X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=index%2Frpnsearch.c;h=47c1c234c84500c4dda68aab85c7a8301fc3cba5;hp=5fa197fe929bc6f30bf9820ee162ab79c181250f;hb=c5971ebf8a88865ed9a1f7c8cf9daa22544f07be;hpb=03419e1f6a4ae8a5b255e7c215da40678c30bb25 diff --git a/index/rpnsearch.c b/index/rpnsearch.c index 5fa197f..47c1c23 100644 --- a/index/rpnsearch.c +++ b/index/rpnsearch.c @@ -1,5 +1,5 @@ /* This file is part of the Zebra server. - Copyright (C) 1994-2010 Index Data + Copyright (C) Index Data Zebra is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free @@ -17,6 +17,9 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#if HAVE_CONFIG_H +#include +#endif #include #include #ifdef WIN32 @@ -68,20 +71,20 @@ void rpn_char_map_prepare(struct zebra_register *reg, zebra_map_t zm, dict_grep_cmap(reg->dict, map_info, rpn_char_map_handler); } -#define TERM_COUNT - -struct grep_info { -#ifdef TERM_COUNT - int *term_no; -#endif +#define TERM_COUNT + +struct grep_info { +#ifdef TERM_COUNT + int *term_no; +#endif ISAM_P *isam_p_buf; - int isam_p_size; + int isam_p_size; int isam_p_indx; int trunc_max; ZebraHandle zh; const char *index_type; ZebraSet termset; -}; +}; static int add_isam_p(const char *name, const char *info, struct grep_info *p) @@ -99,8 +102,8 @@ static int add_isam_p(const char *name, const char *info, if (p->isam_p_indx == p->isam_p_size) { ISAM_P *new_isam_p_buf; -#ifdef TERM_COUNT - int *new_term_no; +#ifdef TERM_COUNT + int *new_term_no; #endif p->isam_p_size = 2*p->isam_p_size + 100; new_isam_p_buf = (ISAM_P *) xmalloc(sizeof(*new_isam_p_buf) * @@ -134,13 +137,13 @@ static int add_isam_p(const char *name, const char *info, int ord = 0; const char *index_name; int len = key_SU_decode(&ord, (const unsigned char *) name); - + zebra_term_untrans (p->zh, p->index_type, term_tmp, name+len); yaz_log(log_level_rpn, "grep: %d %c %s", ord, name[len], term_tmp); zebraExplain_lookup_ord(p->zh->reg->zei, ord, 0 /* index_type */, &db, &index_name); yaz_log(log_level_rpn, "grep: db=%s index=%s", db, index_name); - + resultSetAddTerm(p->zh, p->termset, name[len], db, index_name, term_tmp); } @@ -184,7 +187,7 @@ static void esc_str(char *out_buf, size_t out_size, assert(in_buf); assert(out_size > 20); *out_buf = '\0'; - for (k = 0; k 0 && buf[--i] != '\x01') /* skip length */ + ; + while (i > 0 && buf[--i] != '\x01') /* skip accents */ + ; + return i; /* only basechars left */ +} + +static int term_102_icu(zebra_map_t zm, const char **src, WRBUF term_dict, int space_split, + WRBUF display_term) +{ + int no_terms = 0; + const char *s0 = *src, *s1; + while (*s0 == ' ') + s0++; + s1 = s0; + for (;;) + { + if (*s1 == ' ' && space_split) + break; + else if (*s1 && !strchr(REGEX_CHARS "-", *s1)) + s1++; + else + { + /* EOF or regex reserved char */ + if (s0 != s1) + { + const char *res_buf = 0; + size_t res_len = 0; + const char *display_buf; + size_t display_len; + + zebra_map_tokenize_start(zm, s0, s1 - s0); + + if (zebra_map_tokenize_next(zm, &res_buf, &res_len, + &display_buf, &display_len)) + { + size_t i; + res_len = icu_basechars(res_buf, res_len); + for (i = 0; i < res_len; i++) + { + if (strchr(REGEX_CHARS "\\", res_buf[i])) + wrbuf_putc(term_dict, '\\'); + if (res_buf[i] < 32) + wrbuf_putc(term_dict, '\x01'); + + wrbuf_putc(term_dict, res_buf[i]); + } + wrbuf_write(display_term, display_buf, display_len); + + no_terms++; + } + } + if (*s1 == '\0') + break; + + wrbuf_putc(term_dict, *s1); + wrbuf_putc(display_term, *s1); + + s1++; + s0 = s1; + } + } + if (no_terms) + wrbuf_puts(term_dict, "\x01\x01.*"); + *src = s1; + return no_terms; +} + +static int term_100_icu(zebra_map_t zm, + const char **src, WRBUF term_dict, WRBUF display_term, - int right_trunc) + int mode, + size_t token_number) { - int i; + size_t i; const char *res_buf = 0; size_t res_len = 0; const char *display_buf; size_t display_len; - if (!zebra_map_tokenize_next(zm, &res_buf, &res_len, - &display_buf, &display_len)) + + zebra_map_tokenize_start(zm, *src, strlen(*src)); + for (i = 0; i <= token_number; i++) { - *src += strlen(*src); - return 0; + if (!zebra_map_tokenize_next(zm, &res_buf, &res_len, + &display_buf, &display_len)) + return 0; } wrbuf_write(display_term, display_buf, display_len); - if (right_trunc) - { - /* ICU sort keys seem to be of the form - basechars \x01 accents \x01 length - For now we'll just right truncate from basechars . This - may give false hits due to accents not being used. - */ - i = res_len; - while (--i >= 0 && res_buf[i] != '\x01') - ; - if (i > 0) - { - while (--i >= 0 && res_buf[i] != '\x01') - ; - } - if (i == 0) - { /* did not find base chars at all. Throw error */ - return -1; - } - res_len = i; /* reduce res_len */ + if (mode) + { + res_len = icu_basechars(res_buf, res_len); } + if (mode & 2) + wrbuf_puts(term_dict, ".*"); for (i = 0; i < res_len; i++) { if (strchr(REGEX_CHARS "\\", res_buf[i])) wrbuf_putc(term_dict, '\\'); if (res_buf[i] < 32) - wrbuf_putc(term_dict, 1); - + wrbuf_putc(term_dict, '\x01'); + wrbuf_putc(term_dict, res_buf[i]); } - if (right_trunc) + if (mode & 1) wrbuf_puts(term_dict, ".*"); + else if (mode) + wrbuf_puts(term_dict, "\x01\x01.*"); return 1; } @@ -325,7 +394,7 @@ static int term_100(zebra_map_t zm, wrbuf_putc(display_term, *space_start); wrbuf_putc(term_dict, *space_start); space_start++; - + } /* and reset */ space_start = space_end = 0; @@ -418,7 +487,7 @@ static int term_103(zebra_map_t zm, const char **src, } } *src = s0; - + return i; } @@ -430,8 +499,8 @@ static int term_102(zebra_map_t zm, const char **src, } -/* term_104: handle term, process # and ! */ -static int term_104(zebra_map_t zm, const char **src, +/* term_104: handle term, process ?n * # */ +static int term_104(zebra_map_t zm, const char **src, WRBUF term_dict, int space_split, WRBUF display_term) { const char *s0; @@ -499,8 +568,8 @@ static int term_104(zebra_map_t zm, const char **src, return i; } -/* term_105/106: handle term, where trunc = Process * and ! and right trunc */ -static int term_105(zebra_map_t zm, const char **src, +/* term_105/106: handle term, process * ! and possibly right_truncate */ +static int term_105(zebra_map_t zm, const char **src, WRBUF term_dict, int space_split, WRBUF display_term, int right_truncate) { @@ -573,7 +642,7 @@ static void gen_regular_rel(WRBUF term_dict, int val, int islt) strcpy(dst, "(-[0-9]+|("); else strcpy(dst, "(("); - } + } else { if (!islt) @@ -600,7 +669,7 @@ static void gen_regular_rel(WRBUF term_dict, int val, int islt) if (d == '0') continue; d--; - } + } else { if (d == '9') @@ -608,7 +677,7 @@ static void gen_regular_rel(WRBUF term_dict, int val, int islt) d++; } } - + strcpy(dst + dst_p, numstr); dst_p = strlen(dst) - pos - 1; @@ -628,7 +697,7 @@ static void gen_regular_rel(WRBUF term_dict, int val, int islt) else { if (d != '9') - { + { dst[dst_p++] = '['; dst[dst_p++] = d; dst[dst_p++] = '-'; @@ -638,7 +707,7 @@ static void gen_regular_rel(WRBUF term_dict, int val, int islt) else dst[dst_p++] = d; } - for (i = 0; iisam_p_buf, grep_info->isam_p_indx, wrbuf_buf(display_term), - wrbuf_len(display_term), rank_type, + wrbuf_len(display_term), rank_type, 1 /* preserve pos */, zapt->term->which, rset_nmem, kc, kc->scope, ol, index_type, hits_limit_value, @@ -967,7 +1037,7 @@ static ZEBRA_RES search_term(ZebraHandle zh, } static ZEBRA_RES string_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, + const char **term_sub, WRBUF term_dict, const Odr_oid *attributeSet, NMEM stream, struct grep_info *grep_info, @@ -975,7 +1045,7 @@ static ZEBRA_RES string_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, WRBUF display_term, const char *xpath_use, struct ord_list **ol, - zebra_map_t zm) + zebra_map_t zm, size_t token_number) { int r; AttrType truncation; @@ -999,35 +1069,35 @@ static ZEBRA_RES string_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, yaz_log(log_level_rpn, "truncation value %d", truncation_value); termp = *term_sub; /* start of term for each database */ - + if (zebra_apt_get_ord(zh, zapt, index_type, xpath_use, attributeSet, &ord) != ZEBRA_OK) { *term_sub = 0; return ZEBRA_FAIL; } - + wrbuf_rewind(term_dict); /* new dictionary regexp term */ - + *ol = ord_list_append(stream, *ol, ord); ord_len = key_SU_encode(ord, ord_buf); - + wrbuf_putc(term_dict, '('); - - for (i = 0; ireg->dict, wrbuf_cstr(term_dict), regex_range, - grep_info, &max_pos, + grep_info, &max_pos, ord_len /* number of "exact" chars */, grep_handle); if (r == 1) @@ -1257,7 +1348,7 @@ static ZEBRA_RES grep_info_prepare(ZebraHandle zh, const char *termset_name = 0; if (termset_value_numeric != -2) { - + sprintf(resname, "%d", termset_value_numeric); termset_name = resname; } @@ -1293,18 +1384,18 @@ static ZEBRA_RES search_terms_chrmap(ZebraHandle zh, struct grep_info grep_info; const char *termp = termz; int alloc_sets = 0; - + *num_result_sets = 0; if (grep_info_prepare(zh, zapt, &grep_info, index_type) == ZEBRA_FAIL) return ZEBRA_FAIL; while (1) - { + { ZEBRA_RES res; if (alloc_sets == *num_result_sets) { int add = 10; - RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) * + RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) * sizeof(*rnew)); if (alloc_sets) memcpy(rnew, *result_sets, alloc_sets * sizeof(*rnew)); @@ -1317,7 +1408,8 @@ static ZEBRA_RES search_terms_chrmap(ZebraHandle zh, rank_type, xpath_use, rset_nmem, &(*result_sets)[*num_result_sets], - kc, zm); + kc, zm, + *num_result_sets); if (res != ZEBRA_OK) { int i; @@ -1336,7 +1428,7 @@ static ZEBRA_RES search_terms_chrmap(ZebraHandle zh, grep_info_delete(&grep_info); return ZEBRA_OK; } - + /** \brief Create result set(s) for list of terms \param zh Zebra Handle @@ -1367,8 +1459,6 @@ static ZEBRA_RES search_terms_list(ZebraHandle zh, struct rset_key_control *kc) { zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type); - if (zebra_maps_is_icu(zm)) - zebra_map_tokenize_start(zm, termz, strlen(termz)); return search_terms_chrmap(zh, zapt, termz, attributeSet, hits_limit, stream, index_type, complete_flag, rank_type, xpath_use, @@ -1396,7 +1486,7 @@ static ZEBRA_RES search_position(ZebraHandle zh, char *val; ISAM_P isam_p; zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type); - + attr_init_APT(&position, zapt, 3); position_value = attr_find(&position, NULL); switch(position_value) @@ -1435,7 +1525,7 @@ static ZEBRA_RES search_position(ZebraHandle zh, assert(*val == sizeof(ISAM_P)); memcpy(&isam_p, val+1, sizeof(isam_p)); - *rset = zebra_create_rset_isam(zh, rset_nmem, kc, kc->scope, + *rset = zebra_create_rset_isam(zh, rset_nmem, kc, kc->scope, isam_p, 0); } return ZEBRA_OK; @@ -1465,21 +1555,21 @@ static ZEBRA_RES rpn_search_APT_phrase(ZebraHandle zh, rank_type, xpath_use, rset_nmem, &result_sets, &num_result_sets, kc); - + if (res != ZEBRA_OK) return res; if (num_result_sets > 0) { RSET first_set = 0; - res = search_position(zh, zapt, attributeSet, + res = search_position(zh, zapt, attributeSet, index_type, rset_nmem, &first_set, kc); if (res != ZEBRA_OK) { int i; - for (i = 0; iscope, 2, tmp_set, @@ -1564,7 +1654,7 @@ static ZEBRA_RES rpn_search_APT_or_list(ZebraHandle zh, } } if (num_result_sets == 0) - *rset = rset_create_null(rset_nmem, kc, 0); + *rset = rset_create_null(rset_nmem, kc, 0); else if (num_result_sets == 1) *rset = result_sets[0]; else @@ -1583,9 +1673,9 @@ static ZEBRA_RES rpn_search_APT_and_list(ZebraHandle zh, const Odr_oid *attributeSet, zint hits_limit, NMEM stream, - const char *index_type, + const char *index_type, int complete_flag, - const char *rank_type, + const char *rank_type, const char *xpath_use, NMEM rset_nmem, RSET *rset, @@ -1603,16 +1693,16 @@ static ZEBRA_RES rpn_search_APT_and_list(ZebraHandle zh, kc); if (res != ZEBRA_OK) return res; - for (i = 0; iscope, 2, tmp_set, @@ -1634,7 +1724,7 @@ static ZEBRA_RES rpn_search_APT_and_list(ZebraHandle zh, if (num_result_sets == 0) - *rset = rset_create_null(rset_nmem, kc, 0); + *rset = rset_create_null(rset_nmem, kc, 0); else if (num_result_sets == 1) *rset = result_sets[0]; else @@ -1672,7 +1762,7 @@ static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt, case 1: yaz_log(log_level_rpn, "Relation <"); if (!term_100(zm, term_sub, term_num, 1, display_term)) - { + { wrbuf_destroy(term_num); return 0; } @@ -1710,12 +1800,13 @@ static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt, gen_regular_rel(term_dict, term_value+1, 0); break; case -1: + case 102: case 3: yaz_log(log_level_rpn, "Relation ="); if (!term_100(zm, term_sub, term_num, 1, display_term)) { wrbuf_destroy(term_num); - return 0; + return 0; } term_value = atoi(wrbuf_cstr(term_num)); wrbuf_printf(term_dict, "(0*%d)", term_value); @@ -1727,10 +1818,10 @@ static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt, break; default: *error_code = YAZ_BIB1_UNSUPP_RELATION_ATTRIBUTE; - wrbuf_destroy(term_num); + wrbuf_destroy(term_num); return 0; } - r = dict_lookup_grep(zh->reg->dict, wrbuf_cstr(term_dict), + r = dict_lookup_grep(zh->reg->dict, wrbuf_cstr(term_dict), 0, grep_info, max_pos, 0, grep_handle); if (r == 1) @@ -1743,7 +1834,7 @@ static int numeric_relation(ZebraHandle zh, Z_AttributesPlusTerm *zapt, } static ZEBRA_RES numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, + const char **term_sub, WRBUF term_dict, const Odr_oid *attributeSet, NMEM stream, struct grep_info *grep_info, @@ -1759,25 +1850,25 @@ static ZEBRA_RES numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, int ord, ord_len, i; char ord_buf[32]; zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type); - + *ol = ord_list_create(stream); rpn_char_map_prepare(zh->reg, zm, &rcmi); termp = *term_sub; - + if (zebra_apt_get_ord(zh, zapt, index_type, xpath_use, attributeSet, &ord) != ZEBRA_OK) { return ZEBRA_FAIL; } - + wrbuf_rewind(term_dict); - + *ol = ord_list_append(stream, *ol, ord); - + ord_len = key_SU_encode(ord, ord_buf); - + wrbuf_putc(term_dict, '('); for (i = 0; i < ord_len; i++) { @@ -1785,7 +1876,7 @@ static ZEBRA_RES numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, wrbuf_putc(term_dict, ord_buf[i]); } wrbuf_putc(term_dict, ')'); - + if (!numeric_relation(zh, zapt, &termp, term_dict, attributeSet, grep_info, &max_pos, zm, display_term, &relation_error)) @@ -1803,15 +1894,16 @@ static ZEBRA_RES numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, return ZEBRA_OK; } - + static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char *termz, const Odr_oid *attributeSet, + zint hits_limit, NMEM stream, - const char *index_type, + const char *index_type, int complete_flag, - const char *rank_type, + const char *rank_type, const char *xpath_use, NMEM rset_nmem, RSET *rset, @@ -1823,7 +1915,7 @@ static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh, ZEBRA_RES res; struct grep_info grep_info; int alloc_sets = 0; - zint hits_limit_value; + zint hits_limit_value = hits_limit; const char *term_ref_id_str = 0; zebra_term_limits_APT(zh, zapt, &hits_limit_value, &term_ref_id_str, @@ -1833,14 +1925,14 @@ static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh, if (grep_info_prepare(zh, zapt, &grep_info, index_type) == ZEBRA_FAIL) return ZEBRA_FAIL; while (1) - { + { struct ord_list *ol; WRBUF term_dict = wrbuf_alloc(); WRBUF display_term = wrbuf_alloc(); if (alloc_sets == num_result_sets) { int add = 10; - RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) * + RSET *rnew = (RSET *) nmem_malloc(stream, (alloc_sets+add) * sizeof(*rnew)); if (alloc_sets) memcpy(rnew, result_sets, alloc_sets * sizeof(*rnew)); @@ -1865,7 +1957,7 @@ static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh, grep_info.isam_p_indx, wrbuf_buf(display_term), wrbuf_len(display_term), rank_type, 0 /* preserve position */, - zapt->term->which, rset_nmem, + zapt->term->which, rset_nmem, kc, kc->scope, ol, index_type, hits_limit_value, term_ref_id_str); @@ -1903,7 +1995,7 @@ static ZEBRA_RES rpn_search_APT_local(ZebraHandle zh, { Record rec; zint sysno = atozint(termz); - + if (sysno <= 0) sysno = 0; rec = rec_get(zh->reg->records, sysno); @@ -1923,7 +2015,7 @@ static ZEBRA_RES rpn_search_APT_local(ZebraHandle zh, *rset = rset_create_temp(rset_nmem, kc, kc->scope, res_get(zh->res, "setTmpDir"), 0); rsfd = rset_open(*rset, RSETF_WRITE); - + key.mem[0] = sysno; key.mem[1] = 1; key.len = 2; @@ -1947,7 +2039,7 @@ static ZEBRA_RES rpn_sort_spec(ZebraHandle zh, Z_AttributesPlusTerm *zapt, Z_SortKeySpec *sks; Z_SortKey *sk; char termz[20]; - + attr_init_APT(&sort_relation_type, zapt, 7); sort_relation_value = attr_find(&sort_relation_type, &attributeSet); @@ -1957,7 +2049,7 @@ static ZEBRA_RES rpn_sort_spec(ZebraHandle zh, Z_AttributesPlusTerm *zapt, sort_sequence->specs = (Z_SortKeySpec **) nmem_malloc(stream, sort_sequence->num_specs * sizeof(*sort_sequence->specs)); - for (i = 0; inum_specs; i++) + for (i = 0; i < sort_sequence->num_specs; i++) sort_sequence->specs[i] = 0; } if (zapt->term->which != Z_Term_general) @@ -1988,7 +2080,7 @@ static ZEBRA_RES rpn_sort_spec(ZebraHandle zh, Z_AttributesPlusTerm *zapt, *sks->sortRelation = Z_SortKeySpec_ascending; else if (sort_relation_value == 2) *sks->sortRelation = Z_SortKeySpec_descending; - else + else *sks->sortRelation = Z_SortKeySpec_ascending; sks->caseSensitivity = (Odr_int *) @@ -2011,7 +2103,7 @@ static int rpn_check_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const Odr_oid *curAttributeSet = attributeSet; AttrType use; const char *use_string = 0; - + attr_init_APT(&use, zapt, 1); attr_find_ex(&use, &curAttributeSet, &use_string); @@ -2020,27 +2112,27 @@ static int rpn_check_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt, return zebra_parse_xpath_str(use_string, xpath, max, mem); } - - + + static RSET xpath_trunc(ZebraHandle zh, NMEM stream, - const char *index_type, const char *term, + const char *index_type, const char *term, const char *xpath_use, NMEM rset_nmem, struct rset_key_control *kc) { struct grep_info grep_info; - int ord = zebraExplain_lookup_attr_str(zh->reg->zei, + int ord = zebraExplain_lookup_attr_str(zh->reg->zei, zinfo_index_category_index, index_type, xpath_use); if (grep_info_prepare(zh, 0 /* zapt */, &grep_info, "0") == ZEBRA_FAIL) return rset_create_null(rset_nmem, kc, 0); - + if (ord < 0) return rset_create_null(rset_nmem, kc, 0); else { - int i, r, max_pos; + int i, max_pos; char ord_buf[32]; RSET rset; WRBUF term_dict = wrbuf_alloc(); @@ -2049,17 +2141,17 @@ static RSET xpath_trunc(ZebraHandle zh, NMEM stream, const char *flags = "void"; wrbuf_putc(term_dict, '('); - for (i = 0; ireg->dict, wrbuf_cstr(term_dict), 0, - &grep_info, &max_pos, 0, grep_handle); + dict_lookup_grep(zh->reg->dict, wrbuf_cstr(term_dict), 0, + &grep_info, &max_pos, 0, grep_handle); yaz_log(YLOG_DEBUG, "%s %d positions", term, grep_info.isam_p_indx); rset = rset_trunc(zh, grep_info.isam_p_buf, @@ -2091,7 +2183,7 @@ ZEBRA_RES rpn_search_xpath(ZebraHandle zh, } yaz_log(YLOG_DEBUG, "xpath len=%d", xpath_len); - for (i = 0; ireg->dict, 0, 0); - + { int level = xpath_len; int first_path = 1; - + while (--level >= 0) { WRBUF xpath_rev = wrbuf_alloc(); @@ -2161,7 +2253,7 @@ ZEBRA_RES rpn_search_xpath(ZebraHandle zh, { const char *cp = xpath[level].predicate->u.relation.value; wrbuf_putc(wbuf, '='); - + while (*cp) { if (strchr(REGEX_CHARS, *cp)) @@ -2171,12 +2263,12 @@ ZEBRA_RES rpn_search_xpath(ZebraHandle zh, } } rset_attr = xpath_trunc( - zh, stream, "0", wrbuf_cstr(wbuf), - ZEBRA_XPATH_ATTR_NAME, + zh, stream, "0", wrbuf_cstr(wbuf), + ZEBRA_XPATH_ATTR_NAME, rset_nmem, kc); wrbuf_destroy(wbuf); - } - else + } + else { if (!first_path) { @@ -2184,23 +2276,23 @@ ZEBRA_RES rpn_search_xpath(ZebraHandle zh, continue; } } - yaz_log(log_level_rpn, "xpath_rev (%d) = %s", level, + yaz_log(log_level_rpn, "xpath_rev (%d) = %s", level, wrbuf_cstr(xpath_rev)); if (wrbuf_len(xpath_rev)) { - rset_start_tag = xpath_trunc(zh, stream, "0", + rset_start_tag = xpath_trunc(zh, stream, "0", wrbuf_cstr(xpath_rev), - ZEBRA_XPATH_ELM_BEGIN, + ZEBRA_XPATH_ELM_BEGIN, rset_nmem, kc); if (always_matches) rset = rset_start_tag; else { - rset_end_tag = xpath_trunc(zh, stream, "0", + rset_end_tag = xpath_trunc(zh, stream, "0", wrbuf_cstr(xpath_rev), - ZEBRA_XPATH_ELM_END, + ZEBRA_XPATH_ELM_END, rset_nmem, kc); - + rset = rset_create_between(rset_nmem, kc, kc->scope, rset_start_tag, rset, rset_end_tag, rset_attr); @@ -2216,7 +2308,7 @@ ZEBRA_RES rpn_search_xpath(ZebraHandle zh, #define MAX_XPATH_STEPS 10 -static ZEBRA_RES rpn_search_database(ZebraHandle zh, +static ZEBRA_RES rpn_search_database(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const Odr_oid *attributeSet, zint hits_limit, NMEM stream, @@ -2229,7 +2321,7 @@ static ZEBRA_RES rpn_search_APT(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const Odr_oid *attributeSet, zint hits_limit, NMEM stream, Z_SortKeySpecList *sort_sequence, - int num_bases, const char **basenames, + int num_bases, const char **basenames, NMEM rset_nmem, RSET *rset, struct rset_key_control *kc) @@ -2259,12 +2351,12 @@ static ZEBRA_RES rpn_search_APT(ZebraHandle zh, Z_AttributesPlusTerm *zapt, rset_delete(rsets[i]); *rset = 0; } - else + else { if (num_bases == 1) *rset = rsets[0]; else if (num_bases == 0) - *rset = rset_create_null(rset_nmem, kc, 0); + *rset = rset_create_null(rset_nmem, kc, 0); else *rset = rset_create_or(rset_nmem, kc, kc->scope, 0 /* TERMID */, num_bases, rsets); @@ -2272,7 +2364,7 @@ static ZEBRA_RES rpn_search_APT(ZebraHandle zh, Z_AttributesPlusTerm *zapt, return res; } -static ZEBRA_RES rpn_search_database(ZebraHandle zh, +static ZEBRA_RES rpn_search_database(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const Odr_oid *attributeSet, zint hits_limit, NMEM stream, @@ -2299,7 +2391,7 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, } zebra_maps_attr(zh->reg->zebra_maps, zapt, &index_type, &search_type, rank_type, &complete_flag, &sort_flag); - + yaz_log(YLOG_DEBUG, "index_type=%s", index_type); yaz_log(YLOG_DEBUG, "complete_flag=%d", complete_flag); yaz_log(YLOG_DEBUG, "search_type=%s", search_type); @@ -2312,14 +2404,14 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, return rpn_sort_spec(zh, zapt, attributeSet, stream, sort_sequence, rank_type, rset_nmem, rset, kc); /* consider if an X-Path query is used */ - xpath_len = rpn_check_xpath(zh, zapt, attributeSet, + xpath_len = rpn_check_xpath(zh, zapt, attributeSet, xpath, MAX_XPATH_STEPS, stream); if (xpath_len >= 0) { - if (xpath[xpath_len-1].part[0] == '@') + if (xpath[xpath_len-1].part[0] == '@') xpath_use = ZEBRA_XPATH_ATTR_CDATA; /* last step is attribute */ else - xpath_use = ZEBRA_XPATH_CDATA; /* searching for cdata */ + xpath_use = ZEBRA_XPATH_CDATA; /* searching for cdata */ if (1) { @@ -2332,7 +2424,7 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, if (relation_value == 103) /* alwaysmatches */ { *rset = 0; /* signal no "term" set */ - return rpn_search_xpath(zh, stream, rank_type, *rset, + return rpn_search_xpath(zh, stream, rank_type, *rset, xpath_len, xpath, rset_nmem, rset, kc); } } @@ -2340,7 +2432,7 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, /* search using one of the various search type strategies termz is our UTF-8 search term - attributeSet is top-level default attribute set + attributeSet is top-level default attribute set stream is ODR for search reg_id is the register type complete_flag is 1 for complete subfield, 0 for incomplete @@ -2380,7 +2472,8 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, } else if (!strcmp(search_type, "numeric")) { - res = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, stream, + res = rpn_search_APT_numeric(zh, zapt, termz, attributeSet, hits_limit, + stream, index_type, complete_flag, rank_type, xpath_use, rset_nmem, @@ -2395,7 +2488,7 @@ static ZEBRA_RES rpn_search_database(ZebraHandle zh, return res; if (!*rset) return ZEBRA_FAIL; - return rpn_search_xpath(zh, stream, rank_type, *rset, + return rpn_search_xpath(zh, stream, rank_type, *rset, xpath_len, xpath, rset_nmem, rset, kc); } @@ -2429,9 +2522,9 @@ ZEBRA_RES rpn_get_top_approx_limit(ZebraHandle zh, Z_RPNStructure *zs, Z_AttributesPlusTerm *zapt = zs->u.simple->u.attributesPlusTerm; 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; @@ -2441,7 +2534,7 @@ ZEBRA_RES rpn_get_top_approx_limit(ZebraHandle zh, Z_RPNStructure *zs, } ZEBRA_RES rpn_search_top(ZebraHandle zh, Z_RPNStructure *zs, - const Odr_oid *attributeSet, + const Odr_oid *attributeSet, zint hits_limit, NMEM stream, NMEM rset_nmem, Z_SortKeySpecList *sort_sequence, @@ -2455,7 +2548,7 @@ ZEBRA_RES rpn_search_top(ZebraHandle zh, Z_RPNStructure *zs, res = rpn_search_structure(zh, zs, attributeSet, hits_limit, stream, rset_nmem, - sort_sequence, + sort_sequence, num_bases, basenames, &result_sets, &num_result_sets, 0 /* no parent op */, @@ -2463,7 +2556,7 @@ ZEBRA_RES rpn_search_top(ZebraHandle zh, Z_RPNStructure *zs, if (res != ZEBRA_OK) { int i; - for (i = 0; iwhich != zop->which @@ -2563,7 +2656,7 @@ ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs, case Z_Operator_prox: if (zop->u.prox->which != Z_ProximityOperator_known) { - zebra_setError(zh, + zebra_setError(zh, YAZ_BIB1_UNSUPP_PROX_UNIT_CODE, 0); return ZEBRA_FAIL; @@ -2579,9 +2672,9 @@ ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs, { rset = rset_create_prox(rset_nmem, kc, kc->scope, - *num_result_sets, *result_sets, + *num_result_sets, *result_sets, *zop->u.prox->ordered, - (!zop->u.prox->exclusion ? + (!zop->u.prox->exclusion ? 0 : *zop->u.prox->exclusion), *zop->u.prox->relationType, *zop->u.prox->distance ); @@ -2592,7 +2685,7 @@ ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs, return ZEBRA_FAIL; } *num_result_sets = 1; - *result_sets = nmem_malloc(stream, *num_result_sets * + *result_sets = nmem_malloc(stream, *num_result_sets * sizeof(**result_sets)); (*result_sets)[0] = rset; } @@ -2619,7 +2712,7 @@ ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs, rset = resultSetRef(zh, zs->u.simple->u.resultSetId); if (!rset) { - zebra_setError(zh, + zebra_setError(zh, YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST, zs->u.simple->u.resultSetId); return ZEBRA_FAIL; @@ -2632,7 +2725,7 @@ ZEBRA_RES rpn_search_structure(ZebraHandle zh, Z_RPNStructure *zs, return ZEBRA_FAIL; } *num_result_sets = 1; - *result_sets = nmem_malloc(stream, *num_result_sets * + *result_sets = nmem_malloc(stream, *num_result_sets * sizeof(**result_sets)); (*result_sets)[0] = rset; }