X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Frpnscan.c;h=1d6f628ca38b9e5671e554dfbb3d1e99ab24b4f8;hb=ef21c46268e6bbdd5783099a822933ac78cf1a81;hp=5fd6b3835332411d346c4076e2518d03c3b8b060;hpb=d82c0efad7971d102220a8824e1ea674db5b7fe2;p=idzebra-moved-to-github.git diff --git a/index/rpnscan.c b/index/rpnscan.c index 5fd6b38..1d6f628 100644 --- a/index/rpnscan.c +++ b/index/rpnscan.c @@ -1,4 +1,4 @@ -/* $Id: rpnscan.c,v 1.14 2007-10-29 09:25:40 adam Exp $ +/* $Id: rpnscan.c,v 1.25 2008-01-24 16:17:03 adam Exp $ Copyright (C) 1995-2007 Index Data ApS @@ -41,20 +41,32 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA #define RPN_MAX_ORDS 32 -int log_scan = YLOG_LOG; - /* convert APT SCAN term to internal cmap */ static ZEBRA_RES trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, - char *termz, int reg_type) + char *termz, zebra_map_t zm) { - char termz0[IT_MAX_WORD]; + char term_utf8[IT_MAX_WORD]; - if (zapt_term_to_utf8(zh, zapt, termz0) == ZEBRA_FAIL) + if (zapt_term_to_utf8(zh, zapt, term_utf8) == ZEBRA_FAIL) return ZEBRA_FAIL; /* error */ + else if (zebra_maps_is_icu(zm)) + { + const char *res_buf; + size_t res_len; + zebra_map_tokenize_start(zm, term_utf8, strlen(term_utf8)); + + if (zebra_map_tokenize_next(zm, &res_buf, &res_len, 0, 0)) + { + memcpy(termz, res_buf, res_len); + termz[res_len] = '\0'; + } + else + termz[0] = '\0'; + } else { const char **map; - const char *cp = (const char *) termz0; + const char *cp = (const char *) term_utf8; const char *cp_end = cp + strlen(cp); const char *src; int i = 0; @@ -63,7 +75,7 @@ static ZEBRA_RES trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, while ((len = (cp_end - cp)) > 0) { - map = zebra_maps_input(zh->reg->zebra_maps, reg_type, &cp, len, 0); + map = zebra_maps_input(zm, &cp, len, 0); if (**map == *CHR_SPACE) space_map = *map; else @@ -81,31 +93,6 @@ static ZEBRA_RES trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, return ZEBRA_OK; } -static void count_set(ZebraHandle zh, RSET rset, zint *count) -{ - zint psysno = 0; - struct it_key key; - RSFD rfd; - - yaz_log(YLOG_DEBUG, "count_set"); - - rset->hits_limit = zh->approx_limit; - - *count = 0; - rfd = rset_open(rset, RSETF_READ); - while (rset_read(rfd, &key,0 /* never mind terms */)) - { - if (key.mem[0] != psysno) - { - psysno = key.mem[0]; - if (rfd->counted_items >= rset->hits_limit) - break; - } - } - rset_close(rfd); - *count = rset->hits_count; -} - static void get_first_snippet_from_rset(ZebraHandle zh, RSET rset, zebra_snippets *snippets, zint *sysno) @@ -181,12 +168,21 @@ static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem, Z_AttributesPlusTerm *zapt, RSET limit_set, const char *term, - int index_type, + const char *index_type, struct scan2_info_entry *ar, int ord_no, ZebraScanEntry *glist, int pos) { int i; RSET rset = 0; + zint approx_limit = zh->approx_limit; + AttrType global_hits_limit_attr; + int l; + attr_init_APT(&global_hits_limit_attr, zapt, 12); + + l = attr_find(&global_hits_limit_attr, NULL); + if (l != -1) + approx_limit = l; + for (i = 0; i < ord_no; i++) { if (ar[i].isam_p && strcmp(wrbuf_cstr(ar[i].term), term) == 0) @@ -202,7 +198,7 @@ static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem, wrbuf_buf(ar[i].term), wrbuf_len(ar[i].term), NULL, 1, zapt->term->which, nmem, kc, kc->scope, ol, index_type, - 0 /* hits_limit */, + 0 /* hits_limit_value */, 0 /* term_ref_id_str */); if (!rset) rset = rset_t; @@ -231,13 +227,11 @@ static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem, rset = rset_create_and(nmem, kc, kc->scope, 2, rsets); } /* count it */ - count_set(zh, rset, &count); + zebra_count_set(zh, rset, &count, approx_limit); if (pos != -1) { zint sysno; - int code = -1; - zebra_snippets *rec_snippets = zebra_snippets_create(); zebra_snippets *hit_snippets = zebra_snippets_create(); glist[pos].term = 0; @@ -245,22 +239,38 @@ static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem, get_first_snippet_from_rset(zh, rset, hit_snippets, &sysno); if (sysno) - code = zebra_get_rec_snippets(zh, sysno, rec_snippets); - - if (code == 0) { - const struct zebra_snippet_word *w = - zebra_snippets_lookup(rec_snippets, hit_snippets); - if (w) + zebra_snippets *rec_snippets = zebra_snippets_create(); + int code = zebra_get_rec_snippets(zh, sysno, rec_snippets); + if (code == 0) { - glist[pos].display_term = odr_strdup(stream, w->term); + const struct zebra_snippet_word *w = + zebra_snippets_lookup(rec_snippets, hit_snippets); + if (w) + { + glist[pos].display_term = odr_strdup(stream, w->term); + } + else + { + yaz_log(YLOG_WARN, "zebra_snippets_lookup failed for pos=%d", pos); + } } + zebra_snippets_destroy(rec_snippets); + } + if (zebra_term_untrans_iconv(zh, stream->mem, index_type, + &glist[pos].term, term)) + { + /* failed.. use display_term instead (which could be 0) */ + glist[pos].term = glist[pos].display_term; } + if (!glist[pos].term) - zebra_term_untrans_iconv(zh, stream->mem, index_type, - &glist[pos].term, term); + { + yaz_log(YLOG_WARN, "Could not generate scan term for pos=%d", + pos); + glist[pos].term = "None"; + } glist[pos].occurrences = count; - zebra_snippets_destroy(rec_snippets); zebra_snippets_destroy(hit_snippets); } rset_delete(rset); @@ -271,17 +281,19 @@ static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem, } return 0; } - -static ZEBRA_RES rpn_scan_ver2(ZebraHandle zh, ODR stream, NMEM nmem, + +static ZEBRA_RES rpn_scan_norm(ZebraHandle zh, ODR stream, NMEM nmem, struct rset_key_control *kc, Z_AttributesPlusTerm *zapt, int *position, int *num_entries, ZebraScanEntry **list, int *is_partial, RSET limit_set, - int index_type, int ord_no, int *ords) + const char *index_type, + int ord_no, int *ords) { struct scan2_info_entry *ar = nmem_malloc(nmem, sizeof(*ar) * ord_no); struct rpn_char_map_info rcmi; + zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type); int i, dif; int after_pos; int pos = 0; @@ -297,7 +309,7 @@ static ZEBRA_RES rpn_scan_ver2(ZebraHandle zh, ODR stream, NMEM nmem, *num_entries = 0; return ZEBRA_OK; } - rpn_char_map_prepare(zh->reg, index_type, &rcmi); + rpn_char_map_prepare(zh->reg, zm, &rcmi); for (i = 0; i < ord_no; i++) ar[i].term = wrbuf_alloc(); @@ -311,7 +323,7 @@ static ZEBRA_RES rpn_scan_ver2(ZebraHandle zh, ODR stream, NMEM nmem, termz[prefix_len] = 0; strcpy(ar[i].prefix, termz); - if (trans_scan_term(zh, zapt, termz+prefix_len, index_type) == + if (trans_scan_term(zh, zapt, termz+prefix_len, zm) == ZEBRA_FAIL) { for (i = 0; i < ord_no; i++) @@ -381,7 +393,7 @@ static ZEBRA_RES rpn_scan_ver2(ZebraHandle zh, ODR stream, NMEM nmem, termz[prefix_len] = 0; strcpy(ar[i].prefix, termz); - if (trans_scan_term(zh, zapt, termz+prefix_len, index_type) == + if (trans_scan_term(zh, zapt, termz+prefix_len, zm) == ZEBRA_FAIL) return ZEBRA_FAIL; wrbuf_rewind(ar[i].term); @@ -462,12 +474,12 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, const Odr_oid *attributeset, int num_bases, char **basenames, int *position, int *num_entries, ZebraScanEntry **list, - int *is_partial, RSET limit_set) + int *is_partial, const char *set_name) { int base_no; int ords[RPN_MAX_ORDS], ord_no = 0; - unsigned index_type; + const char *index_type; char *search_type = NULL; char rank_type[128]; int complete_flag; @@ -475,6 +487,7 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, NMEM nmem; ZEBRA_RES res; struct rset_key_control *kc = 0; + RSET limit_set = 0; *list = 0; *is_partial = 0; @@ -482,40 +495,41 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, if (!attributeset) attributeset = yaz_oid_attset_bib_1; - if (!limit_set) /* no limit set given already */ + if (!set_name) { /* see if there is a @attr 8=set */ AttrType termset; int termset_value_numeric; - const char *termset_value_string; + const char *termset_value_string = 0; attr_init_APT(&termset, zapt, 8); termset_value_numeric = attr_find_ex(&termset, NULL, &termset_value_string); if (termset_value_numeric != -1) { - char resname[32]; - const char *termset_name = 0; - if (termset_value_numeric != -2) { + char resname[32]; sprintf(resname, "%d", termset_value_numeric); - termset_name = resname; + set_name = odr_strdup(stream, resname); } else - termset_name = termset_value_string; - - limit_set = resultSetRef(zh, termset_name); - - if (!limit_set) - { - zebra_setError(zh, - YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST, - termset_name); - return ZEBRA_FAIL; - } + set_name = odr_strdup(stream, termset_value_string); } } + + if (set_name) + { + limit_set = resultSetRef(zh, set_name); + if (!limit_set) + { + zebra_setError(zh, + YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST, + set_name); + return ZEBRA_FAIL; + } + } + yaz_log(YLOG_DEBUG, "position = %d, num = %d", *position, *num_entries); @@ -531,7 +545,6 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, zebra_setError(zh, YAZ_BIB1_TOO_MANY_DATABASES_SPECIFIED, 0); return ZEBRA_FAIL; } - for (base_no = 0; base_no < num_bases; base_no++) { int ord; @@ -561,9 +574,9 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, nmem = nmem_create(); kc = zebra_key_control_create(zh); - res = rpn_scan_ver2(zh, stream, nmem, kc, zapt, position, num_entries, - list, - is_partial, limit_set, index_type, ord_no, ords); + res = rpn_scan_norm(zh, stream, nmem, kc, zapt, position, num_entries, + list, is_partial, limit_set, + index_type, ord_no, ords); nmem_destroy(nmem); (*kc->dec)(kc); return res;