X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=index%2Fzrpn.c;h=caccbcde3be959f469a137d49b8b7c2b8e6cffec;hp=3df18e18cf4cc143b3044a46c78a09755a949dcc;hb=ecf94fc25324414287d3bbfbeba4440e558edd17;hpb=b551622098261f77647eb99bbbd4b92a3285ea5f diff --git a/index/zrpn.c b/index/zrpn.c index 3df18e1..caccbcd 100644 --- a/index/zrpn.c +++ b/index/zrpn.c @@ -1,6 +1,6 @@ -/* $Id: zrpn.c,v 1.130 2003-03-01 22:45:38 adam Exp $ - Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003 - Index Data Aps +/* $Id: zrpn.c,v 1.173 2005-04-13 13:03:47 adam Exp $ + Copyright (C) 1995-2005 + Index Data ApS This file is part of the Zebra server. @@ -34,17 +34,31 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include -#include -#include -#include -#include +#include + +/* maximum number of terms in an and/or/phrase item */ +#define TERM_LIST_LENGTH_MAX 256 + +static const struct key_control it_ctrl = +{ + sizeof(struct it_key), + 2, /* we have sysnos and seqnos in this key, nothing more */ + key_compare_it, + key_logdump_txt, /* FIXME - clean up these functions */ + key_get_seq, +}; + + +const struct key_control *key_it_ctrl = &it_ctrl; -struct rpn_char_map_info { +struct rpn_char_map_info +{ ZebraMaps zm; int reg_type; }; -typedef struct { +typedef struct +{ int type; int major; int minor; @@ -52,18 +66,21 @@ typedef struct { } AttrType; -static const char **rpn_char_map_handler (void *vp, const char **from, int len) +static int log_level_set = 0; +static int log_level_rpn = 0; + +static const char **rpn_char_map_handler(void *vp, const char **from, int len) { struct rpn_char_map_info *p = (struct rpn_char_map_info *) vp; - const char **out = zebra_maps_input (p->zm, p->reg_type, from, len); + const char **out = zebra_maps_input(p->zm, p->reg_type, from, len, 0); #if 0 if (out && *out) { const char *outp = *out; - yaz_log (LOG_LOG, "---"); + yaz_log(YLOG_LOG, "---"); while (*outp) { - yaz_log (LOG_LOG, "%02X", *outp); + yaz_log(YLOG_LOG, "%02X", *outp); outp++; } } @@ -71,16 +88,16 @@ static const char **rpn_char_map_handler (void *vp, const char **from, int len) return out; } -static void rpn_char_map_prepare (struct zebra_register *reg, int reg_type, - struct rpn_char_map_info *map_info) +static void rpn_char_map_prepare(struct zebra_register *reg, int reg_type, + struct rpn_char_map_info *map_info) { map_info->zm = reg->zebra_maps; map_info->reg_type = reg_type; - dict_grep_cmap (reg->dict, map_info, rpn_char_map_handler); + dict_grep_cmap(reg->dict, map_info, rpn_char_map_handler); } -static int attr_find_ex (AttrType *src, oid_value *attributeSetP, - const char **string_value) +static int attr_find_ex(AttrType *src, oid_value *attributeSetP, + const char **string_value) { int num_attributes; @@ -100,42 +117,42 @@ static int attr_find_ex (AttrType *src, oid_value *attributeSetP, { oident *attrset; - attrset = oid_getentbyoid (element->attributeSet); + attrset = oid_getentbyoid(element->attributeSet); *attributeSetP = attrset->value; } return *element->value.numeric; break; case Z_AttributeValue_complex: if (src->minor >= element->value.complex->num_list) - break; + break; if (element->attributeSet && attributeSetP) { oident *attrset; - attrset = oid_getentbyoid (element->attributeSet); + attrset = oid_getentbyoid(element->attributeSet); *attributeSetP = attrset->value; } if (element->value.complex->list[src->minor]->which == Z_StringOrNumeric_numeric) - { - ++(src->minor); - return - *element->value.complex->list[src->minor-1]->u.numeric; - } - else if (element->value.complex->list[src->minor]->which == - Z_StringOrNumeric_string) - { - if (!string_value) - break; - ++(src->minor); - *string_value = - element->value.complex->list[src->minor-1]->u.string; - return -2; - } - else - break; + { + ++(src->minor); + return + *element->value.complex->list[src->minor-1]->u.numeric; + } + else if (element->value.complex->list[src->minor]->which == + Z_StringOrNumeric_string) + { + if (!string_value) + break; + ++(src->minor); + *string_value = + element->value.complex->list[src->minor-1]->u.string; + return -2; + } + else + break; default: - assert (0); + assert(0); } } ++(src->major); @@ -143,12 +160,12 @@ static int attr_find_ex (AttrType *src, oid_value *attributeSetP, return -1; } -static int attr_find (AttrType *src, oid_value *attributeSetP) +static int attr_find(AttrType *src, oid_value *attributeSetP) { - return attr_find_ex (src, attributeSetP, 0); + return attr_find_ex(src, attributeSetP, 0); } -static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt, +static void attr_init(AttrType *src, Z_AttributesPlusTerm *zapt, int type) { src->zapt = zapt; @@ -163,7 +180,7 @@ struct grep_info { #ifdef TERM_COUNT int *term_no; #endif - ISAMS_P *isam_p_buf; + ISAM_P *isam_p_buf; int isam_p_size; int isam_p_indx; ZebraHandle zh; @@ -171,88 +188,92 @@ struct grep_info { ZebraSet termset; }; -static void term_untrans (ZebraHandle zh, int reg_type, - char *dst, const char *src) +static void term_untrans(ZebraHandle zh, int reg_type, + char *dst, const char *src) { int len = 0; while (*src) { - const char *cp = zebra_maps_output (zh->reg->zebra_maps, - reg_type, &src); - if (!cp && len < IT_MAX_WORD-1) - dst[len++] = *src++; - else - while (*cp && len < IT_MAX_WORD-1) - dst[len++] = *cp++; + const char *cp = zebra_maps_output(zh->reg->zebra_maps, + reg_type, &src); + if (!cp && len < IT_MAX_WORD-1) + dst[len++] = *src++; + else + while (*cp && len < IT_MAX_WORD-1) + dst[len++] = *cp++; } dst[len] = '\0'; } -static void add_isam_p (const char *name, const char *info, - struct grep_info *p) +static void add_isam_p(const char *name, const char *info, + struct grep_info *p) { + if (!log_level_set) + { + log_level_rpn = yaz_log_module_level("rpn"); + log_level_set = 1; + } if (p->isam_p_indx == p->isam_p_size) { - ISAMS_P *new_isam_p_buf; + ISAM_P *new_isam_p_buf; #ifdef TERM_COUNT int *new_term_no; #endif p->isam_p_size = 2*p->isam_p_size + 100; - new_isam_p_buf = (ISAMS_P *) xmalloc (sizeof(*new_isam_p_buf) * - p->isam_p_size); + new_isam_p_buf = (ISAM_P *) xmalloc(sizeof(*new_isam_p_buf) * + p->isam_p_size); if (p->isam_p_buf) { - memcpy (new_isam_p_buf, p->isam_p_buf, + memcpy(new_isam_p_buf, p->isam_p_buf, p->isam_p_indx * sizeof(*p->isam_p_buf)); - xfree (p->isam_p_buf); + xfree(p->isam_p_buf); } p->isam_p_buf = new_isam_p_buf; #ifdef TERM_COUNT - new_term_no = (int *) xmalloc (sizeof(*new_term_no) * - p->isam_p_size); + new_term_no = (int *) xmalloc(sizeof(*new_term_no) * p->isam_p_size); if (p->term_no) { - memcpy (new_term_no, p->isam_p_buf, + memcpy(new_term_no, p->isam_p_buf, p->isam_p_indx * sizeof(*p->term_no)); - xfree (p->term_no); + xfree(p->term_no); } p->term_no = new_term_no; #endif } - assert (*info == sizeof(*p->isam_p_buf)); - memcpy (p->isam_p_buf + p->isam_p_indx, info+1, sizeof(*p->isam_p_buf)); + assert(*info == sizeof(*p->isam_p_buf)); + memcpy(p->isam_p_buf + p->isam_p_indx, info+1, sizeof(*p->isam_p_buf)); #if 1 if (p->termset) { - const char *db; - int set, use; - char term_tmp[IT_MAX_WORD]; - int su_code = 0; - int len = key_SU_decode (&su_code, name); - - term_untrans (p->zh, p->reg_type, term_tmp, name+len+1); - logf (LOG_LOG, "grep: %d %c %s", su_code, name[len], term_tmp); - zebraExplain_lookup_ord (p->zh->reg->zei, - su_code, &db, &set, &use); - logf (LOG_LOG, "grep: set=%d use=%d db=%s", set, use, db); - - resultSetAddTerm (p->zh, p->termset, name[len], db, - set, use, term_tmp); + const char *db; + int set, use; + char term_tmp[IT_MAX_WORD]; + int su_code = 0; + int len = key_SU_decode (&su_code, name); + + term_untrans (p->zh, p->reg_type, term_tmp, name+len+1); + yaz_log(log_level_rpn, "grep: %d %c %s", su_code, name[len], term_tmp); + zebraExplain_lookup_ord (p->zh->reg->zei, + su_code, &db, &set, &use); + yaz_log(log_level_rpn, "grep: set=%d use=%d db=%s", set, use, db); + + resultSetAddTerm(p->zh, p->termset, name[len], db, + set, use, term_tmp); } #endif (p->isam_p_indx)++; } -static int grep_handle (char *name, const char *info, void *p) +static int grep_handle(char *name, const char *info, void *p) { - add_isam_p (name, info, (struct grep_info *) p); + add_isam_p(name, info, (struct grep_info *) p); return 0; } -static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src, - const char *ct1, const char *ct2) +static int term_pre(ZebraMaps zebra_maps, int reg_type, const char **src, + const char *ct1, const char *ct2, int first) { const char *s1, *s0 = *src; const char **map; @@ -260,12 +281,12 @@ static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src, /* skip white space */ while (*s0) { - if (ct1 && strchr (ct1, *s0)) + if (ct1 && strchr(ct1, *s0)) break; - if (ct2 && strchr (ct2, *s0)) + if (ct2 && strchr(ct2, *s0)) break; s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s1, strlen(s1)); + map = zebra_maps_input(zebra_maps, reg_type, &s1, strlen(s1), first); if (**map != *CHR_SPACE) break; s0 = s1; @@ -274,14 +295,41 @@ static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src, return *s0; } + +static void esc_str(char *out_buf, int out_size, + const char *in_buf, int in_size) +{ + int k; + + assert(out_buf); + assert(in_buf); + assert(out_size > 20); + *out_buf = '\0'; + for (k = 0; k 126) + pc = '?'; + else + pc = c; + sprintf(out_buf +strlen(out_buf), "%02X:%c ", c, pc); + if (strlen(out_buf) > out_size-20) + { + strcat(out_buf, ".."); + break; + } + } +} + #define REGEX_CHARS " []()|.*+?!" -/* term_100: handle term, where trunc=none (no operators at all) */ -static int term_100 (ZebraMaps zebra_maps, int reg_type, - const char **src, char *dst, int space_split, - char *dst_term) +/* term_100: handle term, where trunc = none(no operators at all) */ +static int term_100(ZebraMaps zebra_maps, int reg_type, + const char **src, char *dst, int space_split, + char *dst_term) { - const char *s0, *s1; + const char *s0; const char **map; int i = 0; int j = 0; @@ -289,47 +337,61 @@ static int term_100 (ZebraMaps zebra_maps, int reg_type, const char *space_start = 0; const char *space_end = 0; - if (!term_pre (zebra_maps, reg_type, src, NULL, NULL)) + if (!term_pre(zebra_maps, reg_type, src, NULL, NULL, !space_split)) return 0; s0 = *src; while (*s0) { - s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); - if (space_split) + const char *s1 = s0; + int q_map_match = 0; + map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0), + &q_map_match); + if (space_split) + { + if (**map == *CHR_SPACE) + break; + } + else /* complete subfield only. */ + { + if (**map == *CHR_SPACE) + { /* save space mapping for later .. */ + space_start = s1; + space_end = s0; + continue; + } + else if (space_start) + { /* reload last space */ + while (space_start < space_end) + { + if (strchr(REGEX_CHARS, *space_start)) + dst[i++] = '\\'; + dst_term[j++] = *space_start; + dst[i++] = *space_start++; + } + /* and reset */ + space_start = space_end = 0; + } + } + /* add non-space char */ + memcpy(dst_term+j, s1, s0 - s1); + j += (s0 - s1); + if (!q_map_match) { - if (**map == *CHR_SPACE) - break; + while (s1 < s0) + { + if (strchr(REGEX_CHARS, *s1)) + dst[i++] = '\\'; + dst[i++] = *s1++; + } } - else /* complete subfield only. */ + else { - if (**map == *CHR_SPACE) - { /* save space mapping for later .. */ - space_start = s1; - space_end = s0; - continue; - } - else if (space_start) - { /* reload last space */ - while (space_start < space_end) - { - if (strchr (REGEX_CHARS, *space_start)) - dst[i++] = '\\'; - dst_term[j++] = *space_start; - dst[i++] = *space_start++; - } - /* and reset */ - space_start = space_end = 0; - } + char tmpbuf[80]; + esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0])); + + strcpy(dst + i, map[0]); + i += strlen(map[0]); } - /* add non-space char */ - while (s1 < s0) - { - if (strchr(REGEX_CHARS, *s1)) - dst[i++] = '\\'; - dst_term[j++] = *s1; - dst[i++] = *s1++; - } } dst[i] = '\0'; dst_term[j] = '\0'; @@ -337,17 +399,17 @@ static int term_100 (ZebraMaps zebra_maps, int reg_type, return i; } -/* term_101: handle term, where trunc=Process # */ -static int term_101 (ZebraMaps zebra_maps, int reg_type, - const char **src, char *dst, int space_split, - char *dst_term) +/* term_101: handle term, where trunc = Process # */ +static int term_101(ZebraMaps zebra_maps, int reg_type, + const char **src, char *dst, int space_split, + char *dst_term) { - const char *s0, *s1; + const char *s0; const char **map; int i = 0; int j = 0; - if (!term_pre (zebra_maps, reg_type, src, "#", "#")) + if (!term_pre(zebra_maps, reg_type, src, "#", "#", !space_split)) return 0; s0 = *src; while (*s0) @@ -356,21 +418,37 @@ static int term_101 (ZebraMaps zebra_maps, int reg_type, { dst[i++] = '.'; dst[i++] = '*'; - dst_term[j++] = *s0++; + dst_term[j++] = *s0++; } else { - s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); + const char *s1 = s0; + int q_map_match = 0; + map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0), + &q_map_match); if (space_split && **map == *CHR_SPACE) break; - while (s1 < s0) - { - if (strchr(REGEX_CHARS, *s1)) - dst[i++] = '\\'; - dst_term[j++] = *s1; - dst[i++] = *s1++; - } + + /* add non-space char */ + memcpy(dst_term+j, s1, s0 - s1); + j += (s0 - s1); + if (!q_map_match) + { + while (s1 < s0) + { + if (strchr(REGEX_CHARS, *s1)) + dst[i++] = '\\'; + dst[i++] = *s1++; + } + } + else + { + char tmpbuf[80]; + esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0])); + + strcpy(dst + i, map[0]); + i += strlen(map[0]); + } } } dst[i] = '\0'; @@ -379,21 +457,21 @@ static int term_101 (ZebraMaps zebra_maps, int reg_type, return i; } -/* term_103: handle term, where trunc=re-2 (regular expressions) */ -static int term_103 (ZebraMaps zebra_maps, int reg_type, const char **src, - char *dst, int *errors, int space_split, - char *dst_term) +/* term_103: handle term, where trunc = re-2 (regular expressions) */ +static int term_103(ZebraMaps zebra_maps, int reg_type, const char **src, + char *dst, int *errors, int space_split, + char *dst_term) { int i = 0; int j = 0; - const char *s0, *s1; + const char *s0; const char **map; - if (!term_pre (zebra_maps, reg_type, src, "^\\()[].*+?|", "(")) + if (!term_pre(zebra_maps, reg_type, src, "^\\()[].*+?|", "(", !space_split)) return 0; s0 = *src; if (errors && *s0 == '+' && s0[1] && s0[2] == '+' && s0[3] && - isdigit (s0[1])) + isdigit(((const unsigned char *)s0)[1])) { *errors = s0[1] - '0'; s0 += 3; @@ -402,59 +480,76 @@ static int term_103 (ZebraMaps zebra_maps, int reg_type, const char **src, } while (*s0) { - if (strchr ("^\\()[].*+?|-", *s0)) - { - dst_term[j++] = *s0; + if (strchr("^\\()[].*+?|-", *s0)) + { + dst_term[j++] = *s0; dst[i++] = *s0++; - } + } else { - s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); - if (**map == *CHR_SPACE) + const char *s1 = s0; + int q_map_match = 0; + map = zebra_maps_search(zebra_maps, reg_type, &s0, strlen(s0), + &q_map_match); + if (space_split && **map == *CHR_SPACE) break; - while (s1 < s0) - { - if (strchr(REGEX_CHARS, *s1)) - dst[i++] = '\\'; - dst_term[j++] = *s1; - dst[i++] = *s1++; - } + + /* add non-space char */ + memcpy(dst_term+j, s1, s0 - s1); + j += (s0 - s1); + if (!q_map_match) + { + while (s1 < s0) + { + if (strchr(REGEX_CHARS, *s1)) + dst[i++] = '\\'; + dst[i++] = *s1++; + } + } + else + { + char tmpbuf[80]; + esc_str(tmpbuf, sizeof(tmpbuf), map[0], strlen(map[0])); + + strcpy(dst + i, map[0]); + i += strlen(map[0]); + } } } dst[i] = '\0'; dst_term[j] = '\0'; *src = s0; + return i; } -/* term_103: handle term, where trunc=re-1 (regular expressions) */ +/* term_103: handle term, where trunc = re-1 (regular expressions) */ static int term_102 (ZebraMaps zebra_maps, int reg_type, const char **src, - char *dst, int space_split, char *dst_term) + char *dst, int space_split, char *dst_term) { - return term_103 (zebra_maps, reg_type, src, dst, NULL, space_split, - dst_term); + return term_103(zebra_maps, reg_type, src, dst, NULL, space_split, + dst_term); } -/* term_104: handle term, where trunc=Process # and ! */ -static int term_104 (ZebraMaps zebra_maps, int reg_type, - const char **src, char *dst, int space_split, - char *dst_term) +/* term_104: handle term, where trunc = Process # and ! */ +static int term_104(ZebraMaps zebra_maps, int reg_type, + const char **src, char *dst, int space_split, + char *dst_term) { const char *s0, *s1; const char **map; int i = 0; int j = 0; - if (!term_pre (zebra_maps, reg_type, src, "?*#", "?*#")) + if (!term_pre(zebra_maps, reg_type, src, "?*#", "?*#", !space_split)) return 0; s0 = *src; while (*s0) { if (*s0 == '?') { - dst_term[j++] = *s0++; + dst_term[j++] = *s0++; if (*s0 >= '0' && *s0 <= '9') { int limit = 0; @@ -481,23 +576,24 @@ static int term_104 (ZebraMaps zebra_maps, int reg_type, { dst[i++] = '.'; dst[i++] = '*'; - dst_term[j++] = *s0++; + dst_term[j++] = *s0++; } else if (*s0 == '#') - { + { dst[i++] = '.'; - dst_term[j++] = *s0++; - } + dst_term[j++] = *s0++; + } + else { s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); + map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0); if (space_split && **map == *CHR_SPACE) break; while (s1 < s0) { if (strchr(REGEX_CHARS, *s1)) dst[i++] = '\\'; - dst_term[j++] = *s1; + dst_term[j++] = *s1; dst[i++] = *s1++; } } @@ -508,17 +604,17 @@ static int term_104 (ZebraMaps zebra_maps, int reg_type, return i; } -/* term_105/106: handle term, where trunc=Process * and ! and right trunc */ +/* term_105/106: handle term, where trunc = Process * and ! and right trunc */ static int term_105 (ZebraMaps zebra_maps, int reg_type, - const char **src, char *dst, int space_split, - char *dst_term, int right_truncate) + const char **src, char *dst, int space_split, + char *dst_term, int right_truncate) { const char *s0, *s1; const char **map; int i = 0; int j = 0; - if (!term_pre (zebra_maps, reg_type, src, "*!", "*!")) + if (!term_pre(zebra_maps, reg_type, src, "*!", "*!", !space_split)) return 0; s0 = *src; while (*s0) @@ -527,23 +623,24 @@ static int term_105 (ZebraMaps zebra_maps, int reg_type, { dst[i++] = '.'; dst[i++] = '*'; - dst_term[j++] = *s0++; + dst_term[j++] = *s0++; } else if (*s0 == '!') - { + { dst[i++] = '.'; - dst_term[j++] = *s0++; - } + dst_term[j++] = *s0++; + } + else { s1 = s0; - map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); + map = zebra_maps_input(zebra_maps, reg_type, &s0, strlen(s0), 0); if (space_split && **map == *CHR_SPACE) break; while (s1 < s0) { if (strchr(REGEX_CHARS, *s1)) dst[i++] = '\\'; - dst_term[j++] = *s1; + dst_term[j++] = *s1; dst[i++] = *s1++; } } @@ -565,38 +662,38 @@ static int term_105 (ZebraMaps zebra_maps, int reg_type, * val: border value (inclusive) * islt: 1 if <=; 0 if >=. */ -static void gen_regular_rel (char *dst, int val, int islt) +static void gen_regular_rel(char *dst, int val, int islt) { int dst_p; int w, d, i; int pos = 0; char numstr[20]; - logf (LOG_DEBUG, "gen_regular_rel. val=%d, islt=%d", val, islt); + yaz_log(YLOG_DEBUG, "gen_regular_rel. val=%d, islt=%d", val, islt); if (val >= 0) { if (islt) - strcpy (dst, "(-[0-9]+|("); + strcpy(dst, "(-[0-9]+|("); else - strcpy (dst, "(("); + strcpy(dst, "(("); } else { if (!islt) { - strcpy (dst, "([0-9]+|-("); - dst_p = strlen (dst); + strcpy(dst, "([0-9]+|-("); + dst_p = strlen(dst); islt = 1; } else { - strcpy (dst, "(-("); + strcpy(dst, "(-("); islt = 0; } val = -val; } - dst_p = strlen (dst); - sprintf (numstr, "%d", val); + dst_p = strlen(dst); + sprintf(numstr, "%d", val); for (w = strlen(numstr); --w >= 0; pos++) { d = numstr[w]; @@ -616,7 +713,7 @@ static void gen_regular_rel (char *dst, int val, int islt) } } - strcpy (dst + dst_p, numstr); + strcpy(dst + dst_p, numstr); dst_p = strlen(dst) - pos - 1; if (islt) @@ -658,25 +755,25 @@ static void gen_regular_rel (char *dst, int val, int islt) dst[dst_p] = '\0'; if (islt) { - /* match everything less than 10^(pos-1) */ - strcat (dst, "0*"); - for (i=1; ireg->zebra_maps, reg_type, - term_sub, term_component, - space_split, term_dst)) + term_sub, term_component, + space_split, term_dst)) return 0; - logf (LOG_DEBUG, "Relation <"); - - *term_tmp++ = '('; - for (i = 0; term_component[i]; ) - { - int j = 0; + yaz_log(log_level_rpn, "Relation <"); + + *term_tmp++ = '('; + for (i = 0; term_component[i]; ) + { + int j = 0; - if (i) - *term_tmp++ = '|'; - while (j < i) - string_rel_add_char (&term_tmp, term_component, &j); + if (i) + *term_tmp++ = '|'; + while (j < i) + string_rel_add_char (&term_tmp, term_component, &j); - *term_tmp++ = '['; + *term_tmp++ = '['; - *term_tmp++ = '^'; - string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = '-'; + *term_tmp++ = '^'; + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = '-'; - *term_tmp++ = ']'; - *term_tmp++ = '.'; - *term_tmp++ = '*'; - } - *term_tmp++ = ')'; - *term_tmp = '\0'; + *term_tmp++ = ']'; + *term_tmp++ = '.'; + *term_tmp++ = '*'; + + if ((term_tmp - term_dict) > IT_MAX_WORD) + break; + } + *term_tmp++ = ')'; + *term_tmp = '\0'; break; case 2: if (!term_100 (zh->reg->zebra_maps, reg_type, - term_sub, term_component, - space_split, term_dst)) + term_sub, term_component, + space_split, term_dst)) return 0; - logf (LOG_DEBUG, "Relation <="); + yaz_log(log_level_rpn, "Relation <="); - *term_tmp++ = '('; - for (i = 0; term_component[i]; ) - { - int j = 0; + *term_tmp++ = '('; + for (i = 0; term_component[i]; ) + { + int j = 0; - while (j < i) - string_rel_add_char (&term_tmp, term_component, &j); - *term_tmp++ = '['; + while (j < i) + string_rel_add_char (&term_tmp, term_component, &j); + *term_tmp++ = '['; - *term_tmp++ = '^'; - string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = '-'; + *term_tmp++ = '^'; + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = '-'; - *term_tmp++ = ']'; - *term_tmp++ = '.'; - *term_tmp++ = '*'; + *term_tmp++ = ']'; + *term_tmp++ = '.'; + *term_tmp++ = '*'; - *term_tmp++ = '|'; - } - for (i = 0; term_component[i]; ) - string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = ')'; - *term_tmp = '\0'; + *term_tmp++ = '|'; + + if ((term_tmp - term_dict) > IT_MAX_WORD) + break; + } + for (i = 0; term_component[i]; ) + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = ')'; + *term_tmp = '\0'; break; case 5: if (!term_100 (zh->reg->zebra_maps, reg_type, - term_sub, term_component, space_split, term_dst)) + term_sub, term_component, space_split, term_dst)) return 0; - logf (LOG_DEBUG, "Relation >"); + yaz_log(log_level_rpn, "Relation >"); - *term_tmp++ = '('; - for (i = 0; term_component[i];) - { - int j = 0; + *term_tmp++ = '('; + for (i = 0; term_component[i];) + { + int j = 0; - while (j < i) - string_rel_add_char (&term_tmp, term_component, &j); - *term_tmp++ = '['; - - *term_tmp++ = '^'; - *term_tmp++ = '-'; - string_rel_add_char (&term_tmp, term_component, &i); + while (j < i) + string_rel_add_char (&term_tmp, term_component, &j); + *term_tmp++ = '['; + + *term_tmp++ = '^'; + *term_tmp++ = '-'; + string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = ']'; - *term_tmp++ = '.'; - *term_tmp++ = '*'; + *term_tmp++ = ']'; + *term_tmp++ = '.'; + *term_tmp++ = '*'; - *term_tmp++ = '|'; - } - for (i = 0; term_component[i];) - string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = '.'; - *term_tmp++ = '+'; - *term_tmp++ = ')'; - *term_tmp = '\0'; + *term_tmp++ = '|'; + + if ((term_tmp - term_dict) > IT_MAX_WORD) + break; + } + for (i = 0; term_component[i];) + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = '.'; + *term_tmp++ = '+'; + *term_tmp++ = ')'; + *term_tmp = '\0'; break; case 4: if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub, - term_component, space_split, term_dst)) + term_component, space_split, term_dst)) return 0; - logf (LOG_DEBUG, "Relation >="); + yaz_log(log_level_rpn, "Relation >="); - *term_tmp++ = '('; - for (i = 0; term_component[i];) - { - int j = 0; + *term_tmp++ = '('; + for (i = 0; term_component[i];) + { + int j = 0; - if (i) - *term_tmp++ = '|'; - while (j < i) - string_rel_add_char (&term_tmp, term_component, &j); - *term_tmp++ = '['; + if (i) + *term_tmp++ = '|'; + while (j < i) + string_rel_add_char (&term_tmp, term_component, &j); + *term_tmp++ = '['; - if (term_component[i+1]) - { - *term_tmp++ = '^'; - *term_tmp++ = '-'; - string_rel_add_char (&term_tmp, term_component, &i); - } - else - { - string_rel_add_char (&term_tmp, term_component, &i); - *term_tmp++ = '-'; - } - *term_tmp++ = ']'; - *term_tmp++ = '.'; - *term_tmp++ = '*'; - } - *term_tmp++ = ')'; - *term_tmp = '\0'; + if (term_component[i+1]) + { + *term_tmp++ = '^'; + *term_tmp++ = '-'; + string_rel_add_char (&term_tmp, term_component, &i); + } + else + { + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = '-'; + } + *term_tmp++ = ']'; + *term_tmp++ = '.'; + *term_tmp++ = '*'; + + if ((term_tmp - term_dict) > IT_MAX_WORD) + break; + } + *term_tmp++ = ')'; + *term_tmp = '\0'; break; case 3: default: - logf (LOG_DEBUG, "Relation ="); + yaz_log(log_level_rpn, "Relation ="); if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub, - term_component, space_split, term_dst)) + term_component, space_split, term_dst)) return 0; - strcat (term_tmp, "("); - strcat (term_tmp, term_component); - strcat (term_tmp, ")"); + strcat(term_tmp, "("); + strcat(term_tmp, term_component); + strcat(term_tmp, ")"); } return 1; } -static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, +static int string_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char **term_sub, oid_value attributeSet, NMEM stream, struct grep_info *grep_info, @@ -856,37 +965,44 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, int num_bases, char **basenames, char *term_dst, int xpath_use); -static RSET term_trunc (ZebraHandle zh, Z_AttributesPlusTerm *zapt, +static RSET term_trunc(ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char **term_sub, oid_value attributeSet, NMEM stream, struct grep_info *grep_info, int reg_type, int complete_flag, int num_bases, char **basenames, char *term_dst, - const char *rank_type, int xpath_use) + const char *rank_type, int xpath_use, + NMEM rset_nmem) { int r; grep_info->isam_p_indx = 0; - r = string_term (zh, zapt, term_sub, attributeSet, stream, grep_info, + r = string_term(zh, zapt, term_sub, attributeSet, stream, grep_info, reg_type, complete_flag, num_bases, basenames, term_dst, xpath_use); if (r < 1) return 0; - logf (LOG_DEBUG, "term: %s", term_dst); - return rset_trunc (zh, grep_info->isam_p_buf, + yaz_log(log_level_rpn, "term: %s", term_dst); + return rset_trunc(zh, grep_info->isam_p_buf, grep_info->isam_p_indx, term_dst, strlen(term_dst), rank_type, 1 /* preserve pos */, - zapt->term->which); + zapt->term->which, rset_nmem, + key_it_ctrl,key_it_ctrl->scope); +} +static char *nmem_strdup_i(NMEM nmem, int v) +{ + char val_str[64]; + sprintf (val_str, "%d", v); + return nmem_strdup(nmem, val_str); } - -static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, - oid_value attributeSet, NMEM stream, - struct grep_info *grep_info, - int reg_type, int complete_flag, - int num_bases, char **basenames, - char *term_dst, int xpath_use) +static int string_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, + const char **term_sub, + oid_value attributeSet, NMEM stream, + struct grep_info *grep_info, + int reg_type, int complete_flag, + int num_bases, char **basenames, + char *term_dst, int xpath_use) { char term_dict[2*IT_MAX_WORD+4000]; int j, r, base_no; @@ -900,18 +1016,26 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, struct rpn_char_map_info rcmi; int space_split = complete_flag ? 0 : 1; + int bases_ok = 0; /* no of databases with OK attribute */ + int errCode = 0; /* err code (if any is not OK) */ + char *errString = 0; /* addinfo */ + rpn_char_map_prepare (zh->reg, reg_type, &rcmi); attr_init (&use, zapt, 1); use_value = attr_find_ex (&use, &curAttributeSet, &use_string); - logf (LOG_DEBUG, "string_term, use value %d", use_value); + yaz_log(log_level_rpn, "string_term, use value %d", use_value); attr_init (&truncation, zapt, 5); truncation_value = attr_find (&truncation, NULL); - logf (LOG_DEBUG, "truncation value %d", truncation_value); + yaz_log(log_level_rpn, "truncation value %d", truncation_value); if (use_value == -1) /* no attribute - assumy "any" */ use_value = 1016; for (base_no = 0; base_no < num_bases; base_no++) { + int ord = -1; + int attr_ok = 0; + int regex_range = 0; + int init_pos = 0; attent attp; data1_local_attribute id_xpath_attr; data1_local_attribute *local_attr; @@ -925,34 +1049,61 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, zh->errString = basenames[base_no]; return -1; } - if (use_value == -2) /* string attribute (assume IDXPATH/any) */ + if (xpath_use > 0 && use_value == -2) { - use_value = xpath_use; + /* xpath mode and we have a string attribute */ attp.local_attributes = &id_xpath_attr; attp.attset_ordinal = VAL_IDXPATH; id_xpath_attr.next = 0; + + use_value = xpath_use; /* xpath_use as use-attribute now */ id_xpath_attr.local = use_value; } - else if (curAttributeSet == VAL_IDXPATH) + else if (curAttributeSet == VAL_IDXPATH && use_value >= 0) { + /* X-Path attribute, use numeric value directly */ attp.local_attributes = &id_xpath_attr; attp.attset_ordinal = VAL_IDXPATH; id_xpath_attr.next = 0; id_xpath_attr.local = use_value; } - else + else if (use_string && + (ord = zebraExplain_lookup_attr_str(zh->reg->zei, + use_string)) >= 0) + { + /* we have a match for a raw string attribute */ + char ord_buf[32]; + int i, ord_len; + + if (prefix_len) + term_dict[prefix_len++] = '|'; + else + term_dict[prefix_len++] = '('; + + ord_len = key_SU_encode (ord, ord_buf); + for (i = 0; ierrCode = 114; - zh->errString = nmem_strdup (stream, val_str); + errCode = 114; + if (use_string) + errString = nmem_strdup(stream, use_string); + else + errString = nmem_strdup_i (stream, use_value); } else { @@ -964,167 +1115,160 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, oident.value = curAttributeSet; oid_ent_to_oid (&oident, oid); - zh->errCode = 121; - zh->errString = nmem_strdup (stream, oident.desc); + errCode = 121; + errString = nmem_strdup (stream, oident.desc); } - return -1; - } - } - for (local_attr = attp.local_attributes; local_attr; - local_attr = local_attr->next) - { - int ord; - char ord_buf[32]; - int i, ord_len; - - ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal, - local_attr->local); - if (ord < 0) continue; - if (prefix_len) - term_dict[prefix_len++] = '|'; - else - term_dict[prefix_len++] = '('; - - ord_len = key_SU_encode (ord, ord_buf); - for (i = 0; inext) + { + char ord_buf[32]; + int i, ord_len; + + ord = zebraExplain_lookup_attr_su(zh->reg->zei, + attp.attset_ordinal, + local_attr->local); + if (ord < 0) + continue; + if (prefix_len) + term_dict[prefix_len++] = '|'; + else + term_dict[prefix_len++] = '('; + + ord_len = key_SU_encode (ord, ord_buf); + for (i = 0; ierrCode = 114; - zh->errString = nmem_strdup (stream, val_str); - return -1; +#if 1 + bases_ok++; +#else + errCode = 114; + errString = nmem_strdup_i(stream, use_value); + continue; +#endif } + else + { + bases_ok++; /* this has OK attributes */ + attr_ok = 1; + } + term_dict[prefix_len++] = ')'; term_dict[prefix_len++] = 1; term_dict[prefix_len++] = reg_type; - logf (LOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]); + yaz_log(log_level_rpn, "reg_type = %d", term_dict[prefix_len-1]); term_dict[prefix_len] = '\0'; - j = prefix_len; - switch (truncation_value) - { - case -1: /* not specified */ - case 100: /* do not truncate */ - if (!string_relation (zh, zapt, &termp, term_dict, - attributeSet, - reg_type, space_split, term_dst)) - return 0; - logf (LOG_LOG, "dict_lookup_grep: %s", term_dict+prefix_len); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, - grep_info, &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep fail %d", r); - break; - case 1: /* right truncation */ - term_dict[j++] = '('; - if (!term_100 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ".*)"); - dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - break; - case 2: /* keft truncation */ - term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*'; - if (!term_100 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - break; - case 3: /* left&right truncation */ - term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*'; - if (!term_100 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ".*)"); - dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); + j = prefix_len; + switch (truncation_value) + { + case -1: /* not specified */ + case 100: /* do not truncate */ + if (!string_relation (zh, zapt, &termp, term_dict, + attributeSet, + reg_type, space_split, term_dst)) + return 0; break; + case 1: /* right truncation */ + term_dict[j++] = '('; + if (!term_100(zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ".*)"); + break; + case 2: /* keft truncation */ + term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*'; + if (!term_100(zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ")"); + break; + case 3: /* left&right truncation */ + term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*'; + if (!term_100(zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ".*)"); + break; + case 101: /* process # in term */ + term_dict[j++] = '('; + if (!term_101(zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ")"); + break; + case 102: /* Regexp-1 */ + term_dict[j++] = '('; + if (!term_102(zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ")"); + break; + case 103: /* Regexp-2 */ + r = 1; + term_dict[j++] = '('; + init_pos = 2; + if (!term_103 (zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, ®ex_range, + space_split, term_dst)) + return 0; + strcat(term_dict, ")"); + case 104: /* process # and ! in term */ + term_dict[j++] = '('; + if (!term_104 (zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat(term_dict, ")"); + break; + case 105: /* process * and ! in term */ + term_dict[j++] = '('; + if (!term_105 (zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst, 1)) + return 0; + strcat(term_dict, ")"); + break; + case 106: /* process * and ! in term */ + term_dict[j++] = '('; + if (!term_105 (zh->reg->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst, 0)) + return 0; + strcat(term_dict, ")"); + break; + default: zh->errCode = 120; + zh->errString = nmem_strdup_i(stream, truncation_value); return -1; - case 101: /* process # in term */ - term_dict[j++] = '('; - if (!term_101 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=#: %d", r); - break; - case 102: /* Regexp-1 */ - term_dict[j++] = '('; - if (!term_102 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - logf (LOG_DEBUG, "Regexp-1 tolerance=%d", r); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=regular: %d", - r); - break; - case 103: /* Regexp-2 */ - r = 1; - term_dict[j++] = '('; - if (!term_103 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, &r, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - logf (LOG_DEBUG, "Regexp-2 tolerance=%d", r); - r = dict_lookup_grep (zh->reg->dict, term_dict, r, grep_info, - &max_pos, 2, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=eregular: %d", - r); - break; - case 104: /* process # and ! in term */ - term_dict[j++] = '('; - if (!term_104 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=#/!: %d", r); - break; - case 105: /* process * and ! in term */ - term_dict[j++] = '('; - if (!term_105 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst, 1)) - return 0; - strcat (term_dict, ")"); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=*/!: %d", r); - break; - case 106: /* process * and ! in term */ - term_dict[j++] = '('; - if (!term_105 (zh->reg->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst, 0)) - return 0; - strcat (term_dict, ")"); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); - if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=*/!: %d", r); - break; } + if (attr_ok) + { + char buf[80]; + const char *input = term_dict + prefix_len; + esc_str(buf, sizeof(buf), input, strlen(input)); + } + if (attr_ok) + { + yaz_log(log_level_rpn, "dict_lookup_grep: %s", term_dict+prefix_len); + r = dict_lookup_grep(zh->reg->dict, term_dict, regex_range, + grep_info, &max_pos, init_pos, + grep_handle); + if (r) + yaz_log(YLOG_WARN, "dict_lookup_grep fail %d", r); + } + } + if (!bases_ok) + { + zh->errCode = errCode; + zh->errString = errString; + return -1; } *term_sub = termp; - logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx); + yaz_log(YLOG_DEBUG, "%d positions", grep_info->isam_p_indx); return 1; } @@ -1181,7 +1325,7 @@ static int zapt_term_to_utf8 (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } /* convert APT SCAN term to internal cmap */ -static int trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, +static int trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, char *termz, int reg_type) { char termz0[IT_MAX_WORD]; @@ -1197,261 +1341,29 @@ static int trans_scan_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, int i = 0; const char *space_map = NULL; int len; - - while ((len = (cp_end - cp)) > 0) - { - map = zebra_maps_input (zh->reg->zebra_maps, reg_type, &cp, len); - if (**map == *CHR_SPACE) - space_map = *map; - else - { - if (i && space_map) - for (src = space_map; *src; src++) - termz[i++] = *src; - space_map = NULL; - for (src = *map; *src; src++) - termz[i++] = *src; - } - } - termz[i] = '\0'; - } - return 0; -} - -static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no, - int ordered, int exclusion, int relation, int distance) -{ - int i; - RSFD *rsfd; - int *more; - struct it_key **buf; - RSET result; - char prox_term[1024]; - int length_prox_term = 0; - int min_nn = 10000000; - int term_index; - int term_type = Z_Term_characterString; - const char *flags = NULL; - - rsfd = (RSFD *) xmalloc (sizeof(*rsfd)*rset_no); - more = (int *) xmalloc (sizeof(*more)*rset_no); - buf = (struct it_key **) xmalloc (sizeof(*buf)*rset_no); - - *prox_term = '\0'; - for (i = 0; ino_rset_terms; j++) - { - const char *nflags = rset[i]->rset_terms[j]->flags; - char *term = rset[i]->rset_terms[j]->name; - int lterm = strlen(term); - if (lterm + length_prox_term < sizeof(prox_term)-1) - { - if (length_prox_term) - prox_term[length_prox_term++] = ' '; - strcpy (prox_term + length_prox_term, term); - length_prox_term += lterm; - } - if (min_nn > rset[i]->rset_terms[j]->nn) - min_nn = rset[i]->rset_terms[j]->nn; - flags = nflags; - term_type = rset[i]->rset_terms[j]->type; - - /* only if all term types are of type characterString .. */ - /* the resulting term is of that type */ - if (term_type != Z_Term_characterString) - term_type = Z_Term_general; - } - } - for (i = 0; inn = 0; - result = rset_create (rset_kind_null, &parms); - } - else if (ordered && relation == 3 && exclusion == 0 && distance == 1) - { - /* special proximity case = phrase search ... */ - rset_temp_parms parms; - RSFD rsfd_result; - - parms.rset_term = rset_term_create (prox_term, length_prox_term, - flags, term_type); - parms.rset_term->nn = min_nn; - parms.cmp = key_compare_it; - parms.key_size = sizeof (struct it_key); - parms.temp_path = res_get (zh->res, "setTmpDir"); - result = rset_create (rset_kind_temp, &parms); - rsfd_result = rset_open (result, RSETF_WRITE); - - while (*more) - { - for (i = 1; i 1) - { - more[i-1] = rset_read (rset[i-1], rsfd[i-1], - buf[i-1], &term_index); - break; - } - else if (cmp == 1) - { - if (buf[i-1]->seqno+1 != buf[i]->seqno) - { - more[i-1] = rset_read (rset[i-1], rsfd[i-1], - buf[i-1], &term_index); - break; - } - } - else - { - more[i] = rset_read (rset[i], rsfd[i], buf[i], - &term_index); - break; - } - } - if (i == rset_no) - { - rset_write (result, rsfd_result, buf[0]); - more[0] = rset_read (*rset, *rsfd, *buf, &term_index); - } - } - rset_close (result, rsfd_result); - } - else if (rset_no == 2) - { - /* generic proximity case (two input sets only) ... */ - rset_temp_parms parms; - RSFD rsfd_result; - - yaz_log (LOG_LOG, "generic prox, dist=%d, relation=%d, ordered=%d" - ", exclusion=%d", - distance, relation, ordered, exclusion); - parms.rset_term = rset_term_create (prox_term, length_prox_term, - flags, term_type); - parms.rset_term->nn = min_nn; - parms.cmp = key_compare_it; - parms.key_size = sizeof (struct it_key); - parms.temp_path = res_get (zh->res, "setTmpDir"); - result = rset_create (rset_kind_temp, &parms); - rsfd_result = rset_open (result, RSETF_WRITE); - - while (more[0] && more[1]) - { - int cmp = key_compare_it (buf[0], buf[1]); - if (cmp < -1) - more[0] = rset_read (rset[0], rsfd[0], buf[0], &term_index); - else if (cmp > 1) - more[1] = rset_read (rset[1], rsfd[1], buf[1], &term_index); - else - { - int sysno = buf[0]->sysno; - int seqno[500]; - int n = 0; - - seqno[n++] = buf[0]->seqno; - while ((more[0] = rset_read (rset[0], rsfd[0], buf[0], - &term_index)) && - sysno == buf[0]->sysno) - if (n < 500) - seqno[n++] = buf[0]->seqno; - do - { - for (i = 0; iseqno - seqno[i]; - int excl = exclusion; - if (!ordered && diff < 0) - diff = -diff; - switch (relation) - { - case 1: /* < */ - if (diff < distance && diff >= 0) - excl = !excl; - break; - case 2: /* <= */ - if (diff <= distance && diff >= 0) - excl = !excl; - break; - case 3: /* == */ - if (diff == distance && diff >= 0) - excl = !excl; - break; - case 4: /* >= */ - if (diff >= distance && diff >= 0) - excl = !excl; - break; - case 5: /* > */ - if (diff > distance && diff >= 0) - excl = !excl; - break; - case 6: /* != */ - if (diff != distance && diff >= 0) - excl = !excl; - break; - } - if (excl) - { - rset_write (result, rsfd_result, buf[1]); - break; - } - } - } while ((more[1] = rset_read (rset[1], rsfd[1], buf[1], - &term_index)) && - sysno == buf[1]->sysno); - } - } - rset_close (result, rsfd_result); - } - else - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (prox_term, length_prox_term, - flags, term_type); - parms.rset_term->nn = 0; - result = rset_create (rset_kind_null, &parms); - } - for (i = 0; i 0) + { + map = zebra_maps_input (zh->reg->zebra_maps, reg_type, &cp, len, 0); + if (**map == *CHR_SPACE) + space_map = *map; + else + { + if (i && space_map) + for (src = space_map; *src; src++) + termz[i++] = *src; + space_map = NULL; + for (src = *map; *src; src++) + termz[i++] = *src; + } + } + termz[i] = '\0'; } - xfree (buf); - xfree (more); - xfree (rsfd); - return result; + return 0; } - char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz, NMEM stream, unsigned reg_id) + const char *termz, NMEM stream, unsigned reg_id) { WRBUF wrbuf = 0; AttrType truncation; @@ -1464,33 +1376,33 @@ char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, switch (truncation_value) { default: - ex_list = ""; - break; + ex_list = ""; + break; case 101: - ex_list = "#"; - break; + ex_list = "#"; + break; case 102: case 103: - ex_list = 0; - break; + ex_list = 0; + break; case 104: - ex_list = "!#"; - break; + ex_list = "!#"; + break; case 105: - ex_list = "!*"; - break; + ex_list = "!*"; + break; } if (ex_list) - wrbuf = zebra_replace(zh->reg->zebra_maps, reg_id, ex_list, - termz, strlen(termz)); + wrbuf = zebra_replace(zh->reg->zebra_maps, reg_id, ex_list, + termz, strlen(termz)); if (!wrbuf) - return nmem_strdup(stream, termz); + return nmem_strdup(stream, termz); else { - char *buf = (char*) nmem_malloc (stream, wrbuf_len(wrbuf)+1); - memcpy (buf, wrbuf_buf(wrbuf), wrbuf_len(wrbuf)); - buf[wrbuf_len(wrbuf)] = '\0'; - return buf; + char *buf = (char*) nmem_malloc(stream, wrbuf_len(wrbuf)+1); + memcpy (buf, wrbuf_buf(wrbuf), wrbuf_len(wrbuf)); + buf[wrbuf_len(wrbuf)] = '\0'; + return buf; } } @@ -1503,10 +1415,10 @@ static void grep_info_delete (struct grep_info *grep_info) } static int grep_info_prepare (ZebraHandle zh, - Z_AttributesPlusTerm *zapt, - struct grep_info *grep_info, - int reg_type, - NMEM stream) + Z_AttributesPlusTerm *zapt, + struct grep_info *grep_info, + int reg_type, + NMEM stream) { AttrType termset; int termset_value_numeric; @@ -1525,203 +1437,164 @@ static int grep_info_prepare (ZebraHandle zh, return 0; attr_init (&termset, zapt, 8); termset_value_numeric = - attr_find_ex (&termset, NULL, &termset_value_string); + 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]; + const char *termset_name = 0; + if (termset_value_numeric != -2) + { - sprintf (resname, "%d", termset_value_numeric); - termset_name = resname; - } - else - termset_name = termset_value_string; - logf (LOG_LOG, "creating termset set %s", termset_name); - grep_info->termset = resultSetAdd (zh, termset_name, 1); - if (!grep_info->termset) - { - zh->errCode = 128; - zh->errString = nmem_strdup (stream, termset_name); - return -1; - } + sprintf (resname, "%d", termset_value_numeric); + termset_name = resname; + } + else + termset_name = termset_value_string; + yaz_log(log_level_rpn, "creating termset set %s", termset_name); + grep_info->termset = resultSetAdd (zh, termset_name, 1); + if (!grep_info->termset) + { + zh->errCode = 128; + zh->errString = nmem_strdup (stream, termset_name); + return -1; + } } return 0; } - + static RSET rpn_search_APT_phrase (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz_org, + const char *termz_org, oid_value attributeSet, - NMEM stream, - int reg_type, int complete_flag, - const char *rank_type, int xpath_use, - int num_bases, char **basenames) + NMEM stream, + int reg_type, int complete_flag, + const char *rank_type, int xpath_use, + int num_bases, char **basenames, + NMEM rset_nmem) { char term_dst[IT_MAX_WORD+1]; - RSET rset[60], result; - int i, rset_no = 0; + RSET rset[TERM_LIST_LENGTH_MAX], result; + size_t rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; *term_dst = 0; if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream)) - return 0; - while (1) + return 0; + for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++) { - logf (LOG_DEBUG, "APT_phrase termp=%s", termp); - rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + yaz_log(log_level_rpn, "APT_phrase termp=%s", termp); + rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst, rank_type, - xpath_use); + xpath_use,rset_nmem); if (!rset[rset_no]) break; - if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) - break; } grep_info_delete (&grep_info); if (rset_no == 0) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (termz, -1, rank_type, - zapt->term->which); - return rset_create (rset_kind_null, &parms); - } + return rsnull_create (rset_nmem,key_it_ctrl); else if (rset_no == 1) return (rset[0]); - result = rpn_prox (zh, rset, rset_no, 1, 0, 3, 1); - for (i = 0; iscope, + rset_no, rset, + 1 /* ordered */, 0 /* exclusion */, + 3 /* relation */, 1 /* distance */); return result; } static RSET rpn_search_APT_or_list (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz_org, + const char *termz_org, oid_value attributeSet, - NMEM stream, - int reg_type, int complete_flag, - const char *rank_type, + NMEM stream, + int reg_type, int complete_flag, + const char *rank_type, int xpath_use, - int num_bases, char **basenames) + int num_bases, char **basenames, + NMEM rset_nmem) { char term_dst[IT_MAX_WORD+1]; - RSET rset[60], result; - int i, rset_no = 0; + RSET rset[TERM_LIST_LENGTH_MAX]; + size_t rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream)) - return 0; - while (1) + return 0; + for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++) { - logf (LOG_DEBUG, "APT_or_list termp=%s", termp); - rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + yaz_log(log_level_rpn, "APT_or_list termp=%s", termp); + rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst, rank_type, - xpath_use); + xpath_use,rset_nmem); if (!rset[rset_no]) break; - if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) - break; } grep_info_delete (&grep_info); if (rset_no == 0) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (termz, -1, rank_type, - zapt->term->which); - return rset_create (rset_kind_null, &parms); - } - result = rset[0]; - for (i = 1; iscope, + rset_no, rset); } static RSET rpn_search_APT_and_list (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz_org, + const char *termz_org, oid_value attributeSet, - NMEM stream, - int reg_type, int complete_flag, - const char *rank_type, + NMEM stream, + int reg_type, int complete_flag, + const char *rank_type, int xpath_use, - int num_bases, char **basenames) + int num_bases, char **basenames, + NMEM rset_nmem) { char term_dst[IT_MAX_WORD+1]; - RSET rset[60], result; - int i, rset_no = 0; + RSET rset[TERM_LIST_LENGTH_MAX]; + size_t rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream)) - return 0; - while (1) + return 0; + for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++) { - logf (LOG_DEBUG, "APT_and_list termp=%s", termp); - rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + yaz_log(log_level_rpn, "APT_and_list termp=%s", termp); + rset[rset_no] = term_trunc(zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst, rank_type, - xpath_use); + xpath_use, rset_nmem); if (!rset[rset_no]) break; - assert (rset[rset_no]); - if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) - break; } grep_info_delete (&grep_info); if (rset_no == 0) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (termz, -1, rank_type, - zapt->term->which); - return rset_create (rset_kind_null, &parms); - } - result = rset[0]; - for (i = 1; iscope, + rset_no, rset); } static int numeric_relation (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, - char *term_dict, - oid_value attributeSet, - struct grep_info *grep_info, - int *max_pos, - int reg_type, - char *term_dst) + const char **term_sub, + char *term_dict, + oid_value attributeSet, + struct grep_info *grep_info, + int *max_pos, + int reg_type, + char *term_dst) { AttrType relation; int relation_value; @@ -1732,50 +1605,50 @@ static int numeric_relation (ZebraHandle zh, Z_AttributesPlusTerm *zapt, attr_init (&relation, zapt, 2); relation_value = attr_find (&relation, NULL); - logf (LOG_DEBUG, "numeric relation value=%d", relation_value); + yaz_log(log_level_rpn, "numeric relation value=%d", relation_value); if (!term_100 (zh->reg->zebra_maps, reg_type, term_sub, term_tmp, 1, - term_dst)) - return 0; + term_dst)) + return 0; term_value = atoi (term_tmp); switch (relation_value) { case 1: - logf (LOG_DEBUG, "Relation <"); + yaz_log(log_level_rpn, "Relation <"); gen_regular_rel (term_tmp, term_value-1, 1); break; case 2: - logf (LOG_DEBUG, "Relation <="); + yaz_log(log_level_rpn, "Relation <="); gen_regular_rel (term_tmp, term_value, 1); break; case 4: - logf (LOG_DEBUG, "Relation >="); + yaz_log(log_level_rpn, "Relation >="); gen_regular_rel (term_tmp, term_value, 0); break; case 5: - logf (LOG_DEBUG, "Relation >"); + yaz_log(log_level_rpn, "Relation >"); gen_regular_rel (term_tmp, term_value+1, 0); break; case 3: default: - logf (LOG_DEBUG, "Relation ="); - sprintf (term_tmp, "(0*%d)", term_value); + yaz_log(log_level_rpn, "Relation ="); + sprintf (term_tmp, "(0*%d)", term_value); } - logf (LOG_DEBUG, "dict_lookup_grep: %s", term_tmp); - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, grep_info, max_pos, + yaz_log(log_level_rpn, "dict_lookup_grep: %s", term_tmp); + r = dict_lookup_grep(zh->reg->dict, term_dict, 0, grep_info, max_pos, 0, grep_handle); if (r) - logf (LOG_WARN, "dict_lookup_grep fail, rel=gt: %d", r); - logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx); + yaz_log(YLOG_WARN, "dict_lookup_grep fail, rel = gt: %d", r); + yaz_log(log_level_rpn, "%d positions", grep_info->isam_p_indx); return 1; } -static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, - oid_value attributeSet, struct grep_info *grep_info, - int reg_type, int complete_flag, - int num_bases, char **basenames, - char *term_dst, int xpath_use, NMEM stream) +static int numeric_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, + const char **term_sub, + oid_value attributeSet, struct grep_info *grep_info, + int reg_type, int complete_flag, + int num_bases, char **basenames, + char *term_dst, int xpath_use, NMEM stream) { char term_dict[2*IT_MAX_WORD+2]; int r, base_no; @@ -1786,6 +1659,10 @@ static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char *termp; struct rpn_char_map_info rcmi; + int bases_ok = 0; /* no of databases with OK attribute */ + int errCode = 0; /* err code (if any is not OK) */ + char *errString = 0; /* addinfo */ + rpn_char_map_prepare (zh->reg, reg_type, &rcmi); attr_init (&use, zapt, 1); use_value = attr_find_ex (&use, &curAttributeSet, &use_string); @@ -1809,7 +1686,7 @@ static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, id_xpath_attr.next = 0; id_xpath_attr.local = use_value; } - else if (curAttributeSet == VAL_IDXPATH) + else if (curAttributeSet == VAL_IDXPATH) { attp.local_attributes = &id_xpath_attr; attp.attset_ordinal = VAL_IDXPATH; @@ -1818,20 +1695,19 @@ static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } else { - if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value))) + if ((r = att_getentbyatt (zh, &attp, curAttributeSet, use_value, + use_string))) { - logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d", + yaz_log(YLOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d", curAttributeSet, use_value, r); if (r == -1) - { - char val_str[32]; - sprintf (val_str, "%d", use_value); - zh->errString = nmem_strdup (stream, val_str); - zh->errCode = 114; - } + { + errString = nmem_strdup_i(stream, use_value); + errCode = 114; + } else - zh->errCode = 121; - return -1; + errCode = 121; + continue; } } if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no])) @@ -1844,11 +1720,12 @@ static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, local_attr = local_attr->next) { int ord; - char ord_buf[32]; - int i, ord_len; + char ord_buf[32]; + int i, ord_len; - ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal, - local_attr->local); + ord = zebraExplain_lookup_attr_su(zh->reg->zei, + attp.attset_ordinal, + local_attr->local); if (ord < 0) continue; if (prefix_len) @@ -1856,130 +1733,120 @@ static int numeric_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, else term_dict[prefix_len++] = '('; - ord_len = key_SU_encode (ord, ord_buf); - for (i = 0; ierrCode = 114; - zh->errString = nmem_strdup (stream, val_str); - return -1; + errCode = 114; + errString = nmem_strdup_i(stream, use_value); + continue; } + bases_ok++; term_dict[prefix_len++] = ')'; term_dict[prefix_len++] = 1; term_dict[prefix_len++] = reg_type; - logf (LOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]); + yaz_log(YLOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]); term_dict[prefix_len] = '\0'; if (!numeric_relation (zh, zapt, &termp, term_dict, - attributeSet, grep_info, &max_pos, reg_type, - term_dst)) - return 0; + attributeSet, grep_info, &max_pos, reg_type, + term_dst)) + return 0; + } + if (!bases_ok) + { + zh->errCode = errCode; + zh->errString = errString; + return -1; } *term_sub = termp; - logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx); + yaz_log(YLOG_DEBUG, "%d positions", grep_info->isam_p_indx); return 1; } static RSET rpn_search_APT_numeric (ZebraHandle zh, - Z_AttributesPlusTerm *zapt, - const char *termz, - oid_value attributeSet, - NMEM stream, - int reg_type, int complete_flag, - const char *rank_type, int xpath_use, - int num_bases, char **basenames) + Z_AttributesPlusTerm *zapt, + const char *termz, + oid_value attributeSet, + NMEM stream, + int reg_type, int complete_flag, + const char *rank_type, int xpath_use, + int num_bases, char **basenames, + NMEM rset_nmem) { char term_dst[IT_MAX_WORD+1]; const char *termp = termz; - RSET rset[60], result; - int i, r, rset_no = 0; + RSET rset[TERM_LIST_LENGTH_MAX]; + int r; + size_t rset_no = 0; struct grep_info grep_info; + yaz_log(log_level_rpn, "APT_numeric t='%s'",termz); if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream)) - return 0; - while (1) + return 0; + for (; rset_no < sizeof(rset)/sizeof(*rset); rset_no++) { - logf (LOG_DEBUG, "APT_numeric termp=%s", termp); - grep_info.isam_p_indx = 0; - r = numeric_term (zh, zapt, &termp, attributeSet, &grep_info, - reg_type, complete_flag, num_bases, basenames, - term_dst, xpath_use, - stream); + yaz_log(YLOG_DEBUG, "APT_numeric termp=%s", termp); + grep_info.isam_p_indx = 0; + r = numeric_term(zh, zapt, &termp, attributeSet, &grep_info, + reg_type, complete_flag, num_bases, basenames, + term_dst, xpath_use, + stream); if (r < 1) break; - logf (LOG_DEBUG, "term: %s", term_dst); - rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf, + yaz_log(YLOG_DEBUG, "term: %s", term_dst); + rset[rset_no] = rset_trunc(zh, grep_info.isam_p_buf, grep_info.isam_p_indx, term_dst, - strlen(term_dst), rank_type, + strlen(term_dst), rank_type, 0 /* preserve position */, - zapt->term->which); - assert (rset[rset_no]); - if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) - break; + zapt->term->which, rset_nmem, + key_it_ctrl,key_it_ctrl->scope); + if (!rset[rset_no]) + break; } grep_info_delete (&grep_info); if (rset_no == 0) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (term_dst, -1, rank_type, - zapt->term->which); - return rset_create (rset_kind_null, &parms); - } - result = rset[0]; - for (i = 1; iscope, + rset_no, rset); } static RSET rpn_search_APT_local (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz, + const char *termz, oid_value attributeSet, - NMEM stream, - const char *rank_type) + NMEM stream, + const char *rank_type, NMEM rset_nmem) { RSET result; RSFD rsfd; struct it_key key; - rset_temp_parms parms; - - parms.rset_term = rset_term_create (termz, -1, rank_type, - zapt->term->which); - parms.cmp = key_compare_it; - parms.key_size = sizeof (struct it_key); - parms.temp_path = res_get (zh->res, "setTmpDir"); - result = rset_create (rset_kind_temp, &parms); + int sys; + result = rstemp_create( rset_nmem,key_it_ctrl,key_it_ctrl->scope, + res_get (zh->res, "setTmpDir"),0 ); rsfd = rset_open (result, RSETF_WRITE); - key.sysno = atoi (termz); - key.seqno = 1; - if (key.sysno <= 0) - key.sysno = 1; - rset_write (result, rsfd, &key); - rset_close (result, rsfd); + sys = atoi(termz); + if (sys <= 0) + sys = 1; + key.mem[0] = sys; + key.mem[1] = 1; + key.len = 2; + rset_write (rsfd, &key); + rset_close (rsfd); return result; } static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - oid_value attributeSet, NMEM stream, - Z_SortKeySpecList *sort_sequence, - const char *rank_type) + oid_value attributeSet, NMEM stream, + Z_SortKeySpecList *sort_sequence, + const char *rank_type) { - rset_null_parms parms; int i; int sort_relation_value; AttrType sort_relation_type; @@ -2000,104 +1867,75 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!sort_sequence->specs) { - sort_sequence->num_specs = 10; - sort_sequence->specs = (Z_SortKeySpec **) - nmem_malloc (stream, sort_sequence->num_specs * - sizeof(*sort_sequence->specs)); - for (i = 0; inum_specs; i++) - sort_sequence->specs[i] = 0; + sort_sequence->num_specs = 10; + sort_sequence->specs = (Z_SortKeySpec **) + nmem_malloc(stream, sort_sequence->num_specs * + sizeof(*sort_sequence->specs)); + for (i = 0; inum_specs; i++) + sort_sequence->specs[i] = 0; } if (zapt->term->which != Z_Term_general) - i = 0; + i = 0; else - i = atoi_n ((char *) zapt->term->u.general->buf, - zapt->term->u.general->len); + i = atoi_n ((char *) zapt->term->u.general->buf, + zapt->term->u.general->len); if (i >= sort_sequence->num_specs) - i = 0; + i = 0; sprintf (termz, "%d", i); oe.proto = PROTO_Z3950; oe.oclass = CLASS_ATTSET; oe.value = attributeSet; if (!oid_ent_to_oid (&oe, oid)) - return 0; + return 0; - sks = (Z_SortKeySpec *) nmem_malloc (stream, sizeof(*sks)); + sks = (Z_SortKeySpec *) nmem_malloc(stream, sizeof(*sks)); sks->sortElement = (Z_SortElement *) - nmem_malloc (stream, sizeof(*sks->sortElement)); + nmem_malloc(stream, sizeof(*sks->sortElement)); sks->sortElement->which = Z_SortElement_generic; sk = sks->sortElement->u.generic = (Z_SortKey *) - nmem_malloc (stream, sizeof(*sk)); + nmem_malloc(stream, sizeof(*sk)); sk->which = Z_SortKey_sortAttributes; sk->u.sortAttributes = (Z_SortAttributes *) - nmem_malloc (stream, sizeof(*sk->u.sortAttributes)); + nmem_malloc(stream, sizeof(*sk->u.sortAttributes)); sk->u.sortAttributes->id = oid; sk->u.sortAttributes->list = (Z_AttributeList *) - nmem_malloc (stream, sizeof(*sk->u.sortAttributes->list)); + nmem_malloc(stream, sizeof(*sk->u.sortAttributes->list)); sk->u.sortAttributes->list->num_attributes = 1; sk->u.sortAttributes->list->attributes = (Z_AttributeElement **) - nmem_malloc (stream, sizeof(*sk->u.sortAttributes->list->attributes)); + nmem_malloc(stream, sizeof(*sk->u.sortAttributes->list->attributes)); ae = *sk->u.sortAttributes->list->attributes = (Z_AttributeElement *) - nmem_malloc (stream, sizeof(**sk->u.sortAttributes->list->attributes)); + nmem_malloc(stream, sizeof(**sk->u.sortAttributes->list->attributes)); ae->attributeSet = 0; - ae->attributeType = (int *) - nmem_malloc (stream, sizeof(*ae->attributeType)); + ae->attributeType = (int *) + nmem_malloc(stream, sizeof(*ae->attributeType)); *ae->attributeType = 1; ae->which = Z_AttributeValue_numeric; ae->value.numeric = (int *) - nmem_malloc (stream, sizeof(*ae->value.numeric)); + nmem_malloc(stream, sizeof(*ae->value.numeric)); *ae->value.numeric = use_value; sks->sortRelation = (int *) - nmem_malloc (stream, sizeof(*sks->sortRelation)); + nmem_malloc(stream, sizeof(*sks->sortRelation)); if (sort_relation_value == 1) - *sks->sortRelation = Z_SortRelation_ascending; + *sks->sortRelation = Z_SortKeySpec_ascending; else if (sort_relation_value == 2) - *sks->sortRelation = Z_SortRelation_descending; + *sks->sortRelation = Z_SortKeySpec_descending; else - *sks->sortRelation = Z_SortRelation_ascending; + *sks->sortRelation = Z_SortKeySpec_ascending; sks->caseSensitivity = (int *) - nmem_malloc (stream, sizeof(*sks->caseSensitivity)); + nmem_malloc(stream, sizeof(*sks->caseSensitivity)); *sks->caseSensitivity = 0; sks->which = Z_SortKeySpec_null; sks->u.null = odr_nullval (); sort_sequence->specs[i] = sks; - - parms.rset_term = rset_term_create (termz, -1, rank_type, - zapt->term->which); - return rset_create (rset_kind_null, &parms); + return rsnull_create (NULL,key_it_ctrl); + /* FIXME - nmem?? */ } -/* pop - moved to xpath.c */ -#if 0 - -struct xpath_predicate { - int which; - union { -#define XPATH_PREDICATE_RELATION 1 - struct { - char *name; - char *op; - char *value; - } relation; -#define XPATH_PREDICATE_BOOLEAN 2 - struct { - const char *op; - struct xpath_predicate *left; - struct xpath_predicate *right; - } boolean; - } u; -}; - -struct xpath_location_step { - char *part; - struct xpath_predicate *predicate; -}; - -#endif static int parse_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt, oid_value attributeSet, @@ -2120,37 +1958,23 @@ static int parse_xpath(ZebraHandle zh, Z_AttributesPlusTerm *zapt, static RSET xpath_trunc(ZebraHandle zh, NMEM stream, int reg_type, const char *term, int use, - oid_value curAttributeSet) + oid_value curAttributeSet, NMEM rset_nmem) { RSET rset; struct grep_info grep_info; char term_dict[2048]; char ord_buf[32]; int prefix_len = 0; - int ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, use); + int ord = zebraExplain_lookup_attr_su(zh->reg->zei, curAttributeSet, use); int ord_len, i, r, max_pos; int term_type = Z_Term_characterString; const char *flags = "void"; if (grep_info_prepare (zh, 0 /* zapt */, &grep_info, '0', stream)) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (term, strlen(term), - flags, term_type); - parms.rset_term->nn = 0; - return rset_create (rset_kind_null, &parms); - } + return rsnull_create (rset_nmem,key_it_ctrl); if (ord < 0) - { - rset_null_parms parms; - - parms.rset_term = rset_term_create (term, strlen(term), - flags, term_type); - parms.rset_term->nn = 0; - return rset_create (rset_kind_null, &parms); - } + return rsnull_create (rset_nmem,key_it_ctrl); if (prefix_len) term_dict[prefix_len++] = '|'; else @@ -2166,16 +1990,17 @@ static RSET xpath_trunc(ZebraHandle zh, NMEM stream, term_dict[prefix_len++] = 1; term_dict[prefix_len++] = reg_type; - strcpy (term_dict+prefix_len, term); + strcpy(term_dict+prefix_len, term); grep_info.isam_p_indx = 0; - r = dict_lookup_grep (zh->reg->dict, term_dict, 0, + r = dict_lookup_grep(zh->reg->dict, term_dict, 0, &grep_info, &max_pos, 0, grep_handle); - yaz_log (LOG_LOG, "%s %d positions", term, + yaz_log (YLOG_DEBUG, "%s %d positions", term, grep_info.isam_p_indx); - rset = rset_trunc (zh, grep_info.isam_p_buf, + rset = rset_trunc(zh, grep_info.isam_p_buf, grep_info.isam_p_indx, term, strlen(term), - flags, 1, term_type); + flags, 1, term_type,rset_nmem, + key_it_ctrl, key_it_ctrl->scope); grep_info_delete (&grep_info); return rset; } @@ -2184,7 +2009,8 @@ static RSET rpn_search_xpath (ZebraHandle zh, oid_value attributeSet, int num_bases, char **basenames, NMEM stream, const char *rank_type, RSET rset, - int xpath_len, struct xpath_location_step *xpath) + int xpath_len, struct xpath_location_step *xpath, + NMEM rset_nmem) { oid_value curAttributeSet = attributeSet; int base_no; @@ -2193,10 +2019,10 @@ static RSET rpn_search_xpath (ZebraHandle zh, if (xpath_len < 0) return rset; - yaz_log (LOG_LOG, "len=%d", xpath_len); + yaz_log (YLOG_DEBUG, "xpath len=%d", xpath_len); for (i = 0; i none - a[@attr=value]/b[@other=othervalue] + a[@attr = value]/b[@other = othervalue] /e/@a val range(e/,range(@a,freetext(w,1015,val),@a),e/) /a/b val range(b/a/,freetext(w,1016,val),b/a/) /a/b/@c val range(b/a/,range(@c,freetext(w,1016,val),@c),b/a/) - /a/b[@c=y] val range(b/a/,freetext(w,1016,val),b/a/,@c=y) - /a[@c=y]/b val range(a/,range(b/a/,freetext(w,1016,val),b/a/),a/,@c=y) - /a[@c=x]/b[@c=y] range(a/,range(b/a/,freetext(w,1016,val),b/a/,@c=y),a/,@c=x) + /a/b[@c = y] val range(b/a/,freetext(w,1016,val),b/a/,@c = y) + /a[@c = y]/b val range(a/,range(b/a/,freetext(w,1016,val),b/a/),a/,@c = y) + /a[@c = x]/b[@c = y] range(a/,range(b/a/,freetext(w,1016,val),b/a/,@c = y),a/,@c = x) */ @@ -2238,7 +2064,6 @@ static RSET rpn_search_xpath (ZebraHandle zh, { char xpath_rev[128]; int i, len; - rset_between_parms parms; RSET rset_start_tag = 0, rset_end_tag = 0, rset_attr = 0; *xpath_rev = 0; @@ -2277,41 +2102,45 @@ static RSET rpn_search_xpath (ZebraHandle zh, xpath[level].predicate->which == XPATH_PREDICATE_RELATION && xpath[level].predicate->u.relation.name[0]) { - char predicate_str[128]; - - strcpy (predicate_str, - xpath[level].predicate->u.relation.name+1); + WRBUF wbuf = wrbuf_alloc(); + wrbuf_puts(wbuf, xpath[level].predicate->u.relation.name+1); if (xpath[level].predicate->u.relation.value) { - strcat (predicate_str, "="); - strcat (predicate_str, - xpath[level].predicate->u.relation.value); + const char *cp = xpath[level].predicate->u.relation.value; + wrbuf_putc(wbuf, '='); + + while (*cp) + { + if (strchr(REGEX_CHARS, *cp)) + wrbuf_putc(wbuf, '\\'); + wrbuf_putc(wbuf, *cp); + cp++; + } } - rset_attr = xpath_trunc ( - zh, stream, '0', predicate_str, 3, curAttributeSet); + wrbuf_puts(wbuf, ""); + rset_attr = xpath_trunc( + zh, stream, '0', wrbuf_buf(wbuf), 3, + curAttributeSet,rset_nmem); + wrbuf_free(wbuf, 1); } else { if (!first_path) continue; } - yaz_log (LOG_LOG, "xpath_rev (%d) = %s", level, xpath_rev); - if (strlen(xpath_rev)) - { - rset_start_tag = xpath_trunc(zh, stream, - '0', xpath_rev, 1, curAttributeSet); + yaz_log (log_level_rpn, "xpath_rev (%d) = %s", level, xpath_rev); + if (strlen(xpath_rev)) + { + rset_start_tag = xpath_trunc(zh, stream, '0', + xpath_rev, 1, curAttributeSet, rset_nmem); - rset_end_tag = xpath_trunc(zh, stream, - '0', xpath_rev, 2, curAttributeSet); - - parms.key_size = sizeof(struct it_key); - parms.cmp = key_compare_it; - parms.rset_l = rset_start_tag; - parms.rset_m = rset; - parms.rset_r = rset_end_tag; - parms.rset_attr = rset_attr; - parms.printer = key_print_it; - rset = rset_create (rset_kind_between, &parms); + rset_end_tag = xpath_trunc(zh, stream, '0', + xpath_rev, 2, curAttributeSet, rset_nmem); + + rset = rsbetween_create(rset_nmem, key_it_ctrl, + key_it_ctrl->scope, + rset_start_tag, rset, + rset_end_tag, rset_attr); } first_path = 0; } @@ -2324,8 +2153,9 @@ static RSET rpn_search_xpath (ZebraHandle zh, static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, oid_value attributeSet, NMEM stream, - Z_SortKeySpecList *sort_sequence, - int num_bases, char **basenames) + Z_SortKeySpecList *sort_sequence, + int num_bases, char **basenames, + NMEM rset_nmem) { unsigned reg_id; char *search_type = NULL; @@ -2338,20 +2168,25 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, int xpath_use = 0; struct xpath_location_step xpath[10]; + if (!log_level_set) + { + log_level_rpn = yaz_log_module_level("rpn"); + log_level_set = 1; + } zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type, - rank_type, &complete_flag, &sort_flag); + rank_type, &complete_flag, &sort_flag); - logf (LOG_DEBUG, "reg_id=%c", reg_id); - logf (LOG_DEBUG, "complete_flag=%d", complete_flag); - logf (LOG_DEBUG, "search_type=%s", search_type); - logf (LOG_DEBUG, "rank_type=%s", rank_type); + yaz_log(YLOG_DEBUG, "reg_id=%c", reg_id); + yaz_log(YLOG_DEBUG, "complete_flag=%d", complete_flag); + yaz_log(YLOG_DEBUG, "search_type=%s", search_type); + yaz_log(YLOG_DEBUG, "rank_type=%s", rank_type); if (zapt_term_to_utf8(zh, zapt, termz)) return 0; if (sort_flag) - return rpn_sort_spec (zh, zapt, attributeSet, stream, sort_sequence, - rank_type); + return rpn_sort_spec (zh, zapt, attributeSet, stream, sort_sequence, + rank_type); xpath_len = parse_xpath(zh, zapt, attributeSet, xpath, 10, stream); if (xpath_len >= 0) { @@ -2362,36 +2197,36 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!strcmp (search_type, "phrase")) { - rset = rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream, - reg_id, complete_flag, rank_type, + rset = rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream, + reg_id, complete_flag, rank_type, xpath_use, - num_bases, basenames); + num_bases, basenames, rset_nmem); } else if (!strcmp (search_type, "and-list")) { - rset = rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream, - reg_id, complete_flag, rank_type, + rset = rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream, + reg_id, complete_flag, rank_type, xpath_use, - num_bases, basenames); + num_bases, basenames, rset_nmem); } else if (!strcmp (search_type, "or-list")) { - rset = rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream, - reg_id, complete_flag, rank_type, + rset = rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream, + reg_id, complete_flag, rank_type, xpath_use, - num_bases, basenames); + num_bases, basenames, rset_nmem); } else if (!strcmp (search_type, "local")) { rset = rpn_search_APT_local (zh, zapt, termz, attributeSet, stream, - rank_type); + rank_type, rset_nmem); } else if (!strcmp (search_type, "numeric")) { - rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream, - reg_id, complete_flag, rank_type, + rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream, + reg_id, complete_flag, rank_type, xpath_use, - num_bases, basenames); + num_bases, basenames, rset_nmem); } else if (!strcmp (search_type, "always")) { @@ -2400,48 +2235,51 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, else zh->errCode = 118; return rpn_search_xpath (zh, attributeSet, num_bases, basenames, - stream, rank_type, rset, xpath_len, xpath); + stream, rank_type, rset, + xpath_len, xpath, rset_nmem); } static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, - oid_value attributeSet, NMEM stream, - Z_SortKeySpecList *sort_sequence, + oid_value attributeSet, + NMEM stream, NMEM rset_nmem, + Z_SortKeySpecList *sort_sequence, int num_bases, char **basenames) { RSET r = NULL; if (zs->which == Z_RPNStructure_complex) { Z_Operator *zop = zs->u.complex->roperator; - rset_bool_parms bool_parms; + RSET rsets[2]; /* l and r argument */ - bool_parms.rset_l = rpn_search_structure (zh, zs->u.complex->s1, - attributeSet, stream, - sort_sequence, - num_bases, basenames); - if (bool_parms.rset_l == NULL) + rsets[0]=rpn_search_structure (zh, zs->u.complex->s1, + attributeSet, stream, rset_nmem, + sort_sequence, + num_bases, basenames); + if (rsets[0] == NULL) return NULL; - bool_parms.rset_r = rpn_search_structure (zh, zs->u.complex->s2, - attributeSet, stream, - sort_sequence, - num_bases, basenames); - if (bool_parms.rset_r == NULL) + rsets[1]=rpn_search_structure (zh, zs->u.complex->s2, + attributeSet, stream, rset_nmem, + sort_sequence, + num_bases, basenames); + if (rsets[1] == NULL) { - rset_delete (bool_parms.rset_l); + rset_delete (rsets[0]); return NULL; } - bool_parms.key_size = sizeof(struct it_key); - bool_parms.cmp = key_compare_it; switch (zop->which) { case Z_Operator_and: - r = rset_create (rset_kind_and, &bool_parms); + r = rsmulti_and_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope, + 2, rsets); break; case Z_Operator_or: - r = rset_create (rset_kind_or, &bool_parms); + r = rsmulti_or_create(rset_nmem, key_it_ctrl, key_it_ctrl->scope, + 2, rsets); break; case Z_Operator_and_not: - r = rset_create (rset_kind_not, &bool_parms); + r = rsbool_create_not(rset_nmem,key_it_ctrl, key_it_ctrl->scope, + rsets[0],rsets[1]); break; case Z_Operator_prox: if (zop->u.prox->which != Z_ProximityOperator_known) @@ -2451,28 +2289,23 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, } if (*zop->u.prox->u.known != Z_ProxUnit_word) { - char *val = (char *) nmem_malloc (stream, 16); + char *val = (char *) nmem_malloc(stream, 16); zh->errCode = 132; zh->errString = val; sprintf (val, "%d", *zop->u.prox->u.known); return NULL; } - else - { - RSET rsets[2]; - - rsets[0] = bool_parms.rset_l; - rsets[1] = bool_parms.rset_r; - - r = rpn_prox (zh, rsets, 2, - *zop->u.prox->ordered, - (!zop->u.prox->exclusion ? 0 : - *zop->u.prox->exclusion), - *zop->u.prox->relationType, - *zop->u.prox->distance); - rset_delete (rsets[0]); - rset_delete (rsets[1]); - } + else + { + /* new / old prox */ + r = rsprox_create(rset_nmem,key_it_ctrl,key_it_ctrl->scope, + 2, rsets, + *zop->u.prox->ordered, + (!zop->u.prox->exclusion ? + 0 : *zop->u.prox->exclusion), + *zop->u.prox->relationType, + *zop->u.prox->distance ); + } break; default: zh->errCode = 110; @@ -2483,23 +2316,25 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, { if (zs->u.simple->which == Z_Operand_APT) { - logf (LOG_DEBUG, "rpn_search_APT"); + yaz_log(YLOG_DEBUG, "rpn_search_APT"); r = rpn_search_APT (zh, zs->u.simple->u.attributesPlusTerm, attributeSet, stream, sort_sequence, - num_bases, basenames); + num_bases, basenames,rset_nmem); } else if (zs->u.simple->which == Z_Operand_resultSetId) { - logf (LOG_DEBUG, "rpn_search_ref"); + yaz_log(YLOG_DEBUG, "rpn_search_ref"); r = resultSetRef (zh, zs->u.simple->u.resultSetId); - if (!r) - { - r = rset_create (rset_kind_null, NULL); - zh->errCode = 30; - zh->errString = - nmem_strdup (stream, zs->u.simple->u.resultSetId); - return 0; - } + if (!r) + { + r = rsnull_create (rset_nmem,key_it_ctrl); + zh->errCode = 30; + zh->errString = + nmem_strdup (stream, zs->u.simple->u.resultSetId); + return 0; + } + else + rset_dup(r); } else { @@ -2516,10 +2351,10 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, } -RSET rpn_search (ZebraHandle zh, NMEM nmem, - Z_RPNQuery *rpn, int num_bases, char **basenames, - const char *setname, - ZebraSet sset) +RSET rpn_search(ZebraHandle zh, NMEM nmem, NMEM rset_nmem, + Z_RPNQuery *rpn, int num_bases, char **basenames, + const char *setname, + ZebraSet sset) { RSET rset; oident *attrset; @@ -2532,45 +2367,46 @@ RSET rpn_search (ZebraHandle zh, NMEM nmem, zh->hits = 0; sort_sequence = (Z_SortKeySpecList *) - nmem_malloc (nmem, sizeof(*sort_sequence)); - sort_sequence->num_specs = 10; + nmem_malloc(nmem, sizeof(*sort_sequence)); + sort_sequence->num_specs = 10; /* FIXME - Hard-coded number */ sort_sequence->specs = (Z_SortKeySpec **) - nmem_malloc (nmem, sort_sequence->num_specs * - sizeof(*sort_sequence->specs)); + nmem_malloc(nmem, sort_sequence->num_specs * + sizeof(*sort_sequence->specs)); for (i = 0; inum_specs; i++) - sort_sequence->specs[i] = 0; + sort_sequence->specs[i] = 0; attrset = oid_getentbyoid (rpn->attributeSetId); attributeSet = attrset->value; rset = rpn_search_structure (zh, rpn->RPNStructure, attributeSet, - nmem, sort_sequence, num_bases, basenames); + nmem, rset_nmem, + sort_sequence, num_bases, basenames); if (!rset) - return 0; + return 0; if (zh->errCode) - logf (LOG_DEBUG, "search error: %d", zh->errCode); + yaz_log(YLOG_DEBUG, "search error: %d", zh->errCode); for (i = 0; sort_sequence->specs[i]; i++) - ; + ; sort_sequence->num_specs = i; if (!i) - resultSetRank (zh, sset, rset); + resultSetRank (zh, sset, rset, rset_nmem); else { - logf (LOG_DEBUG, "resultSetSortSingle in rpn_search"); - resultSetSortSingle (zh, nmem, sset, rset, - sort_sequence, &sort_status); - if (zh->errCode) - { - logf (LOG_DEBUG, "resultSetSortSingle status = %d", zh->errCode); - } + yaz_log(YLOG_DEBUG, "resultSetSortSingle in rpn_search"); + resultSetSortSingle (zh, nmem, sset, rset, + sort_sequence, &sort_status); + if (zh->errCode) + { + yaz_log(YLOG_DEBUG, "resultSetSortSingle status = %d", zh->errCode); + } } return rset; } struct scan_info_entry { char *term; - ISAMS_P isam_p; + ISAM_P isam_p; }; struct scan_info { @@ -2592,15 +2428,15 @@ static int scan_handle (char *name, const char *info, int pos, void *client) else idx = - pos - 1; scan_info->list[idx].term = (char *) - odr_malloc (scan_info->odr, strlen(name + len_prefix)+1); - strcpy (scan_info->list[idx].term, name + len_prefix); - assert (*info == sizeof(ISAMS_P)); - memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAMS_P)); + odr_malloc(scan_info->odr, strlen(name + len_prefix)+1); + strcpy(scan_info->list[idx].term, name + len_prefix); + assert (*info == sizeof(ISAM_P)); + memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAM_P)); return 0; } static void scan_term_untrans (ZebraHandle zh, NMEM stream, int reg_type, - char **dst, const char *src) + char **dst, const char *src) { char term_src[IT_MAX_WORD]; char term_dst[IT_MAX_WORD]; @@ -2622,45 +2458,44 @@ static void scan_term_untrans (ZebraHandle zh, NMEM stream, int reg_type, len = 0; else len = outbuf - term_dst; - *dst = nmem_malloc (stream, len + 1); + *dst = nmem_malloc(stream, len + 1); if (len > 0) memcpy (*dst, term_dst, len); (*dst)[len] = '\0'; } else - *dst = nmem_strdup (stream, term_src); + *dst = nmem_strdup(stream, term_src); } static void count_set (RSET r, int *count) { - int psysno = 0; + zint psysno = 0; int kno = 0; struct it_key key; RSFD rfd; - int term_index; - logf (LOG_DEBUG, "count_set"); + yaz_log(YLOG_DEBUG, "count_set"); *count = 0; rfd = rset_open (r, RSETF_READ); - while (rset_read (r, rfd, &key, &term_index)) + while (rset_read (rfd, &key,0 /* never mind terms */)) { - if (key.sysno != psysno) + if (key.mem[0] != psysno) { - psysno = key.sysno; + psysno = key.mem[0]; (*count)++; } kno++; } - rset_close (r, rfd); - logf (LOG_DEBUG, "%d keys, %d records", kno, *count); + rset_close (rfd); + yaz_log(YLOG_DEBUG, "%d keys, %d records", kno, *count); } -void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, - oid_value attributeset, - int num_bases, char **basenames, - int *position, int *num_entries, ZebraScanEntry **list, - int *is_partial) +void rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, + oid_value attributeset, + int num_bases, char **basenames, + int *position, int *num_entries, ZebraScanEntry **list, + int *is_partial, RSET limit_set, int return_zero) { int i; int pos = *position; @@ -2671,120 +2506,175 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, char termz[IT_MAX_WORD+20]; AttrType use; int use_value; + const char *use_string = 0; struct scan_info *scan_info_array; ZebraScanEntry *glist; int ords[32], ord_no = 0; int ptr[32]; + int bases_ok = 0; /* no of databases with OK attribute */ + int errCode = 0; /* err code (if any is not OK) */ + char *errString = 0; /* addinfo */ + unsigned reg_id; char *search_type = NULL; char rank_type[128]; int complete_flag; int sort_flag; + NMEM rset_nmem = NULL; + *list = 0; if (attributeset == VAL_NONE) attributeset = VAL_BIB1; - yaz_log (LOG_DEBUG, "position = %d, num = %d set=%d", + if (!limit_set) + { + AttrType termset; + int termset_value_numeric; + const char *termset_value_string; + attr_init (&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) + { + + sprintf (resname, "%d", termset_value_numeric); + termset_name = resname; + } + else + termset_name = termset_value_string; + + limit_set = resultSetRef (zh, termset_name); + } + } + + yaz_log (YLOG_DEBUG, "position = %d, num = %d set=%d", pos, num, attributeset); attr_init (&use, zapt, 1); - use_value = attr_find (&use, &attributeset); + use_value = attr_find_ex (&use, &attributeset, &use_string); if (zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type, - rank_type, &complete_flag, &sort_flag)) + rank_type, &complete_flag, &sort_flag)) { - *num_entries = 0; - zh->errCode = 113; - return ; + *num_entries = 0; + zh->errCode = 113; + return ; } - yaz_log (LOG_DEBUG, "use_value = %d", use_value); + yaz_log (YLOG_DEBUG, "use_value = %d", use_value); if (use_value == -1) use_value = 1016; for (base_no = 0; base_no < num_bases && ord_no < 32; base_no++) { - int r; - attent attp; - data1_local_attribute *local_attr; + data1_local_attribute *local_attr; + attent attp; + int ord; - if ((r=att_getentbyatt (zh, &attp, attributeset, use_value))) - { - logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d", - attributeset, use_value); - if (r == -1) - { - char val_str[32]; - sprintf (val_str, "%d", use_value); - zh->errCode = 114; - zh->errString = odr_strdup (stream, val_str); - } - else - zh->errCode = 121; - *num_entries = 0; - return; - } - if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no])) - { - zh->errString = basenames[base_no]; + if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no])) + { + zh->errString = basenames[base_no]; zh->errCode = 109; /* Database unavailable */ *num_entries = 0; return; - } - for (local_attr = attp.local_attributes; local_attr && ord_no < 32; - local_attr = local_attr->next) - { - int ord; + } - ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal, - local_attr->local); - if (ord > 0) - ords[ord_no++] = ord; - } + if (use_string && + (ord = zebraExplain_lookup_attr_str(zh->reg->zei, + use_string)) >= 0) + { + /* we have a match for a raw string attribute */ + if (ord > 0) + ords[ord_no++] = ord; + attp.local_attributes = 0; /* no more attributes */ + } + else + { + int r; + + if ((r = att_getentbyatt (zh, &attp, attributeset, use_value, + use_string))) + { + yaz_log(YLOG_DEBUG, "att_getentbyatt fail. set=%d use=%d", + attributeset, use_value); + if (r == -1) + { + char val_str[32]; + sprintf (val_str, "%d", use_value); + errCode = 114; + errString = odr_strdup (stream, val_str); + } + else + errCode = 121; + continue; + } + } + bases_ok++; + for (local_attr = attp.local_attributes; local_attr && ord_no < 32; + local_attr = local_attr->next) + { + ord = zebraExplain_lookup_attr_su(zh->reg->zei, + attp.attset_ordinal, + local_attr->local); + if (ord > 0) + ords[ord_no++] = ord; + } + } + if (!bases_ok && errCode) + { + zh->errCode = errCode; + zh->errString = errString; + *num_entries = 0; } if (ord_no == 0) { - *num_entries = 0; - zh->errCode = 113; - return; + *num_entries = 0; + return; } /* prepare dictionary scanning */ before = pos-1; after = 1+num-pos; scan_info_array = (struct scan_info *) - odr_malloc (stream, ord_no * sizeof(*scan_info_array)); + odr_malloc(stream, ord_no * sizeof(*scan_info_array)); for (i = 0; i < ord_no; i++) { int j, prefix_len = 0; int before_tmp = before, after_tmp = after; struct scan_info *scan_info = scan_info_array + i; - struct rpn_char_map_info rcmi; + struct rpn_char_map_info rcmi; - rpn_char_map_prepare (zh->reg, reg_id, &rcmi); + rpn_char_map_prepare (zh->reg, reg_id, &rcmi); scan_info->before = before; scan_info->after = after; scan_info->odr = stream; scan_info->list = (struct scan_info_entry *) - odr_malloc (stream, (before+after) * sizeof(*scan_info->list)); + odr_malloc(stream, (before+after) * sizeof(*scan_info->list)); for (j = 0; jlist[j].term = NULL; - prefix_len += key_SU_encode (ords[i], termz + prefix_len); + prefix_len += key_SU_encode (ords[i], termz + prefix_len); termz[prefix_len++] = reg_id; termz[prefix_len] = 0; - strcpy (scan_info->prefix, termz); + strcpy(scan_info->prefix, termz); - if (trans_scan_term (zh, zapt, termz+prefix_len, reg_id)) + if (trans_scan_term(zh, zapt, termz+prefix_len, reg_id)) return ; - - dict_scan (zh->reg->dict, termz, &before_tmp, &after_tmp, - scan_info, scan_handle); + + dict_scan(zh->reg->dict, termz, &before_tmp, &after_tmp, + scan_info, scan_handle); } glist = (ZebraScanEntry *) - odr_malloc (stream, (before+after)*sizeof(*glist)); + odr_malloc(stream, (before+after)*sizeof(*glist)); + + rset_nmem = nmem_create(); /* consider terms after main term */ for (i = 0; i < ord_no; i++) @@ -2801,7 +2691,7 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, for (j = 0; j < ord_no; j++) { if (ptr[j] < before+after && - (tst=scan_info_array[j].list[ptr[j]].term) && + (tst = scan_info_array[j].list[ptr[j]].term) && (!mterm || strcmp (tst, mterm) < 0)) { j0 = j; @@ -2811,107 +2701,118 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, if (j0 == -1) break; scan_term_untrans (zh, stream->mem, reg_id, - &glist[i+before].term, mterm); - rset = rset_trunc (zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1, - glist[i+before].term, strlen(glist[i+before].term), - NULL, 0, zapt->term->which); - + &glist[i+before].term, mterm); + rset = rset_trunc(zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1, + glist[i+before].term, strlen(glist[i+before].term), + NULL, 0, zapt->term->which, rset_nmem, + key_it_ctrl,key_it_ctrl->scope); ptr[j0]++; for (j = j0+1; jterm->which); - - bool_parms.key_size = sizeof(struct it_key); - bool_parms.cmp = key_compare_it; - bool_parms.rset_l = rset; - bool_parms.rset_r = rset2; - - rset = rset_create (rset_kind_or, &bool_parms); - + zapt->term->which,rset_nmem, + key_it_ctrl, key_it_ctrl->scope); + rset = rsmulti_or_create(rset_nmem, key_it_ctrl, + 2, key_it_ctrl->scope, rsets); ptr[j]++; } } - count_set (rset, &glist[i+before].occurrences); - rset_delete (rset); + if (limit_set) + { + RSET rsets[2]; + rsets[0] = rset; + rsets[1] = rset_dup(limit_set); + + rset = rsmulti_and_create(rset_nmem, key_it_ctrl, + key_it_ctrl->scope, 2, rsets); + } + count_set(rset, &glist[i+before].occurrences); + rset_delete(rset); } if (i < after) { - *num_entries -= (after-i); - *is_partial = 1; + *num_entries -= (after-i); + *is_partial = 1; } - + /* consider terms before main term */ for (i = 0; i 0)) - { - j0 = j; - mterm = tst; - } - } - if (j0 == -1) - break; - - scan_term_untrans (zh, stream->mem, reg_id, + int j, j0 = -1; + const char *mterm = NULL; + const char *tst; + RSET rset; + + for (j = 0; j 0)) + { + j0 = j; + mterm = tst; + } + } + if (j0 == -1) + break; + + scan_term_untrans (zh, stream->mem, reg_id, &glist[before-1-i].term, mterm); - - rset = rset_trunc - (zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1, - glist[before-1-i].term, strlen(glist[before-1-i].term), - NULL, 0, zapt->term->which); - - ptr[j0]++; - - for (j = j0+1; jterm->which); - - bool_parms.key_size = sizeof(struct it_key); - bool_parms.cmp = key_compare_it; - bool_parms.rset_l = rset; - bool_parms.rset_r = rset2; - - rset = rset_create (rset_kind_or, &bool_parms); - - ptr[j]++; - } - } - count_set (rset, &glist[before-1-i].occurrences); - rset_delete (rset); + + rset = rset_trunc + (zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1, + glist[before-1-i].term, strlen(glist[before-1-i].term), + NULL, 0, zapt->term->which,rset_nmem, + key_it_ctrl,key_it_ctrl->scope); + + ptr[j0]++; + + for (j = j0+1; jterm->which, rset_nmem, + key_it_ctrl, key_it_ctrl->scope); + rset = rsmulti_or_create(rset_nmem, key_it_ctrl, + 2, key_it_ctrl->scope, rsets); + + ptr[j]++; + } + } + if (limit_set) + { + RSET rsets[2]; + rsets[0] = rset; + rsets[1] = rset_dup(limit_set); + + rset = rsmulti_and_create(rset_nmem, key_it_ctrl, + key_it_ctrl->scope, 2, rsets); + } + count_set (rset, &glist[before-1-i].occurrences); + rset_delete (rset); } i = before-i; if (i) @@ -2920,11 +2821,13 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, *position -= i; *num_entries -= i; } + + nmem_destroy(rset_nmem); *list = glist + i; /* list is set to first 'real' entry */ - logf (LOG_DEBUG, "position = %d, num_entries = %d", - *position, *num_entries); + yaz_log(YLOG_DEBUG, "position = %d, num_entries = %d", + *position, *num_entries); if (zh->errCode) - logf (LOG_DEBUG, "scan error: %d", zh->errCode); + yaz_log(YLOG_DEBUG, "scan error: %d", zh->errCode); } - +