X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=dict%2Flookgrep.c;h=03193ed839155bd65f7e8288525e842b2e1c9fd3;hb=07daef4fb241a9106fb995bb87a823708babcb10;hp=39d2b3f4dcfe0e28a28fd43a4be71dd51955bc3d;hpb=55a5cde7eb23fb9aa5a8386d34bb1b6e131c19d8;p=idzebra-moved-to-github.git diff --git a/dict/lookgrep.c b/dict/lookgrep.c index 39d2b3f..03193ed 100644 --- a/dict/lookgrep.c +++ b/dict/lookgrep.c @@ -1,82 +1,26 @@ -/* - * Copyright (C) 1994-1998, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: lookgrep.c,v $ - * Revision 1.21 1998-06-24 12:16:12 adam - * Support for relations on text operands. Open range support in - * DFA module (i.e. [-j], [g-]). - * - * Revision 1.20 1997/10/27 14:33:03 adam - * Moved towards generic character mapping depending on "structure" - * field in abstract syntax file. Fixed a few memory leaks. Fixed - * bug with negative integers when doing searches with relational - * operators. - * - * Revision 1.19 1997/09/18 08:59:18 adam - * Extra generic handle for the character mapping routines. - * - * Revision 1.18 1997/09/05 15:29:58 adam - * Changed prototype for chr_map_input - added const. - * Added support for C++, headers uses extern "C" for public definitions. - * - * Revision 1.17 1996/06/04 10:20:06 adam - * Added support for character mapping. - * - * Revision 1.16 1996/05/24 14:46:04 adam - * Added dict_grep_cmap function to define user-mapping in grep lookups. - * - * Revision 1.15 1996/03/20 09:35:18 adam - * Function dict_lookup_grep got extra parameter, init_pos, which marks - * from which position in pattern approximate pattern matching should occur. - * - * Revision 1.14 1996/02/02 13:43:51 adam - * The public functions simply use char instead of Dict_char to represent - * search strings. Dict_char is used internally only. - * - * Revision 1.13 1996/01/08 09:09:30 adam - * Function dfa_parse got 'const' string argument. - * - * Revision 1.12 1995/12/11 09:04:48 adam - * Bug fix: the lookup/scan/lookgrep didn't handle empty dictionary. - * - * Revision 1.11 1995/12/06 14:43:02 adam - * New function: dict_delete. - * - * Revision 1.10 1995/11/16 17:00:44 adam - * Changed stupid log. - * - * Revision 1.9 1995/10/27 13:58:09 adam - * Makes 'Database unavailable' diagnostic. - * - * Revision 1.8 1995/10/19 14:57:21 adam - * New feature: grep lookup saves length of longest prefix match. - * - * Revision 1.7 1995/10/17 18:01:22 adam - * Userfunc may return non-zero in which case the the grepping stops - * immediately. - * - * Revision 1.6 1995/10/09 16:18:32 adam - * Function dict_lookup_grep got extra client data parameter. - * - * Revision 1.5 1995/09/14 11:52:59 adam - * Grep handle function parameter info is const now. - * - * Revision 1.4 1995/01/24 16:01:02 adam - * Added -ansi to CFLAGS. - * Use new API of dfa module. - * - * Revision 1.3 1994/10/05 12:16:50 adam - * Pagesize is a resource now. - * - * Revision 1.2 1994/10/04 12:08:07 adam - * Some bug fixes and some optimizations. - * - * Revision 1.1 1994/10/03 17:23:04 adam - * First version of dictionary lookup with regular expressions and errors. - * - */ +/* $Id: lookgrep.c,v 1.30 2005-11-01 10:16:08 adam Exp $ + Copyright (C) 1995-2005 + Index Data ApS + +This file is part of the Zebra server. + +Zebra is free software; you can redistribute it and/or modify it under +the terms of the GNU General Public License as published by the Free +Software Foundation; either version 2, or (at your option) any later +version. + +Zebra is distributed in the hope that it will be useful, but WITHOUT ANY +WARRANTY; without even the implied warranty of MERCHANTABILITY or +FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +for more details. + +You should have received a copy of the GNU General Public License +along with Zebra; see the file LICENSE.zebra. If not, write to the +Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA +02111-1307, USA. +*/ + + #include #include @@ -84,12 +28,22 @@ #include #include -#include +#include "dict-p.h" typedef unsigned MatchWord; #define WORD_BITS 32 #define MAX_LENGTH 1024 +/* This code is based + * Sun Wu and Udi Manber: Fast Text Searching Allowing Errors. + * Communications of the ACM, pp. 83-91, Vol. 35, No. 10, Oct. 1992, USA. + * PostScript version of the paper in its submitted form: agrep1.ps) + * recommended reading to understand AGREP ! + * + * http://www.tgries.de/agrep/#AGREP1PS + * http://www.tgries.de/agrep/doc/agrep1ps.zip + */ + typedef struct { int n; /* no of MatchWord needed */ int range; /* max no. of errors */ @@ -118,13 +72,13 @@ static INLINE MatchWord get_bit (MatchContext *mc, MatchWord *m, int ch, static MatchContext *mk_MatchContext (struct DFA *dfa, int range) { - MatchContext *mc = xmalloc (sizeof(*mc)); + MatchContext *mc = (MatchContext *) xmalloc (sizeof(*mc)); int s; mc->n = (dfa->no_states+WORD_BITS) / WORD_BITS; mc->range = range; mc->fact = (range+1)*mc->n; - mc->match_mask = xcalloc (mc->n, sizeof(*mc->match_mask)); + mc->match_mask = (MatchWord *) xcalloc (mc->n, sizeof(*mc->match_mask)); for (s = 0; sno_states; s++) if (dfa->states[s]->rule_no) @@ -301,7 +255,7 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, dict_bf_readp (dict->dbf, ptr, &p); lo = 0; hi = DICT_nodir(p)-1; - indxp = (short*) ((char*) p+DICT_pagesize(dict)-sizeof(short)); + indxp = (short*) ((char*) p+DICT_bsize(p)-sizeof(short)); while (lo <= hi) { @@ -401,8 +355,7 @@ static int dict_grep (Dict dict, Dict_ptr ptr, MatchContext *mc, init_pos)) return 1; dict_bf_readp (dict->dbf, ptr, &p); - indxp = (short*) ((char*) p+DICT_pagesize(dict) - -sizeof(short)); + indxp = (short*) ((char*) p+DICT_bsize(p)-sizeof(short)); } } } @@ -430,10 +383,10 @@ int dict_lookup_grep (Dict dict, const char *pattern, int range, void *client, dfa_verbose = 1; #endif - logf (LOG_DEBUG, "dict_lookup_grep range=%d", range); + yaz_log (YLOG_DEBUG, "dict_lookup_grep range=%d", range); for (i = 0; pattern[i]; i++) { - logf (LOG_DEBUG, " %3d %c", pattern[i], + yaz_log (YLOG_DEBUG, " %3d %c", pattern[i], (pattern[i] > ' ' && pattern[i] < 127) ? pattern[i] : '?'); } @@ -442,6 +395,7 @@ int dict_lookup_grep (Dict dict, const char *pattern, int range, void *client, i = dfa_parse (dfa, &this_pattern); if (i || *this_pattern) { + yaz_log (YLOG_WARN, "dfa_parse fail=%d", i); dfa_delete (&dfa); return -1; } @@ -449,7 +403,7 @@ int dict_lookup_grep (Dict dict, const char *pattern, int range, void *client, mc = mk_MatchContext (dfa, range); - Rj = xcalloc ((MAX_LENGTH+1) * mc->n, sizeof(*Rj)); + Rj = (MatchWord *) xcalloc ((MAX_LENGTH+1) * mc->n, sizeof(*Rj)); set_bit (mc, Rj, 0, 0); for (d = 1; d<=mc->range; d++) @@ -468,12 +422,13 @@ int dict_lookup_grep (Dict dict, const char *pattern, int range, void *client, } } *max_pos = 0; - if (dict->head.last > 1) - i = dict_grep (dict, 1, mc, Rj, 0, client, userfunc, prefix, + if (dict->head.root) + i = dict_grep (dict, dict->head.root, mc, Rj, 0, client, + userfunc, prefix, dfa, max_pos, init_pos); else i = 0; - logf (LOG_DEBUG, "max_pos = %d", *max_pos); + yaz_log (YLOG_DEBUG, "max_pos = %d", *max_pos); dfa_delete (&dfa); xfree (Rj); rm_MatchContext (&mc);