X-Git-Url: http://lists.indexdata.com/cgi-bin?a=blobdiff_plain;f=src%2Fpazpar2.c;h=31cf43157f6c0e8fc04e35156039c6ad44a36eab;hb=235995289021a11b33312c91a55ccfbca605348b;hp=3a696c60f02cdb45e3989581629971c41d7e1d5d;hpb=aabc654ef7b6c5a6740e95b9580223726a86e43a;p=pazpar2-moved-to-github.git diff --git a/src/pazpar2.c b/src/pazpar2.c index 3a696c6..31cf431 100644 --- a/src/pazpar2.c +++ b/src/pazpar2.c @@ -1,4 +1,4 @@ -/* $Id: pazpar2.c,v 1.44 2007-01-18 18:11:19 quinn Exp $ */ +/* $Id: pazpar2.c,v 1.56 2007-03-28 12:05:18 marc Exp $ */ #include #include @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -38,6 +39,8 @@ #include "reclists.h" #include "relevance.h" #include "config.h" +#include "database.h" +#include "settings.h" #define MAX_CHUNK 15 @@ -45,7 +48,7 @@ static void client_fatal(struct client *cl); static void connection_destroy(struct connection *co); static int client_prep_connection(struct client *cl); static void ingest_records(struct client *cl, Z_Records *r); -static struct conf_retrievalprofile *database_retrieval_profile(struct database *db); +//static struct conf_retrievalprofile *database_retrieval_profile(struct database *db); void session_alert_watch(struct session *s, int what); IOCHAN channel_list = 0; // Master list of connections we're handling events to @@ -53,9 +56,6 @@ IOCHAN channel_list = 0; // Master list of connections we're handling events to static struct connection *connection_freelist = 0; static struct client *client_freelist = 0; -static struct host *hosts = 0; // The hosts we know about -static struct database *databases = 0; // The databases we know about - static char *client_states[] = { "Client_Connecting", "Client_Connected", @@ -74,6 +74,8 @@ struct parameters global_parameters = { "", "", + "", + "", 0, 0, 30, @@ -121,6 +123,7 @@ static int send_apdu(struct client *c, Z_APDU *a) static void send_init(IOCHAN i) { + struct connection *co = iochan_getdata(i); struct client *cl = co->client; Z_APDU *a = zget_APDU(global_parameters.odr_out, Z_APDU_initRequest); @@ -136,6 +139,18 @@ static void send_init(IOCHAN i) ODR_MASK_SET(a->u.initRequest->protocolVersion, Z_ProtocolVersion_1); ODR_MASK_SET(a->u.initRequest->protocolVersion, Z_ProtocolVersion_2); ODR_MASK_SET(a->u.initRequest->protocolVersion, Z_ProtocolVersion_3); + + + /* add virtual host if tunneling through Z39.50 proxy */ + + if (0 < strlen(global_parameters.zproxy_override) + && 0 < strlen(cl->database->url)) + yaz_oi_set_string_oidval(&a->u.initRequest->otherInfo, + global_parameters.odr_out, VAL_PROXY, + 1, cl->database->url); + + + if (send_apdu(cl, a) >= 0) { iochan_setflags(i, EVENT_INPUT); @@ -176,7 +191,7 @@ static void send_search(IOCHAN i) for (ndb = 0; db->databases[ndb]; ndb++) databaselist[ndb] = db->databases[ndb]; - a->u.presentRequest->preferredRecordSyntax = + a->u.searchRequest->preferredRecordSyntax = yaz_oidval_to_z3950oid(global_parameters.odr_out, CLASS_RECSYN, VAL_USMARC); a->u.searchRequest->smallSetUpperBound = &ssub; @@ -385,8 +400,6 @@ static void add_facet(struct session *s, const char *type, const char *value) termlist_insert(s->termlists[i].termlist, value); } -int yaz_marc_write_xml(); - static xmlDoc *normalize_record(struct client *cl, Z_External *rec) { struct conf_retrievalprofile *rprofile = cl->database->rprofile; @@ -417,7 +430,7 @@ static xmlDoc *normalize_record(struct client *cl, Z_External *rec) yaz_log(YLOG_WARN, "Failed to encode as XML"); return 0; } - rdoc = xmlNewDoc("1.0"); + rdoc = xmlNewDoc((xmlChar *) "1.0"); xmlDocSetRootElement(rdoc, res); } else @@ -509,7 +522,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) return 0; root = xmlDocGetRootElement(xdoc); - if (!(mergekey = xmlGetProp(root, "mergekey"))) + if (!(mergekey = xmlGetProp(root, (xmlChar *) "mergekey"))) { yaz_log(YLOG_WARN, "No mergekey found in record"); xmlFreeDoc(xdoc); @@ -523,11 +536,12 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) sizeof(struct record_metadata*) * service->num_metadata); memset(res->metadata, 0, sizeof(struct record_metadata*) * service->num_metadata); - mergekey_norm = nmem_strdup(se->nmem, (char*) mergekey); + mergekey_norm = (xmlChar *) nmem_strdup(se->nmem, (char*) mergekey); xmlFree(mergekey); - normalize_mergekey(mergekey_norm, 0); + normalize_mergekey((char *) mergekey_norm, 0); - cluster = reclist_insert(se->reclist, res, mergekey_norm, &se->total_merged); + cluster = reclist_insert(se->reclist, res, (char *) mergekey_norm, + &se->total_merged); if (global_parameters.dump_records) yaz_log(YLOG_LOG, "Cluster id %d from %s (#%d)", cluster->recid, cl->database->url, cl->records); @@ -549,7 +563,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) if (n->type != XML_ELEMENT_NODE) continue; - if (!strcmp(n->name, "metadata")) + if (!strcmp((const char *) n->name, "metadata")) { struct conf_metadata *md = 0; struct conf_sortkey *sk = 0; @@ -557,7 +571,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) int imeta; int first, last; - type = xmlGetProp(n, "type"); + type = xmlGetProp(n, (xmlChar *) "type"); value = xmlNodeListGetString(xdoc, n->children, 0); if (!type || !value) @@ -565,7 +579,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) // First, find out what field we're looking at for (imeta = 0; imeta < service->num_metadata; imeta++) - if (!strcmp(type, service->metadata[imeta].name)) + if (!strcmp((const char *) type, service->metadata[imeta].name)) { md = &service->metadata[imeta]; if (md->sortkey_offset >= 0) @@ -590,7 +604,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) if (md->type == Metadata_type_generic) { char *p, *pe; - for (p = value; *p && isspace(*p); p++) + for (p = (char *) value; *p && isspace(*p); p++) ; for (pe = p + strlen(p) - 1; pe > p && strchr(" ,/.:([", *pe); pe--) @@ -600,7 +614,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) } else if (md->type == Metadata_type_year) { - if (extract_years(value, &first, &last) < 0) + if (extract_years((char *) value, &first, &last) < 0) continue; } else @@ -617,7 +631,7 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) { struct record_metadata *mnode; for (mnode = *wheretoput; mnode; mnode = mnode->next) - if (!strcmp(mnode->data.text, newm->data.text)) + if (!strcmp((const char *) mnode->data.text, newm->data.text)) break; if (!mnode) { @@ -678,22 +692,23 @@ static struct record *ingest_record(struct client *cl, Z_External *rec) yaz_log(YLOG_WARN, "Don't know how to merge on element name %s", md->name); if (md->rank) - relevance_countwords(se->relevance, cluster, value, md->rank); + relevance_countwords(se->relevance, cluster, + (char *) value, md->rank); if (md->termlist) { if (md->type == Metadata_type_year) { char year[64]; sprintf(year, "%d", last); - add_facet(se, type, year); + add_facet(se, (char *) type, year); if (first != last) { sprintf(year, "%d", first); - add_facet(se, type, year); + add_facet(se, (char *) type, year); } } else - add_facet(se, type, value); + add_facet(se, (char *) type, (char *) value); } xmlFree(type); xmlFree(value); @@ -946,18 +961,40 @@ static struct connection *connection_create(struct client *cl) int res; void *addr; - yaz_log(YLOG_DEBUG, "Connection create %s", cl->database->url); + if (!(link = cs_create(tcpip_type, 0, PROTO_Z3950))) - { - yaz_log(YLOG_FATAL|YLOG_ERRNO, "Failed to create comstack"); - exit(1); - } + { + yaz_log(YLOG_FATAL|YLOG_ERRNO, "Failed to create comstack"); + exit(1); + } + + if (0 == strlen(global_parameters.zproxy_override)){ + /* no Z39.50 proxy needed - direct connect */ + yaz_log(YLOG_DEBUG, "Connection create %s", cl->database->url); + + if (!(addr = cs_straddr(link, cl->database->host->ipport))) + { + yaz_log(YLOG_WARN|YLOG_ERRNO, + "Lookup of IP address %s failed", + cl->database->host->ipport); + return 0; + } + + } else { + /* Z39.50 proxy connect */ + yaz_log(YLOG_DEBUG, "Connection create %s proxy %s", + cl->database->url, global_parameters.zproxy_override); - if (!(addr = cs_straddr(link, cl->database->host->ipport))) - { - yaz_log(YLOG_WARN|YLOG_ERRNO, "Lookup of IP address %s failed?", - cl->database->host->ipport); - return 0; + yaz_log(YLOG_LOG, "Connection cs_create_host %s proxy %s", + cl->database->url, global_parameters.zproxy_override); + + if (!(addr = cs_straddr(link, global_parameters.zproxy_override))) + { + yaz_log(YLOG_WARN|YLOG_ERRNO, + "Lookup of IP address %s failed", + global_parameters.zproxy_override); + return 0; + } } res = cs_connect(link, addr); @@ -983,7 +1020,7 @@ static struct connection *connection_create(struct client *cl) cl->connection = new; new->link = link; - new->iochan = iochan_create(cs_fileno(link), handler, 0); + new->iochan = iochan_create(cs_fileno(link), 0, handler, 0); iochan_setdata(new->iochan, new); new->iochan->next = channel_list; channel_list = new->iochan; @@ -1044,6 +1081,8 @@ static int client_prep_connection(struct client *cl) return 0; } +#ifdef GAGA // Moved to database.c + // This function will most likely vanish when a proper target profile mechanism is // introduced. void load_simpletargets(const char *fn) @@ -1079,7 +1118,7 @@ void load_simpletargets(const char *fn) yaz_log(YLOG_LOG, "Target: %s, '%s'", url, db); for (host = hosts; host; host = host->next) - if (!strcmp(url, host->hostport)) + if (!strcmp((const char *) url, host->hostport)) break; if (!host) { @@ -1148,6 +1187,8 @@ void load_simpletargets(const char *fn) fclose(f); } +#endif + static void pull_terms(NMEM nmem, struct ccl_rpn_node *n, char **termlist, int *num) { switch (n->kind) @@ -1240,40 +1281,25 @@ void session_alert_watch(struct session *s, int what) s->watchlist[what].data = 0; } -// This needs to be extended with selection criteria -static struct conf_retrievalprofile *database_retrieval_profile(struct database *db) +//callback for grep_databases +static void select_targets_callback(void *context, struct database *db) { - if (!config) - { - yaz_log(YLOG_FATAL, "Must load configuration (-f)"); - exit(1); - } - if (!config->retrievalprofiles) - { - yaz_log(YLOG_FATAL, "No retrieval profiles defined"); - } - return config->retrievalprofiles; + struct session *se = (struct session*) context; + struct client *cl = client_create(); + cl->database = db; + cl->session = se; + cl->next = se->clients; + se->clients = cl; } // This should be extended with parameters to control selection criteria // Associates a set of clients with a session; -int select_targets(struct session *se) +int select_targets(struct session *se, struct database_criterion *crit) { - struct database *db; - int c = 0; - while (se->clients) client_destroy(se->clients); - for (db = databases; db; db = db->next) - { - struct client *cl = client_create(); - cl->database = db; - cl->session = se; - cl->next = se->clients; - se->clients = cl; - c++; - } - return c; + + return grep_databases(se, crit, select_targets_callback); } int session_active_clients(struct session *s) @@ -1291,22 +1317,62 @@ int session_active_clients(struct session *s) return res; } -char *search(struct session *se, char *query) +// parses crit1=val1,crit2=val2|val3,... +static struct database_criterion *parse_filter(NMEM m, const char *buf) +{ + struct database_criterion *res = 0; + char **values; + int num; + int i; + + if (!buf || !*buf) + return 0; + nmem_strsplit(m, ",", buf, &values, &num); + for (i = 0; i < num; i++) + { + char **subvalues; + int subnum; + int subi; + struct database_criterion *new = nmem_malloc(m, sizeof(*new)); + char *eq = strchr(values[i], '='); + if (!eq) + { + yaz_log(YLOG_WARN, "Missing equal-sign in filter"); + return 0; + } + *(eq++) = '\0'; + new->name = values[i]; + nmem_strsplit(m, "|", eq, &subvalues, &subnum); + new->values = 0; + for (subi = 0; subi < subnum; subi++) + { + struct database_criterion_value *newv = nmem_malloc(m, sizeof(*newv)); + newv->value = subvalues[subi]; + newv->next = new->values; + new->values = newv; + } + new->next = res; + res = new; + } + return res; +} + +char *search(struct session *se, char *query, char *filter) { int live_channels = 0; struct client *cl; + struct database_criterion *criteria; yaz_log(YLOG_DEBUG, "Search"); + nmem_reset(se->nmem); + criteria = parse_filter(se->nmem, filter); strcpy(se->query, query); se->requestid++; - nmem_reset(se->nmem); + // Release any existing clients + select_targets(se, criteria); for (cl = se->clients; cl; cl = cl->next) { - cl->hits = -1; - cl->records = 0; - cl->diagnostic = 0; - if (client_prep_connection(cl)) live_channels++; } @@ -1333,7 +1399,7 @@ void destroy_session(struct session *s) while (s->clients) client_destroy(s->clients); nmem_destroy(s->nmem); - wrbuf_free(s->wrbuf, 1); + wrbuf_destroy(s->wrbuf); } struct session *new_session() @@ -1359,8 +1425,6 @@ struct session *new_session() session->watchlist[i].fun = 0; } - select_targets(session); - return session; } @@ -1390,7 +1454,7 @@ struct termlist_score **termlist(struct session *s, const char *name, int *num) int i; for (i = 0; i < s->num_termlists; i++) - if (!strcmp(s->termlists[i].name, name)) + if (!strcmp((const char *) s->termlists[i].name, name)) return termlist_highscore(s->termlists[i].termlist, num); return 0; } @@ -1549,9 +1613,44 @@ static void start_proxy(void) else return; - http_set_proxyaddr(hp); + http_set_proxyaddr(hp, ser->myurl ? ser->myurl : ""); +} + +static void start_zproxy(void) +{ + struct conf_server *ser = global_parameters.server; + + if (*global_parameters.zproxy_override){ + yaz_log(YLOG_LOG, "Z39.50 proxy %s", + global_parameters.zproxy_override); + return; + } + + else if (ser->zproxy_host || ser->zproxy_port) + { + char hp[128] = ""; + + strcpy(hp, ser->zproxy_host ? ser->zproxy_host : ""); + if (ser->zproxy_port) + { + if (*hp) + strcat(hp, ":"); + else + strcat(hp, "@:"); + + sprintf(hp + strlen(hp), "%d", ser->zproxy_port); + } + strcpy(global_parameters.zproxy_override, hp); + yaz_log(YLOG_LOG, "Z39.50 proxy %s", + global_parameters.zproxy_override); + + } + else + return; } + + int main(int argc, char **argv) { int ret; @@ -1562,7 +1661,7 @@ int main(int argc, char **argv) yaz_log_init(YLOG_DEFAULT_LEVEL, "pazpar2", 0); - while ((ret = options("f:x:h:p:C:s:d", argv, argc, &arg)) != -2) + while ((ret = options("t:f:x:h:p:z:C:s:d", argv, argc, &arg)) != -2) { switch (ret) { case 'f': @@ -1578,6 +1677,12 @@ int main(int argc, char **argv) case 'p': strcpy(global_parameters.proxy_override, arg); break; + case 'z': + strcpy(global_parameters.zproxy_override, arg); + break; + case 't': + strcpy(global_parameters.settings_path, arg); + break; case 's': load_simpletargets(arg); break; @@ -1591,6 +1696,7 @@ int main(int argc, char **argv) " -C cclconfig\n" " -s simpletargetfile\n" " -p hostname[:portno] (HTTP proxy)\n" + " -z hostname[:portno] (Z39.50 proxy)\n" " -d (show internal records)\n"); exit(1); } @@ -1605,7 +1711,12 @@ int main(int argc, char **argv) start_http_listener(); start_proxy(); - global_parameters.ccl_filter = load_cclfile("../etc/default.bib"); + start_zproxy(); + + if (*global_parameters.settings_path) + settings_read(global_parameters.settings_path); + if (!global_parameters.ccl_filter) + global_parameters.ccl_filter = load_cclfile("../etc/default.bib"); global_parameters.yaz_marc = yaz_marc_create(); yaz_marc_subfield_str(global_parameters.yaz_marc, "\t"); global_parameters.odr_in = odr_createmem(ODR_DECODE);