enum conf_metadata_type {
Metadata_type_generic, // Generic text field
- Metadata_type_number, // A number
Metadata_type_year, // A number
Metadata_type_date // A number
};
if (!strcmp(merge_key, (*p)->record->merge_key))
{
struct record_cluster *existing = (*p)->record;
+ struct record *re = existing->records;
+
+ for (; re; re = re->next)
+ {
+ if (record_compare(record, re, service))
+ {
+ yaz_mutex_leave(l->mutex);
+ return 0;
+ }
+ }
record->next = existing->records;
existing->records = record;
cluster = existing;
}
-struct record_metadata * record_metadata_insert(NMEM nmem,
- struct record_metadata ** rmd,
- union data_types data)
+int record_compare(struct record *r1, struct record *r2,
+ struct conf_service *service)
{
- struct record_metadata * tmp_rmd = 0;
- // assert(nmem);
-
- if(!rmd)
- return 0;
-
- // construct new record_metadata
- tmp_rmd = nmem_malloc(nmem, sizeof(struct record_metadata));
- tmp_rmd->data = data;
-
-
- // insert in *rmd's place, moving *rmd one down the list
- tmp_rmd->next = *rmd;
- *rmd = tmp_rmd;
-
- return *rmd;
-}
-
-struct record_metadata * record_add_metadata_field_id(NMEM nmem,
- struct record * record,
- int field_id,
- union data_types data)
-{
- if (field_id < 0 || !record || !record->metadata)
- return 0;
-
- return record_metadata_insert(nmem, &(record->metadata[field_id]), data);
-}
-
-
-struct record_metadata * record_add_metadata(NMEM nmem,
- struct record * record,
- struct conf_service * service,
- const char * name,
- union data_types data)
-{
- int field_id = 0;
-
- if (!record || !record->metadata || !service || !name)
- return 0;
-
- field_id = conf_service_metadata_field_id(service, name);
-
- if (-1 == field_id)
- return 0;
-
- return record_metadata_insert(nmem, &(record->metadata[field_id]), data);
-}
-
-
-
-
-
-
-union data_types * record_assign_sortkey_field_id(NMEM nmem,
- struct record * record,
- int field_id,
- union data_types data)
-{
- if (field_id < 0 || !record || !record->sortkeys)
- return 0;
-
- return data_types_assign(nmem, &(record->sortkeys[field_id]), data);
-}
-
-
-
-union data_types * record_assign_sortkey(NMEM nmem,
- struct record * record,
- struct conf_service * service,
- const char * name,
- union data_types data)
-{
- int field_id = 0;
-
- if (!record || !service || !name)
- return 0;
-
- field_id = conf_service_sortkey_field_id(service, name);
-
- if (!(-1 < field_id) || !(field_id < service->num_sortkeys))
- return 0;
-
- return record_assign_sortkey_field_id(nmem, record, field_id, data);
+ int i;
+ for (i = 0; i < service->num_metadata; i++)
+ {
+ struct conf_metadata *ser_md = &service->metadata[i];
+ enum conf_metadata_type type = ser_md->type;
+
+ struct record_metadata *m1 = r1->metadata[i];
+ struct record_metadata *m2 = r2->metadata[i];
+ while (m1 && m2)
+ {
+ switch (type)
+ {
+ case Metadata_type_generic:
+ if (strcmp(m1->data.text.disp, m2->data.text.disp))
+ return 0;
+ break;
+ case Metadata_type_date:
+ case Metadata_type_year:
+ if (m1->data.number.min != m2->data.number.min ||
+ m1->data.number.max != m2->data.number.max)
+ return 0;
+ break;
+ }
+ m1 = m1->next;
+ m2 = m2->next;
+ }
+ if (m1 || m2)
+ return 0;
+ }
+ return 1;
}
-
-
/*
* Local variables:
* c-basic-offset: 4
struct record_metadata * record_metadata_create(NMEM nmem);
-struct record_metadata * record_metadata_insert(NMEM nmem,
- struct record_metadata ** rmd,
- union data_types data);
-
-
-struct record_metadata * record_add_metadata_field_id(NMEM nmem,
- struct record * record,
- int field_id,
- union data_types data);
-
-
-struct record_metadata * record_add_metadata(NMEM nmem,
- struct record * record,
- struct conf_service * service,
- const char * name,
- union data_types data);
-
-
-union data_types * record_assign_sortkey_field_id(NMEM nmem,
- struct record * record,
- int field_id,
- union data_types data);
-
-
-union data_types * record_assign_sortkey(NMEM nmem,
- struct record * record,
- struct conf_service * service,
- const char * name,
- union data_types data);
-
-
-
+int record_compare(struct record *r1, struct record *r2, struct conf_service *service);
struct record_cluster
{
struct record *records;
};
-
-
-
#endif // RECORD_H
/*
struct client_list *next;
};
+struct session_sorted_results {
+ const char *field;
+ int increasing;
+ struct session_sorted_results *next;
+};
+
/* session counting (1) , disable client counting (0) */
static YAZ_MUTEX g_session_mutex = 0;
static int no_sessions = 0;
void search_sort(struct session *se, const char *field, int increasing)
{
+ struct session_sorted_results *sr;
struct client_list *l;
- struct timeval tval;
session_enter(se);
+
+ /* see if we already have sorted for this critieria */
+ for (sr = se->sorted_results; sr; sr = sr->next)
+ {
+ if (!strcmp(field, sr->field) && increasing == sr->increasing)
+ break;
+ }
+ if (sr)
+ {
+ yaz_log(YLOG_LOG, "search_sort: field=%s increasing=%d already fetched",
+ field, increasing);
+ session_leave(se);
+ return;
+ }
+ yaz_log(YLOG_LOG, "search_sort: field=%s increasing=%d must fetch",
+ field, increasing);
+ sr = nmem_malloc(se->nmem, sizeof(*sr));
+ sr->field = nmem_strdup(se->nmem, field);
+ sr->increasing = increasing;
+ sr->next = se->sorted_results;
+ se->sorted_results = sr;
+
for (l = se->clients; l; l = l->next)
{
struct client *cl = l->client;
break;
}
}
-
+
if (strategy_plus_sort)
{
+ struct timeval tval;
if (client_prep_connection(cl, se->service->z3950_operation_timeout,
se->service->z3950_session_timeout,
se->service->server->iochan_man,
char **array;
int num;
nmem_strsplit(se->nmem, ":", strategy_plus_sort, &array, &num);
-
+
if (num == 2)
{
const char *sort_spec = array[1];
nmem_reset(se->nmem);
se->total_records = se->total_merged = 0;
se->num_termlists = 0;
+
+ /* reset list of sorted results and clear to relevance search */
+ se->sorted_results = nmem_malloc(se->nmem, sizeof(*se->sorted_results));
+ se->sorted_results->field = nmem_strdup(se->nmem, "relevance");
+ se->sorted_results->increasing = 0;
+ se->sorted_results->next = 0;
+
live_channels = select_targets(se, filter);
if (!live_channels)
{
}
session_enter(se);
if (client_get_session(cl) == se)
- ret = ingest_to_cluster(cl, xdoc, root, record_no, mergekey_norm);
+ ingest_to_cluster(cl, xdoc, root, record_no, mergekey_norm);
session_leave(se);
xmlFreeDoc(xdoc);
xmlNode *n;
xmlChar *type = 0;
xmlChar *value = 0;
- struct session_database *sdb = client_get_database(cl);
struct session *se = client_get_session(cl);
struct conf_service *service = se->service;
struct record *record = record_create(se->nmem,
service->num_metadata,
service->num_sortkeys, cl,
record_no);
+
+ for (n = root->children; n; n = n->next)
+ {
+ if (type)
+ xmlFree(type);
+ if (value)
+ xmlFree(value);
+ type = value = 0;
+
+ if (n->type != XML_ELEMENT_NODE)
+ continue;
+ if (!strcmp((const char *) n->name, "metadata"))
+ {
+ struct conf_metadata *ser_md = 0;
+ struct record_metadata **wheretoput = 0;
+ struct record_metadata *rec_md = 0;
+ int md_field_id = -1;
+
+ type = xmlGetProp(n, (xmlChar *) "type");
+ value = xmlNodeListGetString(xdoc, n->children, 1);
+
+ if (!type || !value || !*value)
+ continue;
+
+ md_field_id
+ = conf_service_metadata_field_id(service, (const char *) type);
+ if (md_field_id < 0)
+ {
+ if (se->number_of_warnings_unknown_metadata == 0)
+ {
+ session_log(se, YLOG_WARN,
+ "Ignoring unknown metadata element: %s", type);
+ }
+ se->number_of_warnings_unknown_metadata++;
+ continue;
+ }
+
+ ser_md = &service->metadata[md_field_id];
+
+ // non-merged metadata
+ rec_md = record_metadata_init(se->nmem, (const char *) value,
+ ser_md->type, n->properties);
+ if (!rec_md)
+ {
+ session_log(se, YLOG_WARN, "bad metadata data '%s' "
+ "for element '%s'", value, type);
+ continue;
+ }
+ wheretoput = &record->metadata[md_field_id];
+ while (*wheretoput)
+ wheretoput = &(*wheretoput)->next;
+ *wheretoput = rec_md;
+ }
+ }
+
struct record_cluster *cluster = reclist_insert(se->reclist,
service,
record,
mergekey_norm,
&se->total_merged);
+ if (!cluster)
+ return -1;
- const char *use_term_factor_str = session_setting_oneval(sdb, PZ_TERMLIST_TERM_FACTOR);
- int use_term_factor = 0;
- int term_factor = 1;
- if (use_term_factor_str && use_term_factor_str[0] != 0)
- use_term_factor = atoi(use_term_factor_str);
- if (use_term_factor) {
+ struct session_database *sdb = client_get_database(cl);
+ int term_factor = 1;
+ const char *use_term_factor_str =
+ session_setting_oneval(sdb, PZ_TERMLIST_TERM_FACTOR);
+ if (use_term_factor_str && use_term_factor_str[0] == '1')
+ {
int maxrecs = client_get_maxrecs(cl);
int hits = (int) client_get_hits(cl);
term_factor = MAX(hits, maxrecs) / MAX(1, maxrecs);
yaz_log(YLOG_DEBUG, "Using term factor: %d (%d / %d)", term_factor, MAX(hits, maxrecs), MAX(1, maxrecs));
}
- if (!cluster)
- return -1;
if (global_parameters.dump_records)
session_log(se, YLOG_LOG, "Cluster id %s from %s (#%d)", cluster->recid,
sdb->database->id, record_no);
+
+
relevance_newrec(se->relevance, cluster);
// now parsing XML record and adding data to cluster or record metadata
md_field_id
= conf_service_metadata_field_id(service, (const char *) type);
if (md_field_id < 0)
- {
- if (se->number_of_warnings_unknown_metadata == 0)
- {
- session_log(se, YLOG_WARN,
- "Ignoring unknown metadata element: %s", type);
- }
- se->number_of_warnings_unknown_metadata++;
continue;
- }
ser_md = &service->metadata[md_field_id];
- if (ser_md->sortkey_offset >= 0){
+ if (ser_md->sortkey_offset >= 0)
+ {
sk_field_id = ser_md->sortkey_offset;
ser_sk = &service->sortkeys[sk_field_id];
}
- // non-merged metadata
- rec_md = record_metadata_init(se->nmem, (const char *) value,
- ser_md->type, n->properties);
- if (!rec_md)
- {
- session_log(se, YLOG_WARN, "bad metadata data '%s' "
- "for element '%s'", value, type);
- continue;
- }
- wheretoput = &record->metadata[md_field_id];
- while (*wheretoput)
- wheretoput = &(*wheretoput)->next;
- *wheretoput = rec_md;
-
// merged metadata
rec_md = record_metadata_init(se->nmem, (const char *) value,
ser_md->type, 0);
normalize_cache_t normalize_cache;
YAZ_MUTEX session_mutex;
unsigned session_id;
+ struct session_sorted_results *sorted_results;
};
struct statistics {
http://localhost:9763/search.pz2?command=init&clear=1
http://localhost:9763/search.pz2?session=1&command=settings&pz%3Atermlist_term_count%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=10&use_url_proxy%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=0&pz%3Apiggyback%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1&pz%3Apreferred%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1&pz%3Acclmap%3Asu%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Dsubject&pz%3Asru%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=solr&use_thumbnails%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=0&pz%3Acclmap%3Adate%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Ddate&medium%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=web&pz%3Aname%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=ocs_test&pz%3Acclmap%3Aissn%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=u%3D8&pz%3Acclmap%3Ati%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Dtitle&pz%3Acclmap%3Aau%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Dauthor&pz%3Axslt%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=solr-pz2.xsl&pz%3Acclmap%3Aterm%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Dtext+s%3Dal&pz%3Acclmap%3Aisbn%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=1%3Disbn&pz%3Aqueryencoding%5Bocs-test.indexdata.com%2Fsolr%2Fselect%5D=UTF-8
http://localhost:9763/search.pz2?session=1&command=search&query=water
-1 http://localhost:9763/search.pz2?session=1&command=show&block=preferred
+2 http://localhost:9763/search.pz2?session=1&command=show&block=preferred
http://localhost:9763/search.pz2?session=1&command=termlist&name=xtargets%2Csubject%2Cauthor%2Cdate%2Cmedium
http://localhost:9763/search.pz2?session=1&command=search&query=computer
2 http://localhost:9763/search.pz2?session=1&command=show&block=1
2 http://localhost:9763/search.pz2?session=1&command=show&block=1&sort=title:1
+1 http://localhost:9763/search.pz2?session=1&command=show&block=1&sort=title:0
--- /dev/null
+<?xml version="1.0" encoding="UTF-8"?>
+<show><status>OK</status>
+<activeclients>0</activeclients>
+<merged>3</merged>
+<total>3</total>
+<start>0</start>
+<num>3</num>
+<hit>
+
+<md-title>OIL/GAS DRILLING</md-title>
+<md-description>This database contains information on oil and gas drilling such as well name, operator, driller, location, depth, copies of logs run, permits, samples (cuttings, core), completion records</md-description><location id="my" name="marcserver">
+<md-title>OIL/GAS DRILLING</md-title>
+<md-description tag="520">This database contains information on oil and gas drilling such as well name, operator, driller, location, depth, copies of logs run, permits, samples (cuttings, core), completion records</md-description>
+<md-description tag="513">1907-PRESENT</md-description></location>
+<recid>title oil gas drilling author medium book</recid>
+</hit>
+<hit>
+
+<md-title>GROUNDWATER RESOURCE MAPS - COUNTY SERIES</md-title>
+<md-description>A series of 1:250,000 scale maps showing well yield, well depth, and depth to bedrock for a large number of bedrock wells inventoried by the Maine Geological Survey in the mid-to late 1970's comprises this data set. Some series also show bedrock topography and potentiometric surface. Geographic coverage is restricted to Southern Maine</md-description><location id="my" name="marcserver">
+<md-title>GROUNDWATER RESOURCE MAPS - COUNTY SERIES</md-title>
+<md-description tag="520">A series of 1:250,000 scale maps showing well yield, well depth, and depth to bedrock for a large number of bedrock wells inventoried by the Maine Geological Survey in the mid-to late 1970's comprises this data set. Some series also show bedrock topography and potentiometric surface. Geographic coverage is restricted to Southern Maine</md-description>
+<md-description tag="513">1972-1978</md-description></location>
+<recid>title groundwater resource maps county series author medium book</recid>
+</hit>
+<hit>
+
+<md-title>BIBLIOGRAPHY OF MAINE GEOLOGY</md-title>
+<md-description>This data base is a computer based bibliography of marine geology. It allows searching by topic and geographic location, similar to GEOREF. It is currently under development to replace the printed Bibliography of Marine Geology</md-description><location id="my" name="marcserver">
+<md-title>BIBLIOGRAPHY OF MAINE GEOLOGY</md-title>
+<md-description tag="520">This data base is a computer based bibliography of marine geology. It allows searching by topic and geographic location, similar to GEOREF. It is currently under development to replace the printed Bibliography of Marine Geology</md-description>
+<md-description tag="513">1692-PRESENT</md-description></location>
+<recid>title bibliography of maine geology author medium book</recid>
+</hit>
+</show>
\ No newline at end of file
projects / pazpar2-moved-to-github.git / commitdiff