X-Git-Url: http://git.indexdata.com/?p=yaz-moved-to-github.git;a=blobdiff_plain;f=src%2Fsolr.c;h=0fd1480ebd9ca754f3b8450dac4c09ed0a133b59;hp=807326e7e172775232addf7f34603179263e2f77;hb=34356a0306446525cc3b2a5ddc268d77e5ade288;hpb=e53c4f335802aa7e8d617b4dafc0908bc50cabce diff --git a/src/solr.c b/src/solr.c index 807326e..0fd1480 100644 --- a/src/solr.c +++ b/src/solr.c @@ -1,11 +1,14 @@ /* This file is part of the YAZ toolkit. - * Copyright (C) 1995-2010 Index Data + * Copyright (C) 1995-2011 Index Data * See the file LICENSE for details. */ /** - * \file srwutil.c - * \brief Implements SRW/SRU utilities. + * \file solr.c + * \brief Implements SOAP Webservice decoding/encoding */ +#if HAVE_CONFIG_H +#include +#endif #include #include @@ -18,41 +21,36 @@ #include "sru-p.h" +#define SOLR_MAX_PARAMETERS 100 + #if YAZ_HAVE_XML2 #include #include -#define SOLR_MAX_PARAMETERS 100 - -const char *xml_node_attribute_value_get(xmlNodePtr ptr, const char *node_name, const char *attribute_name) { - - struct _xmlAttr *attr; - // check if the node name matches - if (strcmp((const char*) ptr->name, node_name)) - return 0; - // check if the attribute name and return the value - for (attr = ptr->properties; attr; attr = attr->next) - if (attr->children && attr->children->type == XML_TEXT_NODE) { - if (!strcmp((const char *) attr->name, attribute_name)) - return (const char *) attr->children->content; - } - return 0; -} - - -static int match_xml_node_attribute(xmlNodePtr ptr, const char *node_name, const char *attribute_name, const char *value) +static int match_xml_node_attribute( + xmlNodePtr ptr, + const char *node_name, const char *attribute_name, const char *value) { const char *attribute_value; // check if the node name matches if (strcmp((const char*) ptr->name, node_name)) return 0; - attribute_value = xml_node_attribute_value_get(ptr, node_name, attribute_name); - if (attribute_value && !strcmp(attribute_value, value)) + if (attribute_name) + { + attribute_value = yaz_element_attribute_value_get(ptr, node_name, + attribute_name); + if (attribute_value && !strcmp(attribute_value, value)) + return 1; + } + else /* No attribute to check */ return 1; return 0; } -static void yaz_solr_decode_result_docs(ODR o, xmlNodePtr ptr, Odr_int start, Z_SRW_searchRetrieveResponse *sr) { +static void yaz_solr_decode_result_docs(ODR o, xmlNodePtr ptr, + Odr_int start, + Z_SRW_searchRetrieveResponse *sr) +{ xmlNodePtr node; int offset = 0; int i = 0; @@ -94,16 +92,21 @@ static void yaz_solr_decode_result_docs(ODR o, xmlNodePtr ptr, Odr_int start, Z_ } } -static int yaz_solr_decode_result(ODR o, xmlNodePtr ptr, Z_SRW_searchRetrieveResponse *sr) { +static int yaz_solr_decode_result(ODR o, xmlNodePtr ptr, + Z_SRW_searchRetrieveResponse *sr) +{ Odr_int start = 0; struct _xmlAttr *attr; for (attr = ptr->properties; attr; attr = attr->next) - if (attr->children && attr->children->type == XML_TEXT_NODE) { - if (!strcmp((const char *) attr->name, "numFound")) { + if (attr->children && attr->children->type == XML_TEXT_NODE) + { + if (!strcmp((const char *) attr->name, "numFound")) + { sr->numberOfRecords = odr_intdup(o, odr_atoi( (const char *) attr->children->content)); } - else if (!strcmp((const char *) attr->name, "start")) { + else if (!strcmp((const char *) attr->name, "start")) + { start = odr_atoi((const char *) attr->children->content); } } @@ -114,76 +117,55 @@ static int yaz_solr_decode_result(ODR o, xmlNodePtr ptr, Z_SRW_searchRetrieveRe return -1; } -static Z_AttributeList *yaz_solr_use_atttribute_create(ODR o, const char *name) { - Z_AttributeList *attributes= (Z_AttributeList *) odr_malloc(o, sizeof(*attributes)); - Z_AttributeElement ** elements; - attributes->num_attributes = 1; - /* TODO check on name instead - if (!attributes->num_attributes) { - attributes->attributes = (Z_AttributeElement**)odr_nullval(); - return attributes; - } - */ - elements = (Z_AttributeElement**) odr_malloc (o, attributes->num_attributes * sizeof(*elements)); - elements[0] = (Z_AttributeElement*)odr_malloc(o,sizeof(**elements)); - elements[0]->attributeType = odr_malloc(o, sizeof(*elements[0]->attributeType)); - *elements[0]->attributeType = 1; - elements[0]->attributeSet = odr_nullval(); - elements[0]->which = Z_AttributeValue_complex; - elements[0]->value.complex = (Z_ComplexAttribute *) odr_malloc(o, sizeof(Z_ComplexAttribute)); - elements[0]->value.complex->num_list = 1; - elements[0]->value.complex->list = (Z_StringOrNumeric **) odr_malloc(o, 1 * sizeof(Z_StringOrNumeric *)); - elements[0]->value.complex->list[0] = (Z_StringOrNumeric *) odr_malloc(o, sizeof(Z_StringOrNumeric)); - elements[0]->value.complex->list[0]->which = Z_StringOrNumeric_string; - elements[0]->value.complex->list[0]->u.string = (Z_InternationalString *) odr_strdup(o, name); - elements[0]->value.complex->semanticAction = 0; - elements[0]->value.complex->num_semanticAction = 0; - attributes->attributes = elements; - return attributes; -} - - -static const char *get_facet_term_count(xmlNodePtr node, int *freq) { - - const char *term = xml_node_attribute_value_get(node, "int", "name"); +static const char *get_facet_term_count(xmlNodePtr node, int *freq) +{ + const char *term = yaz_element_attribute_value_get(node, "int", "name"); xmlNodePtr child; WRBUF wrbuf = wrbuf_alloc(); if (!term) return term; - for (child = node->children; child ; child = child->next) { + for (child = node->children; child ; child = child->next) + { if (child->type == XML_TEXT_NODE) - wrbuf_puts(wrbuf, (const char *) child->content); + wrbuf_puts(wrbuf, (const char *) child->content); } *freq = atoi(wrbuf_cstr(wrbuf)); wrbuf_destroy(wrbuf); return term; } -Z_FacetField *yaz_solr_decode_facet_field(ODR o, xmlNodePtr ptr, Z_SRW_searchRetrieveResponse *sr) +Z_FacetField *yaz_solr_decode_facet_field(ODR o, xmlNodePtr ptr, + Z_SRW_searchRetrieveResponse *sr) + { - // USE attribute - const char* name = xml_node_attribute_value_get(ptr, "lst", "name"); - Z_AttributeList *list = yaz_solr_use_atttribute_create(o, name); + Z_AttributeList *list; Z_FacetField *facet_field; int num_terms = 0; int index = 0; xmlNodePtr node; - for (node = ptr->children; node; node = node->next) { + // USE attribute + const char* name = yaz_element_attribute_value_get(ptr, "lst", "name"); + list = yaz_use_attribute_create(o, name); + for (node = ptr->children; node; node = node->next) num_terms++; - } facet_field = facet_field_create(o, list, num_terms); index = 0; - for (node = ptr->children; node; node = node->next) { + for (node = ptr->children; node; node = node->next) + { int count = 0; const char *term = get_facet_term_count(node, &count); - facet_field_term_set(o, facet_field, facet_term_create(o, term_create(o, term), count), index); + facet_field_term_set(o, facet_field, + facet_term_create(o, term_create(o, term), count), + index); index++; } return facet_field; } -static int yaz_solr_decode_facet_counts(ODR o, xmlNodePtr root, Z_SRW_searchRetrieveResponse *sr) { +static int yaz_solr_decode_facet_counts(ODR o, xmlNodePtr root, + Z_SRW_searchRetrieveResponse *sr) +{ xmlNodePtr ptr; for (ptr = root->children; ptr; ptr = ptr->next) { @@ -200,7 +182,9 @@ static int yaz_solr_decode_facet_counts(ODR o, xmlNodePtr root, Z_SRW_searchRetr num_facets = 0; for (node = ptr->children; node; node= node->next) { - facet_list_field_set(o, facet_list, yaz_solr_decode_facet_field(o, node, sr), num_facets); + facet_list_field_set(o, facet_list, + yaz_solr_decode_facet_field(o, node, sr), + num_facets); num_facets++; } sr->facetList = facet_list; @@ -210,10 +194,6 @@ static int yaz_solr_decode_facet_counts(ODR o, xmlNodePtr root, Z_SRW_searchRetr return 0; } -static void yaz_solr_decode_facets(ODR o, xmlNodePtr ptr, Z_SRW_searchRetrieveResponse *sr) { - if (match_xml_node_attribute(ptr, "lst", "name", "facet_counts")) - yaz_solr_decode_facet_counts(o, ptr->children, sr); -} #endif int yaz_solr_decode_response(ODR o, Z_HTTP_Response *hres, Z_SRW_PDU **pdup) @@ -252,7 +232,11 @@ int yaz_solr_decode_response(ODR o, Z_HTTP_Response *hres, Z_SRW_PDU **pdup) if (ptr->type == XML_ELEMENT_NODE && !strcmp((const char *) ptr->name, "result")) rc_result = yaz_solr_decode_result(o, ptr, sr); - if (match_xml_node_attribute(ptr, "lst", "name", "facet_counts")) + /* TODO The check on hits is a work-around to avoid garbled facets on zero results from the SOLR server. + * The work-around works because the results is before the facets in the xml. */ + if (rc_result == 0 && *sr->numberOfRecords > 0 && + match_xml_node_attribute(ptr, "lst", "name", + "facet_counts")) rc_facets = yaz_solr_decode_facet_counts(o, ptr, sr); } ret = rc_result + rc_facets; @@ -268,31 +252,59 @@ int yaz_solr_decode_response(ODR o, Z_HTTP_Response *hres, Z_SRW_PDU **pdup) #endif } -static void yaz_solr_encode_facet_field(ODR encode, char **name, char **value, int *i, Z_FacetField *facet_field, int *limit) { - Z_AttributeList *attribute_list = facet_field->attributes; - struct yaz_facet_attr attr_values; - yaz_facet_attr_init(&attr_values); - yaz_facet_attr_get_z_attributes(attribute_list, &attr_values); - // TODO do we want to support server decided - if (!attr_values.errcode && attr_values.useattr) { - yaz_add_name_value_str(encode, name, value, i, "facet.field", (char *) attr_values.useattr); - // TODO max(attr_values, *limit); - if (attr_values.limit > 0 && attr_values.limit > *limit) { - *limit = attr_values.limit; - } - } -} +static int yaz_solr_encode_facet_field( + ODR encode, char **name, char **value, int *i, + Z_FacetField *facet_field) +{ + Z_AttributeList *attribute_list = facet_field->attributes; + struct yaz_facet_attr attr_values; + yaz_facet_attr_init(&attr_values); + yaz_facet_attr_get_z_attributes(attribute_list, &attr_values); + // TODO do we want to support server decided -static void yaz_solr_encode_facet_list(ODR encode, char **name, char **value, int *i, Z_FacetList *facet_list, int *limit) { + if (attr_values.errcode) + return -1; + if (attr_values.useattr) + { + WRBUF wrbuf = wrbuf_alloc(); + wrbuf_puts(wrbuf, (char *) attr_values.useattr); + yaz_add_name_value_str(encode, name, value, i, + "facet.field", + odr_strdup(encode, wrbuf_cstr(wrbuf))); + if (attr_values.limit > 0) + { + WRBUF wrbuf2 = wrbuf_alloc(); + Odr_int olimit; + wrbuf_puts(wrbuf2, "f."); + wrbuf_puts(wrbuf2, wrbuf_cstr(wrbuf)); + wrbuf_puts(wrbuf2, ".facet.limit"); + olimit = attr_values.limit; + yaz_add_name_value_int(encode, name, value, i, + odr_strdup(encode, wrbuf_cstr(wrbuf2)), + &olimit); + wrbuf_destroy(wrbuf2); + } + wrbuf_destroy(wrbuf); + } + return 0; +} +static int yaz_solr_encode_facet_list( + ODR encode, char **name, char **value, + int *i, Z_FacetList *facet_list) +{ int index; - for (index = 0; index < facet_list->num; index++) { - yaz_solr_encode_facet_field(encode, name, value, i, facet_list->elements[index], limit); - + for (index = 0; index < facet_list->num; index++) + { + int r = yaz_solr_encode_facet_field(encode, name, value, i, + facet_list->elements[index]); + if (r) + return -1; + } + return 0; } - int yaz_solr_encode_request(Z_HTTP_Request *hreq, Z_SRW_PDU *srw_pdu, ODR encode, const char *charset) { @@ -302,16 +314,14 @@ int yaz_solr_encode_request(Z_HTTP_Request *hreq, Z_SRW_PDU *srw_pdu, char *uri_args; char *path; int i = 0; - + z_HTTP_header_add_basic_auth(encode, &hreq->headers, srw_pdu->username, srw_pdu->password); - - switch (srw_pdu->which) + if (srw_pdu->which == Z_SRW_searchRetrieve_request) { - case Z_SRW_searchRetrieve_request: { Z_SRW_searchRetrieveRequest *request = srw_pdu->u.request; solr_op = "select"; - switch(srw_pdu->u.request->query_type) + switch (srw_pdu->u.request->query_type) { case Z_SRW_query_type_pqf: yaz_add_name_value_str(encode, name, value, &i, @@ -335,21 +345,17 @@ int yaz_solr_encode_request(Z_HTTP_Request *hreq, Z_SRW_PDU *srw_pdu, yaz_add_name_value_str(encode, name, value, &i, "fl", request->recordSchema); - if (request->facetList) { + if (request->facetList) + { Z_FacetList *facet_list = request->facetList; - int limit = 0; - Odr_int olimit; yaz_add_name_value_str(encode, name, value, &i, "facet", "true"); - yaz_solr_encode_facet_list(encode, name, value, &i, facet_list, &limit); - olimit = limit; - yaz_add_name_value_int(encode, name, value, &i, "facet.limit", &olimit); - + yaz_add_name_value_str(encode, name, value, &i, "facet.mincount", "1"); + if (yaz_solr_encode_facet_list(encode, name, value, &i, facet_list)) + return -1; } - break; } - default: + else return -1; - } name[i] = 0; yaz_array_to_uri(&uri_args, encode, name, value);