexing :%.*s:", text_len, text);
+
+ /* if there is no text, we do not need to proceed */
+ if (text_len)
+ {
+ xmlChar *look = index_p;
+ xmlChar *bval;
+ xmlChar *eval;
+
+ xmlChar index[256];
+ xmlChar type[256];
+
+ /* assingning text to be indexed */
+ recword->term_buf = (const char *)text;
+ recword->term_len = text_len;
+
+ /* parsing all index name/type pairs */
+ /* may not start with ' ' or ':' */
+ while (*look && ' ' != *look && ':' != *look)
+ {
+ /* setting name and type to zero */
+ *index = '\0';
+ *type = '\0';
+
+ /* parsing one index name */
+ bval = look;
+ while (*look && ':' != *look && ' ' != *look)
+ {
+ look++;
+ }
+ eval = look;
+ strncpy((char *)index, (const char *)bval, eval - bval);
+ index[eval - bval] = '\0';
+
+
+ /* parsing one index type, if existing */
+ if (':' == *look)
+ {
+ look++;
+
+ bval = look;
+ while (*look && ' ' != *look)
+ {
+ look++;
+ }
+ eval = look;
+ strncpy((char *)type, (const char *)bval, eval - bval);
+ type[eval - bval] = '\0';
+ }
+
+ /* actually indexing the text given */
+ dom_log(YLOG_DEBUG, tinfo, 0,
+ "INDEX '%s:%s' '%s'",
+ index ? (const char *) index : "null",
+ type ? (const char *) type : "null",
+ text ? (const char *) text : "null");
+
+ recword->index_name = (const char *)index;
+ if (type && *type)
+ recword->index_type = *type;
+
+ /* writing debug out */
+ if (extctr->flagShowRecords)
+ dom_log(YLOG_LOG, tinfo, 0,
+ "INDEX '%s:%s' '%s'",
+ index ? (const char *) index : "null",
+ type ? (const char *) type : "null",
+ text ? (const char *) text : "null");
+
+ /* actually indexing the text given */
+ recword->index_name = (const char *)index;
+ if (type && *type)
+ recword->index_type = *type;
+ (extctr->tokenAdd)(recword);
+
+ /* eat whitespaces */
+ if (*look && ' ' == *look && *(look+1))
+ {
+ look++;
+ }
+ }
+ }
+ xmlFree(text);
+ }
+}
+
+
+/* DOM filter style indexing */
+static void set_record_info(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
+ xmlNodePtr node,
+ xmlChar * id_p,
+ xmlChar * rank_p,
+ xmlChar * type_p)
+{
+
+ /* writing debug info out */
+ if (extctr->flagShowRecords)
+ dom_log(YLOG_LOG, tinfo, 0,
+ "RECORD id=%s rank=%s type=%s",
+ id_p ? (const char *) id_p : "(null)",
+ rank_p ? (const char *) rank_p : "(null)",
+ type_p ? (const char *) type_p : "(null)");
+
+
+ if (id_p)
+ sscanf((const char *)id_p, "%255s", extctr->match_criteria);
-static const char *zebra_xslt_ns = ZEBRA_SCHEMA_XSLT_NS;
+ if (rank_p)
+ extctr->staticrank = atozint((const char *)rank_p);
-static void index_node(struct filter_info *tinfo, struct recExtractCtrl *ctrl,
- xmlNodePtr ptr, RecWord *recWord)
+ /* if (!strcmp("update", type_str)) */
+ /* index_node(tinfo, ctrl, ptr, recword); */
+ /* else if (!strcmp("delete", type_str)) */
+ /* dom_log(YLOG_WARN, tinfo, ptr, "dom filter delete: to be implemented"); */
+ /* else */
+ /* dom_log(YLOG_WARN, tinfo, ptr, "dom filter: unknown record type '%s'", */
+ /* type_str); */
+ if (tinfo->record_info_invoked == 1)
+ {
+ /* warn about multiple only once */
+ dom_log(YLOG_WARN, tinfo, node, "multiple record elements");
+ }
+ tinfo->record_info_invoked++;
+
+}
+
+
+/* DOM filter style indexing */
+static void process_xml_element_zebra_node(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
+ RecWord* recword,
+ xmlNodePtr node)
{
- for(; ptr; ptr = ptr->next)
+ if (node->type == XML_ELEMENT_NODE && node->ns && node->ns->href
+ && 0 == XML_STRCMP(node->ns->href, zebra_dom_ns))
{
- 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))
+ if (0 == XML_STRCMP(node->name, "index"))
+ {
+ xmlChar *index_p = 0;
+
+ struct _xmlAttr *attr;
+ for (attr = node->properties; attr; attr = attr->next)
+ {
+ if (attr_content_xml(attr, "name", &index_p))
+ {
+ index_value_of(tinfo, extctr, recword,node, index_p);
+ }
+ else
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "bad attribute @%s, expected @name",
+ attr->name);
+ }
+ }
+ }
+ else if (0 == XML_STRCMP(node->name, "record"))
+ {
+ xmlChar *id_p = 0;
+ xmlChar *rank_p = 0;
+ xmlChar *type_p = 0;
+
+ struct _xmlAttr *attr;
+ for (attr = node->properties; attr; attr = attr->next)
+ {
+ if (attr_content_xml(attr, "id", &id_p))
;
- else if (attr_content(attr, "xpath", &xpath_str))
+ else if (attr_content_xml(attr, "rank", &rank_p))
;
- else if (attr_content(attr, "type", &type_str))
+ else if (attr_content_xml(attr, "type", &type_p))
;
else
- yaz_log(YLOG_WARN, "%s: dom filter: bad attribute %s"
- " for <index>",
- tinfo->fname, attr->name);
- }
- if (name_str)
- {
- int prev_type = recWord->index_type; /* save default type */
-
- if (type_str && *type_str)
- recWord->index_type = *type_str; /* type was given */
- recWord->index_name = name_str;
- index_cdata(tinfo, ctrl, ptr->children, recWord);
-
- recWord->index_type = prev_type; /* restore it again */
- }
- }
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "bad attribute @%s, expected @id|@rank|@type",
+ attr->name);
+ }
+
+ if (type_p && 0 != strcmp("update", (const char *)type_p))
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "attribute @%s, only implemented '@type='update'",
+ attr->name);
+ }
+ }
+ set_record_info(tinfo, extctr, node, id_p, rank_p, type_p);
+ }
+ else
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "bad element <%s>,"
+ " expected <record>|<index> in namespace '%s'",
+ node->name, zebra_dom_ns);
+ }
}
}
-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"))
+/* DOM filter style indexing */
+static void process_xml_pi_node(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
+ xmlNodePtr node,
+ xmlChar **index_pp)
+{
+ /* if right PI name, continue parsing PI */
+ if (0 == strcmp(zebra_pi_name, (const char *)node->name))
{
- 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;
+ xmlChar *pi_p = node->content;
+ xmlChar *look = pi_p;
+
+ xmlChar *bval;
+ xmlChar *eval;
+
+ /* parsing PI record instructions */
+ if (0 == strncmp((const char *)look, "record", 6))
+ {
+ xmlChar id[256];
+ xmlChar rank[256];
+ xmlChar type[256];
+
+ *id = '\0';
+ *rank = '\0';
+ *type = '\0';
+
+ look += 6;
+
+ /* eat whitespace */
+ while (*look && ' ' == *look && *(look+1))
+ look++;
+
+ /* parse possible id */
+ if (*look && 0 == strncmp((const char *)look, "id=", 3))
+ {
+ look += 3;
+ bval = look;
+ while (*look && ' ' != *look)
+ look++;
+ eval = look;
+ strncpy((char *)id, (const char *)bval, eval - bval);
+ id[eval - bval] = '\0';
+ }
+
+ /* eat whitespace */
+ while (*look && ' ' == *look && *(look+1))
+ look++;
+
+ /* parse possible rank */
+ if (*look && 0 == strncmp((const char *)look, "rank=", 5))
+ {
+ look += 6;
+ bval = look;
+ while (*look && ' ' != *look)
+ look++;
+ eval = look;
+ strncpy((char *)rank, (const char *)bval, eval - bval);
+ rank[eval - bval] = '\0';
+ }
+
+ /* eat whitespace */
+ while (*look && ' ' == *look && *(look+1))
+ look++;
+
+ if (look && '\0' != *look)
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "content '%s', can not parse '%s'",
+ pi_p, look);
+ }
+ else
+ set_record_info(tinfo, extctr, node, id, rank, 0);
+
+ }
+ /* parsing index instruction */
+ else if (0 == strncmp((const char *)look, "index", 5))
+ {
+ look += 5;
+
+ /* eat whitespace */
+ while (*look && ' ' == *look && *(look+1))
+ look++;
+
+ /* export index instructions to outside */
+ *index_pp = look;
+ }
+ else
+ {
+ dom_log(YLOG_WARN, tinfo, node,
+ "content '%s', can not parse '%s'",
+ pi_p, look);
+ }
}
+}
- 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);
+/* 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;
+
+ /* 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, recword, node);
+
+ /* loop through kid nodes */
+ for (node = node->children; node; node = node->next)
+ {
+ /* check and set PI record and index index instructions */
+ if (node->type == XML_PI_NODE)
+ {
+ process_xml_pi_node(tinfo, extctr, node, &index_p);
+ }
+ 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, recword, node, index_p);
+ index_p = 0;
+ }
+ process_xml_element_node(tinfo, extctr, recword,node);
+ }
+ else
+ continue;
+ }
}
-static void extract_doc_alvis(struct filter_info *tinfo,
- struct recExtractCtrl *recctr,
- xmlDocPtr doc)
+/* DOM filter style indexing */
+static void extract_dom_doc_node(struct filter_info *tinfo,
+ struct recExtractCtrl *extctr,
+ xmlDocPtr doc)
{
- if (doc){
- RecWord recWord;
- xmlChar *buf_out;
- int len_out;
- xmlNodePtr root_ptr;
+ /* only need to do the initialization once, reuse recword for all terms */
+ RecWord recword;
+ (*extctr->init)(extctr, &recword);
- (*recctr->init)(recctr, &recWord);
-
- if (recctr->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, recctr, root_ptr, &recWord);
- else
- yaz_log(YLOG_WARN, "No root for index XML record");
+ /*
+ if (extctr->flagShowRecords)
+ {
+ xmlDocDumpMemory(doc, &buf_out, &len_out);
+ fwrite(buf_out, len_out, 1, stdout);
+ xmlFree(buf_out);
}
+ */
+ tinfo->record_info_invoked = 0;
+ process_xml_element_node(tinfo, extctr, &recword, (xmlNodePtr)doc);
}
+
+
static int convert_extract_doc(struct filter_info *tinfo,
struct filter_input *input,
struct recExtractCtrl *p,
xmlDocPtr doc)
{
- /* RecWord recWord; */
xmlChar *buf_out;
int len_out;
const char *params[10];
xmlDocPtr store_doc = 0;
params[0] = 0;
- set_param_str(params, "schema", zebra_xslt_ns, tinfo->odr_record);
+ set_param_str(params, "schema", zebra_dom_ns, tinfo->odr_record);
/* input conversion */
- perform_convert(tinfo, input->convert, params, &doc, 0);
-
- /* (*p->init)(p, &recWord); */
+ perform_convert(tinfo, p, input->convert, params, &doc, 0);
if (tinfo->store)
{
/* store conversion */
store_doc = xmlCopyDoc(doc, 1);
- perform_convert(tinfo, tinfo->store->convert,
+ perform_convert(tinfo, p, tinfo->store->convert,
params, &store_doc, &last_xsp);
}
store_doc ? store_doc : doc, last_xsp);
else
xmlDocDumpMemory(store_doc ? store_doc : doc, &buf_out, &len_out);
- if (p->flagShowRecords)
- fwrite(buf_out, len_out, 1, stdout);
+
+ /* if (p->flagShowRecords)
+ fwrite(buf_out, len_out, 1, stdout); */
+
(*p->setStoreData)(p, buf_out, len_out);
xmlFree(buf_out);
xmlFreeDoc(store_doc);
/* extract conversion */
- perform_convert(tinfo, tinfo->extract->convert, params, &doc, 0);
+ perform_convert(tinfo, p, tinfo->extract->convert, params, &doc, 0);
+ /* finally, do the indexing */
if (doc)
- {
-
- extract_doc_alvis(tinfo, p, doc);
-
- /*
- xmlNodePtr root_ptr;
- if (p->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, p, root_ptr, &recWord);
- else
- {
- yaz_log(YLOG_WARN, "No root for index XML record");
- }
- */
+ extract_dom_doc_node(tinfo, p, doc);
+ if (doc)
xmlFreeDoc(doc);
- }
+ if (tinfo->record_info_invoked == 0)
+ return RECCTRL_EXTRACT_SKIP;
return RECCTRL_EXTRACT_OK;
}
if (p->first_record)
{
- if (input->u.xmlreader.reader)
- xmlFreeTextReader(input->u.xmlreader.reader);
- input->u.xmlreader.reader = xmlReaderForIO(ioread_ex, ioclose_ex,
+ if (input->u.xmlreader.reader)
+ xmlFreeTextReader(input->u.xmlreader.reader);
+ input->u.xmlreader.reader = xmlReaderForIO(ioread_ex, ioclose_ex,
p /* I/O handler */,
0 /* URL */,
0 /* encoding */,
ret = xmlTextReaderRead(input->u.xmlreader.reader);
while (ret == 1)
{
- int type = xmlTextReaderNodeType(input->u.xmlreader.reader);
- int depth = xmlTextReaderDepth(input->u.xmlreader.reader);
- if (type == XML_READER_TYPE_ELEMENT &&
+ int type = xmlTextReaderNodeType(input->u.xmlreader.reader);
+ int depth = xmlTextReaderDepth(input->u.xmlreader.reader);
+ if (type == XML_READER_TYPE_ELEMENT &&
input->u.xmlreader.split_level == depth)
- {
- xmlNodePtr ptr = xmlTextReaderExpand(input->u.xmlreader.reader);
+ {
+ xmlNodePtr ptr
+ = xmlTextReaderExpand(input->u.xmlreader.reader);
if (ptr)
{
xmlNodePtr ptr2 = xmlCopyNode(ptr, 1);
xmlDocSetRootElement(doc, ptr2);
+ /* writing debug info out */
+ if (p->flagShowRecords){
+ xmlChar *buf_out = 0;
+ int len_out = 0;
+ xmlDocDumpMemory(doc, &buf_out, &len_out);
+ yaz_log(YLOG_LOG, "%s: XMLREADER depth: %i\n%s",
+ tinfo->fname ? tinfo->fname : "(none)",
+ depth,
+ buf_out);
+ xmlFree(buf_out);
+ }
+
return convert_extract_doc(tinfo, input, p, doc);
}
else
input->u.xmlreader.reader = 0;
return RECCTRL_EXTRACT_ERROR_GENERIC;
}
- }
- ret = xmlTextReaderRead(input->u.xmlreader.reader);
+ }
+ ret = xmlTextReaderRead(input->u.xmlreader.reader);
}
xmlFreeTextReader(input->u.xmlreader.reader);
input->u.xmlreader.reader = 0;
{
if (p->first_record) /* only one record per stream */
{
- xmlDocPtr doc = xmlReadIO(ioread_ex, ioclose_ex, p /* I/O handler */,
+ xmlDocPtr doc = xmlReadIO(ioread_ex, ioclose_ex,
+ p /* I/O handler */,
0 /* URL */,
0 /* encoding */,
XML_PARSE_XINCLUDE|XML_PARSE_NOENT);
{
int i;
- yaz_log(YLOG_WARN, "MARC: Skipping bad byte %d (0x%02X)",
+ dom_log(YLOG_WARN, tinfo, 0,
+ "MARC: Skipping bad byte %d (0x%02X)",
*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);
+ dom_log(YLOG_WARN, tinfo, 0,
+ "MARC record length < 25, is %d", 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");
+ dom_log(YLOG_WARN, tinfo, 0,
+ "couldn't read whole MARC record");
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",
+ dom_log (YLOG_WARN, tinfo, 0,
+ "parsing of MARC record failed r=%d length=%d",
r, record_length);
return RECCTRL_EXTRACT_ERROR_GENERIC;
}
static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p)
{
- /* const char *esn = zebra_xslt_ns; */
+ /* const char *esn = zebra_dom_ns; */
const char *esn = 0;
const char *params[32];
struct filter_info *tinfo = clientData;
if (p->comp)
{
- if (p->comp->which == Z_RecordComp_simple
- && p->comp->u.simple->which == Z_ElementSetNames_generic)
- {
- esn = p->comp->u.simple->u.generic;
- }
- else if (p->comp->which == Z_RecordComp_complex
- && p->comp->u.complex->generic->elementSpec
- && p->comp->u.complex->generic->elementSpec->which ==
- Z_ElementSpec_elementSetName)
- {
- esn = p->comp->u.complex->generic->elementSpec->u.elementSetName;
- }
+ if (p->comp->which == Z_RecordComp_simple
+ && p->comp->u.simple->which == Z_ElementSetNames_generic)
+ {
+ esn = p->comp->u.simple->u.generic;
+ }
+ else if (p->comp->which == Z_RecordComp_complex
+ && p->comp->u.complex->generic->elementSpec
+ && p->comp->u.complex->generic->elementSpec->which ==
+ Z_ElementSpec_elementSetName)
+ {
+ esn = p->comp->u.complex->generic->elementSpec->u.elementSetName;
+ }
}
retrieve = lookup_retrieve(tinfo, esn);
if (!retrieve)
{
- p->diagnostic =
- YAZ_BIB1_SPECIFIED_ELEMENT_SET_NAME_NOT_VALID_FOR_SPECIFIED_;
- return 0;
+ p->diagnostic =
+ YAZ_BIB1_SPECIFIED_ELEMENT_SET_NAME_NOT_VALID_FOR_SPECIFIED_;
+ return 0;
}
params[0] = 0;
XML_PARSE_XINCLUDE|XML_PARSE_NOENT);
if (!doc)
{
- p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS;
- return 0;
+ p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS;
+ return 0;
}
/* retrieve conversion */
- perform_convert(tinfo, retrieve->convert, params, &doc, &last_xsp);
+ perform_convert(tinfo, 0, retrieve->convert, params, &doc, &last_xsp);
if (!doc)
{
- p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS;
+ p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS;
}
else if (p->input_format == VAL_NONE || p->input_format == VAL_TEXT_XML)
{
- xmlChar *buf_out;
- int len_out;
+ xmlChar *buf_out;
+ int len_out;
if (last_xsp)
xsltSaveResultToString(&buf_out, &len_out, doc, last_xsp);
else
- xmlDocDumpMemory(doc, &buf_out, &len_out);
+ xmlDocDumpMemory(doc, &buf_out, &len_out);
- p->output_format = VAL_TEXT_XML;
- p->rec_len = len_out;
- p->rec_buf = odr_malloc(p->odr, p->rec_len);
- memcpy(p->rec_buf, buf_out, p->rec_len);
- xmlFree(buf_out);
+ p->output_format = VAL_TEXT_XML;
+ p->rec_len = len_out;
+ p->rec_buf = odr_malloc(p->odr, p->rec_len);
+ memcpy(p->rec_buf, buf_out, p->rec_len);
+ xmlFree(buf_out);
}
else if (p->output_format == VAL_SUTRS)
{
- xmlChar *buf_out;
- int len_out;
+ xmlChar *buf_out;
+ int len_out;
if (last_xsp)
xsltSaveResultToString(&buf_out, &len_out, doc, last_xsp);
else
- xmlDocDumpMemory(doc, &buf_out, &len_out);
+ xmlDocDumpMemory(doc, &buf_out, &len_out);
- p->output_format = VAL_SUTRS;
- p->rec_len = len_out;
- p->rec_buf = odr_malloc(p->odr, p->rec_len);
- memcpy(p->rec_buf, buf_out, p->rec_len);
+ p->output_format = VAL_SUTRS;
+ p->rec_len = len_out;
+ p->rec_buf = odr_malloc(p->odr, p->rec_len);
+ memcpy(p->rec_buf, buf_out, p->rec_len);
- xmlFree(buf_out);
+ xmlFree(buf_out);
}
else
{
- p->diagnostic = YAZ_BIB1_RECORD_SYNTAX_UNSUPP;
+ p->diagnostic = YAZ_BIB1_RECORD_SYNTAX_UNSUPP;
}
xmlFreeDoc(doc);
return 0;