X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fzrpn.c;h=e87cf464d07f2051bb55b24711e85e6969cc9e1e;hb=28f3461a79a4569d736f4ffc66c6ae207e38c2ab;hp=a5617dccbfe3612968e066244c1ebab5c30b6acb;hpb=5d51844baabbb37b83afdd9b10db0a6c53cd1905;p=idzebra-moved-to-github.git diff --git a/index/zrpn.c b/index/zrpn.c index a5617dc..e87cf46 100644 --- a/index/zrpn.c +++ b/index/zrpn.c @@ -1,10 +1,69 @@ /* - * Copyright (C) 1995-1998, Index Data I/S + * Copyright (C) 1995-1999, Index Data * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * * $Log: zrpn.c,v $ - * Revision 1.80 1998-06-23 15:33:34 adam + * Revision 1.98 1999-11-30 13:48:04 adam + * Improved installation. Updated for inclusion of YAZ header files. + * + * Revision 1.97 1999/10/14 14:33:50 adam + * Added truncation 5=106. + * + * Revision 1.96 1999/09/23 10:05:05 adam + * Implemented structure=105 searching. + * + * Revision 1.95 1999/09/07 07:19:21 adam + * Work on character mapping. Implemented replace rules. + * + * Revision 1.94 1999/07/20 13:59:18 adam + * Fixed bug that occurred when phrases had 0 hits. + * + * Revision 1.93 1999/06/17 14:38:40 adam + * Bug fix: Scan SEGV'ed when getting unknown use attribute. + * + * Revision 1.92 1999/05/26 07:49:13 adam + * C++ compilation. + * + * Revision 1.91 1999/02/02 14:51:13 adam + * Updated WIN32 code specific sections. Changed header. + * + * Revision 1.90 1998/11/16 16:03:43 adam + * Moved loggin utilities to Yaz. Was implemented in file zlogs.c. + * + * Revision 1.89 1998/11/16 10:11:55 adam + * Added addtional info for error 114 - unsupported use attribute. + * + * Revision 1.88 1998/10/18 07:54:52 adam + * Additional info added for diagnostics 114 (Unsupported use attribute) and + * 121 (Unsupported attribute set). + * + * Revision 1.87 1998/09/28 11:19:12 adam + * Fix for Compiled ASN.1. + * + * Revision 1.86 1998/09/22 10:48:20 adam + * Minor changes in search API. + * + * Revision 1.85 1998/09/22 10:03:43 adam + * Changed result sets to be persistent in the sense that they can + * be re-searched if needed. + * Fixed memory leak in rsm_or. + * + * Revision 1.84 1998/09/18 12:41:00 adam + * Fixed bug with numerical relations. + * + * Revision 1.83 1998/09/02 13:53:19 adam + * Extra parameter decode added to search routines to implement + * persistent queries. + * + * Revision 1.82 1998/06/26 11:16:40 quinn + * Added support (un-optimised) for left and left/right truncation + * + * Revision 1.81 1998/06/24 12:16:14 adam + * Support for relations on text operands. Open range support in + * DFA module (i.e. [-j], [g-]). + * + * Revision 1.80 1998/06/23 15:33:34 adam * Added feature to specify sort criteria in query (type 7 specifies * sort flags). * @@ -281,7 +340,7 @@ */ #include #include -#ifdef WINDOWS +#ifdef WIN32 #include #else #include @@ -302,7 +361,7 @@ struct rpn_char_map_info { static const char **rpn_char_map_handler (void *vp, const char **from, int len) { - struct rpn_char_map_info *p = vp; + struct rpn_char_map_info *p = (struct rpn_char_map_info *) vp; return zebra_maps_input (p->zm, p->reg_type, from, len); } @@ -392,7 +451,7 @@ struct grep_info { #ifdef TERM_COUNT int *term_no; #endif - ISAM_P *isam_p_buf; + ISAMS_P *isam_p_buf; int isam_p_size; int isam_p_indx; ZebraHandle zh; @@ -405,8 +464,11 @@ static void term_untrans (ZebraHandle zh, int reg_type, while (*src) { const char *cp = zebra_maps_output (zh->zebra_maps, reg_type, &src); - while (*cp) - *dst++ = *cp++; + if (!cp) + *dst++ = *src++; + else + while (*cp) + *dst++ = *cp++; } *dst = '\0'; } @@ -416,13 +478,13 @@ static void add_isam_p (const char *name, const char *info, { if (p->isam_p_indx == p->isam_p_size) { - ISAM_P *new_isam_p_buf; + ISAMS_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 = xmalloc (sizeof(*new_isam_p_buf) * - p->isam_p_size); + new_isam_p_buf = (ISAMS_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, @@ -432,8 +494,8 @@ static void add_isam_p (const char *name, const char *info, p->isam_p_buf = new_isam_p_buf; #ifdef TERM_COUNT - new_term_no = 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, @@ -455,7 +517,7 @@ static void add_isam_p (const char *name, const char *info, static int grep_handle (char *name, const char *info, void *p) { - add_isam_p (name, info, p); + add_isam_p (name, info, (struct grep_info *) p); return 0; } @@ -482,6 +544,7 @@ static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src, return *s0; } +/* 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) @@ -514,6 +577,7 @@ 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) @@ -555,7 +619,7 @@ 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) @@ -604,6 +668,7 @@ static int term_103 (ZebraMaps zebra_maps, int reg_type, const char **src, return i; } +/* 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) { @@ -611,6 +676,106 @@ static int term_102 (ZebraMaps zebra_maps, int reg_type, const char **src, 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, "#!", "#!")) + return 0; + s0 = *src; + while (*s0) + { + if (*s0 == '#') + { + dst[i++] = '.'; + dst[i++] = '*'; + dst_term[j++] = *s0++; + } + else if (*s0 == '!') + { + dst[i++] = '.'; + dst_term[j++] = *s0++; + } + { + s1 = s0; + map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); + if (space_split && **map == *CHR_SPACE) + break; + while (s1 < s0) + { + if (!isalnum (*s1)) + dst[i++] = '\\'; + dst_term[j++] = *s1; + dst[i++] = *s1++; + } + } + } + dst[i] = '\0'; + dst_term[j++] = '\0'; + *src = s0; + return i; +} + +/* 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 *s0, *s1; + const char **map; + int i = 0; + int j = 0; + + if (!term_pre (zebra_maps, reg_type, src, "*!", "*!")) + return 0; + s0 = *src; + while (*s0) + { + if (*s0 == '*') + { + dst[i++] = '.'; + dst[i++] = '+'; + dst_term[j++] = *s0++; + } + else if (*s0 == '!') + { + dst[i++] = '.'; + dst_term[j++] = *s0++; + } + { + s1 = s0; + map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0)); + if (space_split && **map == *CHR_SPACE) + break; + while (s1 < s0) + { + if (!isalnum (*s1)) + dst[i++] = '\\'; + dst_term[j++] = *s1; + dst[i++] = *s1++; + } + } + } + if (right_truncate) + { + dst[i++] = '.'; + dst[i++] = '*'; + } + dst[i] = '\0'; + + dst_term[j++] = '\0'; + *src = s0; + return i; +} + + /* gen_regular_rel - generate regular expression from relation * val: border value (inclusive) * islt: 1 if <=; 0 if >=. @@ -640,7 +805,7 @@ static void gen_regular_rel (char *dst, int val, int islt) } else { - strcpy (dst, "((-"); + strcpy (dst, "(-("); islt = 0; } val = -val; @@ -708,11 +873,14 @@ static void gen_regular_rel (char *dst, int val, int islt) dst[dst_p] = '\0'; if (islt) { - for (i=1; i abc ([b-].*|a[c-].*|ab[d-].*|abc.+) + * ([^-a].*|a[^-b].*ab[^-c].*|abc.+) + * >= abc ([b-].*|a[c-].*|ab[c-].*) + * ([^-a].*|a[^-b].*|ab[c-].*) + * < abc ([-0].*|a[-a].*|ab[-b].*) + * ([^a-].*|a[^b-].*|ab[^c-].*) + * <= abc ([-0].*|a[-a].*|ab[-b].*|abc) + * ([^a-].*|a[^b-].*|ab[^c-].*|abc) + */ static int string_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, + int reg_type, int space_split, char *term_dst) { AttrType relation; int relation_value; - int term_value; - int r; + int i; char *term_tmp = term_dict + strlen(term_dict); + char term_component[256]; attr_init (&relation, zapt, 2); relation_value = attr_find (&relation, NULL); @@ -742,63 +923,153 @@ static int string_relation (ZebraHandle zh, Z_AttributesPlusTerm *zapt, switch (relation_value) { case 1: - if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1, - term_dst)) + if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_component, + space_split, term_dst)) return 0; - term_value = atoi (term_tmp); logf (LOG_DEBUG, "Relation <"); - gen_regular_rel (term_tmp, term_value-1, 1); + + *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++ = '['; + + *term_tmp++ = '^'; + string_rel_add_char (&term_tmp, term_component, &i); + *term_tmp++ = '-'; + + *term_tmp++ = ']'; + *term_tmp++ = '.'; + *term_tmp++ = '*'; + } + *term_tmp++ = ')'; + *term_tmp = '\0'; break; case 2: - if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1, - term_dst)) + if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_component, + space_split, term_dst)) return 0; - term_value = atoi (term_tmp); logf (LOG_DEBUG, "Relation <="); - gen_regular_rel (term_tmp, term_value, 1); - break; - case 4: - if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1, - term_dst)) - return 0; - term_value = atoi (term_tmp); - logf (LOG_DEBUG, "Relation >="); - gen_regular_rel (term_tmp, term_value, 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++ = '^'; + string_rel_add_char (&term_tmp, term_component, &i); + *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'; break; case 5: - if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1, - term_dst)) + if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_component, + space_split, term_dst)) return 0; - term_value = atoi (term_tmp); logf (LOG_DEBUG, "Relation >"); - gen_regular_rel (term_tmp, term_value+1, 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); + + *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'; + break; + case 4: + if (!term_100 (zh->zebra_maps, reg_type, 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; + + 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'; break; case 3: default: logf (LOG_DEBUG, "Relation ="); - *term_tmp = '('; - if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp+1, 1, - term_dst)) + if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_component, + space_split, term_dst)) return 0; + strcat (term_tmp, "("); + strcat (term_tmp, term_component); strcat (term_tmp, ")"); } - logf (LOG_DEBUG, "dict_lookup_grep: %s", term_tmp); - r = dict_lookup_grep (zh->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); return 1; } static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char **term_sub, - oid_value attributeSet, struct grep_info *grep_info, + oid_value attributeSet, NMEM stream, + struct grep_info *grep_info, int reg_type, int complete_flag, int num_bases, char **basenames, char *term_dst) { - char term_dict[2*IT_MAX_WORD+2]; + char term_dict[2*IT_MAX_WORD+4000]; int j, r, base_no; AttrType truncation; int truncation_value; @@ -832,9 +1103,25 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, logf (LOG_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->errCode = 114; + zh->errString = nmem_strdup (stream, val_str); + } else + { + int oid[OID_SIZE]; + struct oident oident; + + oident.proto = PROTO_Z3950; + oident.oclass = CLASS_ATTSET; + oident.value = curAttributeSet; + oid_ent_to_oid (&oident, oid); + zh->errCode = 121; + zh->errString = nmem_strdup (stream, oident.desc); + } return -1; } if (zebraExplain_curDatabase (zh->zei, basenames[base_no])) @@ -868,7 +1155,10 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } if (!prefix_len) { - zh->errCode = 114; + char val_str[32]; + sprintf (val_str, "%d", use_value); + zh->errCode = 114; + zh->errString = nmem_strdup (stream, val_str); return -1; } term_dict[prefix_len++] = ')'; @@ -882,20 +1172,14 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, case -1: /* not specified */ case 100: /* do not truncate */ if (!string_relation (zh, zapt, &termp, term_dict, - attributeSet, grep_info, &max_pos, - reg_type, term_dst)) + attributeSet, + reg_type, space_split, term_dst)) return 0; -#if 0 - term_dict[j++] = '('; - if (!term_100 (zh->zebra_maps, reg_type, - &termp, term_dict + j, space_split, term_dst)) - return 0; - strcat (term_dict, ")"); - r = dict_lookup_grep (zh->dict, term_dict, 0, grep_info, - &max_pos, 0, grep_handle); + logf (LOG_DEBUG, "dict_lookup_grep: %s", term_dict+prefix_len); + r = dict_lookup_grep (zh->dict, term_dict, 0, grep_info, &max_pos, + 0, grep_handle); if (r) - logf (LOG_WARN, "dict_lookup_grep err, trunc=none:%d", r); -#endif + logf (LOG_WARN, "dict_lookup_grep fail, rel=gt: %d", r); break; case 1: /* right truncation */ term_dict[j++] = '('; @@ -906,8 +1190,24 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, dict_lookup_grep (zh->dict, term_dict, 0, grep_info, &max_pos, 0, grep_handle); break; - case 2: /* left truncation */ + case 2: /* keft truncation */ + term_dict[j++] = '('; term_dict[j++] = '.'; term_dict[j++] = '*'; + if (!term_100 (zh->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat (term_dict, ")"); + dict_lookup_grep (zh->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->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat (term_dict, ".*)"); + dict_lookup_grep (zh->dict, term_dict, 0, grep_info, + &max_pos, 0, grep_handle); + break; zh->errCode = 120; return -1; case 101: /* process # in term */ @@ -948,6 +1248,39 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, 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->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst)) + return 0; + strcat (term_dict, ")"); + r = dict_lookup_grep (zh->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->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst, 1)) + return 0; + strcat (term_dict, ")"); + r = dict_lookup_grep (zh->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->zebra_maps, reg_type, + &termp, term_dict + j, space_split, term_dst, 0)) + return 0; + strcat (term_dict, ")"); + r = dict_lookup_grep (zh->dict, term_dict, 0, grep_info, + &max_pos, 0, grep_handle); + if (r) + logf (LOG_WARN, "dict_lookup_grep err, trunc=*/!: %d", r); + break; } } *term_sub = termp; @@ -1103,31 +1436,38 @@ static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no) int term_index; const char *flags = NULL; - rsfd = xmalloc (sizeof(*rsfd)*rset_no); - more = xmalloc (sizeof(*more)*rset_no); - buf = xmalloc (sizeof(*buf)*rset_no); + 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 (length_prox_term) - prox_term[length_prox_term++] = ' '; - strcpy (prox_term + length_prox_term, term); - length_prox_term += lterm; + 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; } } + for (i = 0; inn = 0; result = rset_create (rset_kind_null, &parms); } @@ -1147,7 +1488,8 @@ static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no) rset_temp_parms parms; RSFD rsfd_result; - parms.rset_term = rset_term_create (prox_term, -1, flags); + parms.rset_term = rset_term_create (prox_term, length_prox_term, + flags); parms.rset_term->nn = min_nn; parms.key_size = sizeof (struct it_key); parms.temp_path = res_get (zh->res, "setTmpDir"); @@ -1208,19 +1550,66 @@ static RSET rpn_prox (ZebraHandle zh, RSET *rset, int rset_no) return result; } + +char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, + const char *termz, NMEM stream, unsigned reg_id) +{ + WRBUF wrbuf = 0; + AttrType truncation; + int truncation_value; + char *ex_list = 0; + + attr_init (&truncation, zapt, 5); + truncation_value = attr_find (&truncation, NULL); + + switch (truncation_value) + { + default: + ex_list = ""; + break; + case 101: + ex_list = "#"; + break; + case 102: + case 103: + ex_list = 0; + break; + case 104: + ex_list = "!#"; + break; + case 105: + ex_list = "!*"; + break; + } + if (ex_list) + wrbuf = zebra_replace(zh->zebra_maps, reg_id, ex_list, + termz, strlen(termz)); + if (!wrbuf) + 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; + } +} + static RSET rpn_search_APT_phrase (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz, + const char *termz_org, oid_value attributeSet, + NMEM stream, int reg_type, int complete_flag, const char *rank_type, int num_bases, char **basenames) { char term_dst[IT_MAX_WORD+1]; - const char *termp = termz; RSET rset[60], result; int i, r, rset_no = 0; struct grep_info grep_info; + char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); + const char *termp = termz; #ifdef TERM_COUNT grep_info.term_no = 0; @@ -1234,7 +1623,7 @@ static RSET rpn_search_APT_phrase (ZebraHandle zh, { logf (LOG_DEBUG, "APT_phrase termp=%s", termp); grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, &grep_info, + r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst); if (r < 1) @@ -1244,7 +1633,7 @@ static RSET rpn_search_APT_phrase (ZebraHandle zh, grep_info.isam_p_indx, term_dst, strlen(term_dst), rank_type); assert (rset[rset_no]); - if (++rset_no >= sizeof(rset)/sizeof(*rset)) + if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } #ifdef TERM_COUNT @@ -1268,18 +1657,19 @@ static RSET rpn_search_APT_phrase (ZebraHandle zh, static RSET rpn_search_APT_or_list (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz, + const char *termz_org, oid_value attributeSet, + NMEM stream, int reg_type, int complete_flag, const char *rank_type, int num_bases, char **basenames) { char term_dst[IT_MAX_WORD+1]; - const char *termp = termz; RSET rset[60], result; int i, r, rset_no = 0; struct grep_info grep_info; - + char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); + const char *termp = termz; #ifdef TERM_COUNT grep_info.term_no = 0; #endif @@ -1292,7 +1682,7 @@ static RSET rpn_search_APT_or_list (ZebraHandle zh, { logf (LOG_DEBUG, "APT_or_list termp=%s", termp); grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, &grep_info, + r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst); if (r < 1) @@ -1302,7 +1692,7 @@ static RSET rpn_search_APT_or_list (ZebraHandle zh, grep_info.isam_p_indx, term_dst, strlen(term_dst), rank_type); assert (rset[rset_no]); - if (++rset_no >= sizeof(rset)/sizeof(*rset)) + if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } #ifdef TERM_COUNT @@ -1332,17 +1722,19 @@ static RSET rpn_search_APT_or_list (ZebraHandle zh, static RSET rpn_search_APT_and_list (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char *termz, + const char *termz_org, oid_value attributeSet, + NMEM stream, int reg_type, int complete_flag, const char *rank_type, int num_bases, char **basenames) { char term_dst[IT_MAX_WORD+1]; - const char *termp = termz; RSET rset[60], result; int i, r, rset_no = 0; struct grep_info grep_info; + char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); + const char *termp = termz; #ifdef TERM_COUNT grep_info.term_no = 0; @@ -1356,7 +1748,7 @@ static RSET rpn_search_APT_and_list (ZebraHandle zh, { logf (LOG_DEBUG, "APT_and_list termp=%s", termp); grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, &grep_info, + r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, reg_type, complete_flag, num_bases, basenames, term_dst); if (r < 1) @@ -1366,7 +1758,7 @@ static RSET rpn_search_APT_and_list (ZebraHandle zh, grep_info.isam_p_indx, term_dst, strlen(term_dst), rank_type); assert (rset[rset_no]); - if (++rset_no >= sizeof(rset)/sizeof(*rset)) + if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } #ifdef TERM_COUNT @@ -1543,6 +1935,7 @@ 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 num_bases, char **basenames) @@ -1575,7 +1968,7 @@ static RSET rpn_search_APT_numeric (ZebraHandle zh, grep_info.isam_p_indx, term_dst, strlen(term_dst), rank_type); assert (rset[rset_no]); - if (++rset_no >= sizeof(rset)/sizeof(*rset)) + if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } #ifdef TERM_COUNT @@ -1606,6 +1999,7 @@ static RSET rpn_search_APT_numeric (ZebraHandle zh, static RSET rpn_search_APT_local (ZebraHandle zh, Z_AttributesPlusTerm *zapt, const char *termz, oid_value attributeSet, + NMEM stream, const char *rank_type) { RSET result; @@ -1629,7 +2023,7 @@ static RSET rpn_search_APT_local (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - oid_value attributeSet, ODR stream, + oid_value attributeSet, NMEM stream, Z_SortKeySpecList *sort_sequence, const char *rank_type) { @@ -1654,15 +2048,17 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!sort_sequence->specs) { sort_sequence->num_specs = 10; - sort_sequence->specs = odr_malloc (stream, sort_sequence->num_specs * - sizeof(*sort_sequence->specs)); + 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; else - i = atoi_n (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; @@ -1672,29 +2068,35 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!oid_ent_to_oid (&oe, oid)) return 0; - sks = odr_malloc (stream, sizeof(*sks)); - sks->sortElement = odr_malloc (stream, sizeof(*sks->sortElement)); + sks = (Z_SortKeySpec *) nmem_malloc (stream, sizeof(*sks)); + sks->sortElement = (Z_SortElement *) + nmem_malloc (stream, sizeof(*sks->sortElement)); sks->sortElement->which = Z_SortElement_generic; - sk = sks->sortElement->u.generic = odr_malloc (stream, sizeof(*sk)); + sk = sks->sortElement->u.generic = (Z_SortKey *) + nmem_malloc (stream, sizeof(*sk)); sk->which = Z_SortKey_sortAttributes; - sk->u.sortAttributes = odr_malloc (stream, sizeof(*sk->u.sortAttributes)); + sk->u.sortAttributes = (Z_SortAttributes *) + nmem_malloc (stream, sizeof(*sk->u.sortAttributes)); sk->u.sortAttributes->id = oid; - sk->u.sortAttributes->list = - odr_malloc (stream, sizeof(*sk->u.sortAttributes->list)); + sk->u.sortAttributes->list = (Z_AttributeList *) + nmem_malloc (stream, sizeof(*sk->u.sortAttributes->list)); sk->u.sortAttributes->list->num_attributes = 1; - sk->u.sortAttributes->list->attributes = - odr_malloc (stream, sizeof(*sk->u.sortAttributes->list->attributes)); - ae = *sk->u.sortAttributes->list->attributes = - odr_malloc (stream, sizeof(**sk->u.sortAttributes->list->attributes)); + sk->u.sortAttributes->list->attributes = (Z_AttributeElement **) + 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)); ae->attributeSet = 0; - ae->attributeType = odr_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 = odr_malloc (stream, sizeof(*ae->value.numeric)); + ae->value.numeric = (int *) + nmem_malloc (stream, sizeof(*ae->value.numeric)); *ae->value.numeric = use_value; - sks->sortRelation = odr_malloc (stream, sizeof(*sks->sortRelation)); + sks->sortRelation = (int *) + nmem_malloc (stream, sizeof(*sks->sortRelation)); if (sort_relation_value == 1) *sks->sortRelation = Z_SortRelation_ascending; else if (sort_relation_value == 2) @@ -1702,10 +2104,16 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, else *sks->sortRelation = Z_SortRelation_ascending; - sks->caseSensitivity = odr_malloc (stream, sizeof(*sks->caseSensitivity)); + sks->caseSensitivity = (int *) + nmem_malloc (stream, sizeof(*sks->caseSensitivity)); *sks->caseSensitivity = 0; +#ifdef ASN_COMPILED + sks->which = Z_SortKeySpec_null; + sks->u.null = odr_nullval (); +#else sks->missingValueAction = 0; +#endif sort_sequence->specs[i] = sks; @@ -1715,7 +2123,7 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - oid_value attributeSet, ODR stream, + oid_value attributeSet, NMEM stream, Z_SortKeySpecList *sort_sequence, int num_bases, char **basenames) { @@ -1747,30 +2155,30 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!strcmp (search_type, "phrase")) { - return rpn_search_APT_phrase (zh, zapt, termz, attributeSet, + return rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "and-list")) { - return rpn_search_APT_and_list (zh, zapt, termz, attributeSet, + return rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "or-list")) { - return rpn_search_APT_or_list (zh, zapt, termz, attributeSet, + return rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "local")) { - return rpn_search_APT_local (zh, zapt, termz, attributeSet, + return rpn_search_APT_local (zh, zapt, termz, attributeSet, stream, rank_type); } else if (!strcmp (search_type, "numeric")) { - return rpn_search_APT_numeric (zh, zapt, termz, attributeSet, + return rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } @@ -1779,7 +2187,7 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, - oid_value attributeSet, ODR stream, + oid_value attributeSet, NMEM stream, Z_SortKeySpecList *sort_sequence, int num_bases, char **basenames) { @@ -1836,7 +2244,7 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, #ifdef ASN_COMPILED if (*zop->u.prox->u.known != Z_ProxUnit_word) { - char *val = odr_malloc (stream, 16); + char *val = (char *) nmem_malloc (stream, 16); zh->errCode = 132; zh->errString = val; sprintf (val, "%d", *zop->u.prox->u.known); @@ -1845,7 +2253,7 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, #else if (*zop->u.prox->proximityUnitCode != Z_ProxUnit_word) { - char *val = odr_malloc (stream, 16); + char *val = (char *) nmem_malloc (stream, 16); zh->errCode = 132; zh->errString = val; sprintf (val, "%d", *zop->u.prox->proximityUnitCode); @@ -1894,9 +2302,11 @@ static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, return r; } -void rpn_search (ZebraHandle zh, ODR stream, + +RSET rpn_search (ZebraHandle zh, NMEM nmem, Z_RPNQuery *rpn, int num_bases, char **basenames, - const char *setname) + const char *setname, + ZebraSet sset) { RSET rset; oident *attrset; @@ -1904,42 +2314,50 @@ void rpn_search (ZebraHandle zh, ODR stream, Z_SortKeySpecList *sort_sequence; int sort_status, i; - zlog_rpn (rpn); - zh->errCode = 0; zh->errString = NULL; zh->hits = 0; - sort_sequence = odr_malloc (stream, sizeof(*sort_sequence)); + sort_sequence = (Z_SortKeySpecList *) + nmem_malloc (nmem, sizeof(*sort_sequence)); sort_sequence->num_specs = 10; - sort_sequence->specs = odr_malloc (stream, sort_sequence->num_specs * - sizeof(*sort_sequence->specs)); + sort_sequence->specs = (Z_SortKeySpec **) + nmem_malloc (nmem, sort_sequence->num_specs * + sizeof(*sort_sequence->specs)); for (i = 0; inum_specs; i++) sort_sequence->specs[i] = 0; attrset = oid_getentbyoid (rpn->attributeSetId); attributeSet = attrset->value; - rset = rpn_search_structure (zh, rpn->RPNStructure, attributeSet, stream, - sort_sequence, - num_bases, basenames); + rset = rpn_search_structure (zh, rpn->RPNStructure, attributeSet, + nmem, sort_sequence, num_bases, basenames); if (!rset) - return; + return 0; - resultSetAdd (zh, setname, 1, rset, &zh->hits); if (zh->errCode) logf (LOG_DEBUG, "search error: %d", zh->errCode); - + for (i = 0; sort_sequence->specs[i]; i++) ; sort_sequence->num_specs = i; - if (i) - resultSetSort (zh, stream, 1, &setname, setname, sort_sequence, - &sort_status); + if (!i) + resultSetRank (zh, sset, rset); + 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); + } + } + return rset; } struct scan_info_entry { char *term; - ISAM_P isam_p; + ISAMS_P isam_p; }; struct scan_info { @@ -1952,31 +2370,30 @@ struct scan_info { static int scan_handle (char *name, const char *info, int pos, void *client) { int len_prefix, idx; - struct scan_info *scan_info = client; + struct scan_info *scan_info = (struct scan_info *) client; len_prefix = strlen(scan_info->prefix); if (memcmp (name, scan_info->prefix, len_prefix)) return 1; - if (pos > 0) - idx = scan_info->after - pos + scan_info->before; + if (pos > 0) idx = scan_info->after - pos + scan_info->before; else idx = - pos - 1; - scan_info->list[idx].term = odr_malloc (scan_info->odr, - strlen(name + len_prefix)+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(ISAM_P)); - memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAM_P)); + assert (*info == sizeof(ISAMS_P)); + memcpy (&scan_info->list[idx].isam_p, info+1, sizeof(ISAMS_P)); return 0; } -static void scan_term_untrans (ZebraHandle zh, ODR stream, int reg_type, +static void scan_term_untrans (ZebraHandle zh, NMEM stream, int reg_type, char **dst, const char *src) { char term_dst[1024]; term_untrans (zh, reg_type, term_dst, src); - *dst = odr_malloc (stream, strlen(term_dst)+1); + *dst = (char *) nmem_malloc (stream, strlen(term_dst)+1); strcpy (*dst, term_dst); } @@ -2030,11 +2447,11 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, char *rank_type = NULL; int complete_flag; int sort_flag; + *list = 0; if (attributeset == VAL_NONE) attributeset = VAL_BIB1; - zlog_scan (zapt, attributeset); logf (LOG_DEBUG, "position = %d, num = %d", pos, num); attr_init (&use, zapt, 1); @@ -2063,6 +2480,8 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, zh->errCode = 114; else zh->errCode = 121; + *num_entries = 0; + return; } if (zebraExplain_curDatabase (zh->zei, basenames[base_no])) { @@ -2089,7 +2508,8 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, /* prepare dictionary scanning */ before = pos-1; after = 1+num-pos; - scan_info_array = odr_malloc (stream, ord_no * sizeof(*scan_info_array)); + scan_info_array = (struct scan_info *) + odr_malloc (stream, ord_no * sizeof(*scan_info_array)); for (i = 0; i < ord_no; i++) { int j, prefix_len = 0; @@ -2103,8 +2523,8 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, scan_info->after = after; scan_info->odr = stream; - scan_info->list = odr_malloc (stream, (before+after)* - sizeof(*scan_info->list)); + scan_info->list = (struct scan_info_entry *) + odr_malloc (stream, (before+after) * sizeof(*scan_info->list)); for (j = 0; jlist[j].term = NULL; @@ -2118,7 +2538,8 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, dict_scan (zh->dict, termz, &before_tmp, &after_tmp, scan_info, scan_handle); } - glist = odr_malloc (stream, (before+after)*sizeof(*glist)); + glist = (ZebraScanEntry *) + odr_malloc (stream, (before+after)*sizeof(*glist)); /* consider terms after main term */ for (i = 0; i < ord_no; i++) @@ -2144,7 +2565,7 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, } if (j0 == -1) break; - scan_term_untrans (zh, stream, reg_id, + 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), @@ -2208,7 +2629,7 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, if (j0 == -1) break; - scan_term_untrans (zh, stream, reg_id, + scan_term_untrans (zh, stream->mem, reg_id, &glist[before-1-i].term, mterm); rset = rset_trunc