X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fmod_dom.c;h=ea09c31013562878d5cd8d11f38736a0e1c62819;hb=1872e3fc60b482771bbd1cb4b0290b8d6a9ef5d0;hp=d93853279740aad8441f9b114b14fb2ad4e06fe3;hpb=1995ca9c4d5451771944163fc302d21578055241;p=idzebra-moved-to-github.git diff --git a/index/mod_dom.c b/index/mod_dom.c index d938532..ea09c31 100644 --- a/index/mod_dom.c +++ b/index/mod_dom.c @@ -1,5 +1,4 @@ - -/* $Id: mod_dom.c,v 1.28 2007-03-06 08:48:57 adam Exp $ +/* $Id: mod_dom.c,v 1.35 2007-04-16 08:44:31 adam Exp $ Copyright (C) 1995-2007 Index Data ApS @@ -44,6 +43,7 @@ #include #include +#include /* DOM filter style indexing */ #define ZEBRA_DOM_NS "http://indexdata.com/zebra-2.0" @@ -86,14 +86,14 @@ struct filter_input { int type; union { struct { + xmlTextReaderPtr reader; + int split_level; + } xmlreader; + struct { const char *input_charset; yaz_marc_t handle; yaz_iconv_t iconv; } marc; - struct { - xmlTextReaderPtr reader; - int split_level; - } xmlreader; } u; struct filter_input *next; }; @@ -631,6 +631,13 @@ static ZEBRA_RES parse_dom(struct filter_info *tinfo, const char *fname) return ZEBRA_FAIL; } } + if (!tinfo->input_list) + { + struct filter_input *p + = new_input(tinfo, DOM_INPUT_XMLREADER); + p->u.xmlreader.split_level = 0; + p->u.xmlreader.reader = 0; + } return ZEBRA_OK; } @@ -697,12 +704,12 @@ static int ioclose_ex(void *context) /* DOM filter style indexing */ static int attr_content_xml(struct _xmlAttr *attr, const char *name, - xmlChar **dst_content) + const char **dst_content) { if (0 == XML_STRCMP(attr->name, name) && attr->children && attr->children->type == XML_TEXT_NODE) { - *dst_content = (attr->children->content); + *dst_content = (const char *) (attr->children->content); return 1; } return 0; @@ -714,7 +721,7 @@ static void index_value_of(struct filter_info *tinfo, struct recExtractCtrl *extctr, RecWord* recword, xmlNodePtr node, - xmlChar * index_p) + const char *index_p) { if (tinfo->record_info_invoked == 1) { @@ -724,9 +731,9 @@ static void index_value_of(struct filter_info *tinfo, /* if there is no text, we do not need to proceed */ if (text_len) { - xmlChar *look = index_p; - xmlChar *bval; - xmlChar *eval; + const char *look = index_p; + const char *bval; + const char *eval; xmlChar index[256]; xmlChar type[256]; @@ -795,7 +802,7 @@ static void index_value_of(struct filter_info *tinfo, (extctr->tokenAdd)(recword); /* eat whitespaces */ - if (*look && ' ' == *look && *(look+1)) + if (*look && ' ' == *look) { look++; } @@ -810,33 +817,42 @@ static void index_value_of(struct filter_info *tinfo, static void set_record_info(struct filter_info *tinfo, struct recExtractCtrl *extctr, xmlNodePtr node, - xmlChar * id_p, - xmlChar * rank_p, - xmlChar * type_p) + const char * id_p, + const char * rank_p, + const char * type_p) { - /* writing debug info out */ - if (extctr->flagShowRecords) - dom_log(YLOG_LOG, tinfo, 0, + if (extctr && extctr->flagShowRecords) + dom_log(YLOG_LOG, tinfo, node, "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) + if (id_p && *id_p) sscanf((const char *)id_p, "%255s", extctr->match_criteria); - if (rank_p) + if (rank_p && *rank_p) extctr->staticrank = atozint((const char *)rank_p); - /* 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 (type_p && *type_p) + { + enum zebra_recctrl_action_t action = action_update; + if (!strcmp(type_p, "insert")) + action = action_insert; + else if (!strcmp(type_p, "delete")) + action = action_delete; + else if (!strcmp(type_p, "replace")) + action = action_replace; + else if (!strcmp(type_p, "update")) + action = action_update; + else + dom_log(YLOG_WARN, tinfo, node, "bad @type value: %s", type_p); + extctr->action = action; + yaz_log(YLOG_LOG, "In mod_dom.c: setting action to %d", action); + } + if (tinfo->record_info_invoked == 1) { /* warn about multiple only once */ @@ -858,14 +874,14 @@ static void process_xml_element_zebra_node(struct filter_info *tinfo, { if (0 == XML_STRCMP(node->name, "index")) { - xmlChar *index_p = 0; + const char *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); + index_value_of(tinfo, extctr, recword, node, index_p); } else { @@ -877,9 +893,9 @@ static void process_xml_element_zebra_node(struct filter_info *tinfo, } else if (0 == XML_STRCMP(node->name, "record")) { - xmlChar *id_p = 0; - xmlChar *rank_p = 0; - xmlChar *type_p = 0; + const char *id_p = 0; + const char *rank_p = 0; + const char *type_p = 0; struct _xmlAttr *attr; for (attr = node->properties; attr; attr = attr->next) @@ -896,13 +912,6 @@ static void process_xml_element_zebra_node(struct filter_info *tinfo, "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); } @@ -916,80 +925,74 @@ static void process_xml_element_zebra_node(struct filter_info *tinfo, } } +static int attr_content_pi(const char **c_ptr, const char *name, + char *value, size_t value_max) +{ + size_t name_len = strlen(name); + const char *look = *c_ptr; + int ret = 0; + + *value = '\0'; + while (*look && ' ' == *look) + look++; + if (strlen(look) > name_len) + { + if (look[name_len] == '=' && !memcmp(look, name, name_len)) + { + size_t i = 0; + look += name_len+1; + while (*look && ' ' != *look) + { + if (i < value_max-1) + value[i++] = *look; + look++; + } + value[i] = '\0'; + ret = 1; + } + } + while (*look && ' ' == *look) + look++; + *c_ptr = look; + return ret; +} /* DOM filter style indexing */ static void process_xml_pi_node(struct filter_info *tinfo, struct recExtractCtrl *extctr, xmlNodePtr node, - xmlChar **index_pp) + const char **index_pp) { /* if right PI name, continue parsing PI */ if (0 == strcmp(zebra_pi_name, (const char *)node->name)) { xmlChar *pi_p = node->content; - xmlChar *look = pi_p; + const char *look = (const char *) node->content; - 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]; - + char id[256]; + char rank[256]; + char 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); - + while (*look) + if (attr_content_pi(&look, "id", id, sizeof(id))) + ; + else if (attr_content_pi(&look, "rank", rank, sizeof(rank))) + ; + else if (attr_content_pi(&look, "type", type, sizeof(type))) + { + dom_log(YLOG_WARN, tinfo, node, + "content '%s', can not parse '%s'", + pi_p, look); + break; + } + set_record_info(tinfo, extctr, node, id, rank, type); } /* parsing index instruction */ else if (0 == strncmp((const char *)look, "index", 5)) @@ -997,7 +1000,7 @@ static void process_xml_pi_node(struct filter_info *tinfo, look += 5; /* eat whitespace */ - while (*look && ' ' == *look && *(look+1)) + while (*look && ' ' == *look) look++; /* export index instructions to outside */ @@ -1019,7 +1022,7 @@ static void process_xml_element_node(struct filter_info *tinfo, xmlNodePtr node) { /* remember indexing instruction from PI to next element node */ - xmlChar *index_p = 0; + const char *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*/ @@ -1142,8 +1145,9 @@ static int extract_xml_split(struct filter_info *tinfo, p /* I/O handler */, 0 /* URL */, 0 /* encoding */, - XML_PARSE_XINCLUDE| - XML_PARSE_NOENT); + XML_PARSE_XINCLUDE + | XML_PARSE_NOENT + | XML_PARSE_NONET); } if (!input->u.xmlreader.reader) return RECCTRL_EXTRACT_ERROR_GENERIC; @@ -1157,12 +1161,12 @@ static int extract_xml_split(struct filter_info *tinfo, if (type == XML_READER_TYPE_ELEMENT && input->u.xmlreader.split_level == depth) { + xmlNodePtr ptr; + /* per default do not ingest record */ tinfo->record_info_invoked = 0; - - xmlNodePtr ptr - = xmlTextReaderExpand(input->u.xmlreader.reader); - + + ptr = xmlTextReaderExpand(input->u.xmlreader.reader); if (ptr) { /* we have a new document */ @@ -1178,7 +1182,7 @@ static int extract_xml_split(struct filter_info *tinfo, xmlChar *buf_out = 0; int len_out = 0; xmlDocDumpMemory(doc, &buf_out, &len_out); - yaz_log(YLOG_LOG, "%s: XMLREADER depth: %i\n%.*s", + yaz_log(YLOG_LOG, "%s: XMLREADER level: %i\n%.*s", tinfo->fname ? tinfo->fname : "(none)", depth, len_out, buf_out); xmlFree(buf_out); @@ -1210,7 +1214,9 @@ static int extract_xml_full(struct filter_info *tinfo, p /* I/O handler */, 0 /* URL */, 0 /* encoding */, - XML_PARSE_XINCLUDE|XML_PARSE_NOENT); + XML_PARSE_XINCLUDE + | XML_PARSE_NOENT + | XML_PARSE_NONET); if (!doc) { return RECCTRL_EXTRACT_ERROR_GENERIC; @@ -1285,7 +1291,7 @@ static int filter_extract(void *clientData, struct recExtractCtrl *p) if (!input) return RECCTRL_EXTRACT_ERROR_GENERIC; - + odr_reset(tinfo->odr_record); switch(input->type) { @@ -1369,7 +1375,7 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) doc = xmlReadIO(ioread_ret, ioclose_ret, p /* I/O handler */, 0 /* URL */, 0 /* encoding */, - XML_PARSE_XINCLUDE|XML_PARSE_NOENT); + XML_PARSE_XINCLUDE | XML_PARSE_NOENT | XML_PARSE_NONET); if (!doc) { p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; @@ -1382,7 +1388,7 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) { p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; } - else if (p->input_format == VAL_NONE || p->input_format == VAL_TEXT_XML) + else if (!p->input_format || !oid_oidcmp(p->input_format, yaz_oid_xml())) { xmlChar *buf_out; int len_out; @@ -1392,13 +1398,13 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) else xmlDocDumpMemory(doc, &buf_out, &len_out); - p->output_format = VAL_TEXT_XML; + p->output_format = yaz_oid_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) + else if (!oid_oidcmp(p->output_format, yaz_oid_sutrs())) { xmlChar *buf_out; int len_out; @@ -1408,7 +1414,7 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) else xmlDocDumpMemory(doc, &buf_out, &len_out); - p->output_format = VAL_SUTRS; + p->output_format = yaz_oid_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);