X-Git-Url: http://lists.indexdata.com/cgi-bin?a=blobdiff_plain;f=src%2Fsession.c;h=fc021a87510c046ca378cc7968077213adce0324;hb=4d9f530309235e8c78d50f82196f135c3127c064;hp=2adb1ebe1842bc2058241b42fe1de19fa77b9f32;hpb=aeb802e28a6a6450e1bafd95efa828c0165411a4;p=pazpar2-moved-to-github.git diff --git a/src/session.c b/src/session.c index 2adb1eb..fc021a8 100644 --- a/src/session.c +++ b/src/session.c @@ -188,23 +188,51 @@ void pull_terms(NMEM nmem, struct ccl_rpn_node *n, char **termlist, int *num) void add_facet(struct session *s, const char *type, const char *value, int count) { struct conf_service *service = s->service; - pp2_relevance_token_t prt; + pp2_charset_token_t prt; const char *facet_component; WRBUF facet_wrbuf = wrbuf_alloc(); - prt = pp2_relevance_tokenize(service->facet_pct); - - pp2_relevance_first(prt, value, 0); - while ((facet_component = pp2_relevance_token_next(prt))) + WRBUF display_wrbuf = wrbuf_alloc(); + int i; + const char *icu_chain_id = 0; + + for (i = 0; i < service->num_metadata; i++) + if (!strcmp((service->metadata + i)->name, type)) + icu_chain_id = (service->metadata + i)->facetrule; + yaz_log(YLOG_LOG, "icu_chain id=%s", icu_chain_id ? icu_chain_id : "null"); + + if (!icu_chain_id) + icu_chain_id = "facet"; + prt = pp2_charset_token_create(service->charsets, icu_chain_id); + if (!prt) { + yaz_log(YLOG_FATAL, "Unknown ICU chain '%s' for facet of type '%s'", + icu_chain_id, type); + wrbuf_destroy(facet_wrbuf); + wrbuf_destroy(display_wrbuf); + return; + } + pp2_charset_token_first(prt, value, 0); + while ((facet_component = pp2_charset_token_next(prt))) + { + const char *display_component; if (*facet_component) { if (wrbuf_len(facet_wrbuf)) wrbuf_puts(facet_wrbuf, " "); wrbuf_puts(facet_wrbuf, facet_component); } + display_component = pp2_get_display(prt); + if (display_component) + { + if (wrbuf_len(display_wrbuf)) + wrbuf_puts(display_wrbuf, " "); + wrbuf_puts(display_wrbuf, display_component); + } } - pp2_relevance_token_destroy(prt); - + pp2_charset_token_destroy(prt); + + yaz_log(YLOG_LOG, "facet norm=%s", wrbuf_cstr(facet_wrbuf)); + yaz_log(YLOG_LOG, "facet display=%s", wrbuf_cstr(display_wrbuf)); if (wrbuf_len(facet_wrbuf)) { int i; @@ -217,6 +245,7 @@ void add_facet(struct session *s, const char *type, const char *value, int count { session_log(s, YLOG_FATAL, "Too many termlists"); wrbuf_destroy(facet_wrbuf); + wrbuf_destroy(display_wrbuf); return; } @@ -229,10 +258,11 @@ void add_facet(struct session *s, const char *type, const char *value, int count #if 0 session_log(s, YLOG_DEBUG, "Facets for %s: %s norm:%s (%d)", type, value, wrbuf_cstr(facet_wrbuf), count); #endif - termlist_insert(s->termlists[i].termlist, wrbuf_cstr(facet_wrbuf), - count); + termlist_insert(s->termlists[i].termlist, wrbuf_cstr(display_wrbuf), + wrbuf_cstr(facet_wrbuf), count); } wrbuf_destroy(facet_wrbuf); + wrbuf_destroy(display_wrbuf); } static xmlDoc *record_to_xml(struct session *se, @@ -1105,15 +1135,15 @@ static int get_mergekey_from_doc(xmlDoc *doc, xmlNode *root, const char *name, if (value) { const char *norm_str; - pp2_relevance_token_t prt = - pp2_relevance_tokenize(service->mergekey_pct); + pp2_charset_token_t prt = + pp2_charset_token_create(service->charsets, "mergekey"); - pp2_relevance_first(prt, (const char *) value, 0); + pp2_charset_token_first(prt, (const char *) value, 0); if (wrbuf_len(norm_wr) > 0) wrbuf_puts(norm_wr, " "); wrbuf_puts(norm_wr, name); while ((norm_str = - pp2_relevance_token_next(prt))) + pp2_charset_token_next(prt))) { if (*norm_str) { @@ -1122,7 +1152,7 @@ static int get_mergekey_from_doc(xmlDoc *doc, xmlNode *root, const char *name, } } xmlFree(value); - pp2_relevance_token_destroy(prt); + pp2_charset_token_destroy(prt); no_found++; } } @@ -1144,11 +1174,11 @@ static const char *get_mergekey(xmlDoc *doc, struct client *cl, int record_no, if (mergekey) { const char *norm_str; - pp2_relevance_token_t prt = - pp2_relevance_tokenize(service->mergekey_pct); + pp2_charset_token_t prt = + pp2_charset_token_create(service->charsets, "mergekey"); - pp2_relevance_first(prt, (const char *) mergekey, 0); - while ((norm_str = pp2_relevance_token_next(prt))) + pp2_charset_token_first(prt, (const char *) mergekey, 0); + while ((norm_str = pp2_charset_token_next(prt))) { if (*norm_str) { @@ -1157,7 +1187,7 @@ static const char *get_mergekey(xmlDoc *doc, struct client *cl, int record_no, wrbuf_puts(norm_wr, norm_str); } } - pp2_relevance_token_destroy(prt); + pp2_charset_token_destroy(prt); xmlFree(mergekey); } else @@ -1356,7 +1386,7 @@ static int ingest_to_cluster(struct client *cl, // now parsing XML record and adding data to cluster or record metadata for (n = root->children; n; n = n->next) { - pp2_relevance_token_t prt; + pp2_charset_token_t prt; if (type) xmlFree(type); if (value) @@ -1423,16 +1453,15 @@ static int ingest_to_cluster(struct client *cl, // assign cluster or record based on merge action if (ser_md->merge == Metadata_merge_unique) { - struct record_metadata *mnode; - for (mnode = *wheretoput; mnode; mnode = mnode->next) - if (!strcmp((const char *) mnode->data.text.disp, + while (*wheretoput) + { + if (!strcmp((const char *) (*wheretoput)->data.text.disp, rec_md->data.text.disp)) break; - if (!mnode) - { - rec_md->next = *wheretoput; - *wheretoput = rec_md; + wheretoput = &(*wheretoput)->next; } + if (!*wheretoput) + *wheretoput = rec_md; } else if (ser_md->merge == Metadata_merge_longest) { @@ -1452,12 +1481,13 @@ static int ingest_to_cluster(struct client *cl, nmem_malloc(se->nmem, sizeof(union data_types)); - prt = pp2_relevance_tokenize(service->sort_pct); + prt = + pp2_charset_token_create(service->charsets, "sort"); - pp2_relevance_first(prt, rec_md->data.text.disp, - skip_article); + pp2_charset_token_first(prt, rec_md->data.text.disp, + skip_article); - pp2_relevance_token_next(prt); + pp2_charset_token_next(prt); sort_str = pp2_get_sort(prt); @@ -1471,13 +1501,14 @@ static int ingest_to_cluster(struct client *cl, } cluster->sortkeys[sk_field_id]->text.sort = nmem_strdup(se->nmem, sort_str); - pp2_relevance_token_destroy(prt); + pp2_charset_token_destroy(prt); } } } else if (ser_md->merge == Metadata_merge_all) { - rec_md->next = *wheretoput; + while (*wheretoput) + wheretoput = &(*wheretoput)->next; *wheretoput = rec_md; } else if (ser_md->merge == Metadata_merge_range)