X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=rset%2Frstemp.c;h=e32822b2e66b949c37898060bc69897ce3aaeacd;hb=5066a2c2ead5876e528294b9850f3a3fb5fb3d46;hp=6110d2cb27dc83785c61f217c13a3abeeae3bbcf;hpb=33b386a95c0aac273527d596ce1aafa0dc567b7b;p=idzebra-moved-to-github.git diff --git a/rset/rstemp.c b/rset/rstemp.c index 6110d2c..e32822b 100644 --- a/rset/rstemp.c +++ b/rset/rstemp.c @@ -1,97 +1,30 @@ -/* - * Copyright (C) 1994-1998, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: rstemp.c,v $ - * Revision 1.24 1998-03-05 08:36:28 adam - * New result set model. - * - * Revision 1.23 1997/12/18 10:54:25 adam - * New method result set method rs_hits that returns the number of - * hits in result-set (if known). The ranked result set returns real - * number of hits but only when not combined with other operands. - * - * Revision 1.22 1997/10/31 12:38:12 adam - * Bug fix: added missing xfree() call. - * - * Revision 1.21 1997/09/17 12:19:23 adam - * Zebra version corresponds to YAZ version 1.4. - * Changed Zebra server so that it doesn't depend on global common_resource. - * - * Revision 1.20 1997/09/09 13:38:17 adam - * Partial port to WIN95/NT. - * - * Revision 1.19 1997/09/04 13:58:57 adam - * Added O_BINARY for open calls. - * - * Revision 1.18 1996/10/29 13:54:52 adam - * Changed name of setting tempSetDir to setTmpDir. - * - * Revision 1.17 1995/12/11 09:15:28 adam - * New set types: sand/sor/snot - ranked versions of and/or/not in - * ranked/semi-ranked result sets. - * Note: the snot not finished yet. - * New rset member: flag. - * Bug fix: r_delete in rsrel.c did free bad memory block. - * - * Revision 1.16 1995/11/28 14:47:02 adam - * New setting: tempSetPath. Location of temporary result sets. - * - * Revision 1.15 1995/10/12 12:41:58 adam - * Private info (buf) moved from struct rset_control to struct rset. - * Bug fixes in relevance. - * - * Revision 1.14 1995/10/10 14:00:04 adam - * Function rset_open changed its wflag parameter to general flags. - * - * Revision 1.13 1995/10/06 14:38:06 adam - * New result set method: r_score. - * Local no (sysno) and score is transferred to retrieveCtrl. - * - * Revision 1.12 1995/09/28 09:52:11 adam - * xfree/xmalloc used everywhere. - * - * Revision 1.11 1995/09/18 14:17:56 adam - * Bug fixes. - * - * Revision 1.10 1995/09/15 14:45:39 adam - * Bug fixes. - * - * Revision 1.9 1995/09/15 09:20:42 adam - * Bug fixes. - * - * Revision 1.8 1995/09/08 14:52:42 adam - * Work on relevance feedback. - * - * Revision 1.7 1995/09/07 13:58:44 adam - * New parameter: result-set file descriptor (RSFD) to support multiple - * positions within the same result-set. - * Boolean operators: and, or, not implemented. - * - * Revision 1.6 1995/09/06 16:11:56 adam - * More work on boolean sets. - * - * Revision 1.5 1995/09/05 16:36:59 adam - * Minor changes. - * - * Revision 1.4 1995/09/05 11:43:24 adam - * Complete version of temporary sets. Not tested yet though. - * - * Revision 1.3 1995/09/04 15:20:40 adam - * More work on temp sets. is_open member removed. - * - * Revision 1.2 1995/09/04 09:10:56 adam - * Minor changes. - * - * Revision 1.1 1994/11/04 13:21:30 quinn - * Working. - * - */ +/* $Id: rstemp.c,v 1.33 2002-08-02 19:26:57 adam Exp $ + Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002 + 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 -#ifdef WINDOWS +#ifdef WIN32 #include #else #include @@ -140,45 +73,50 @@ struct rset_temp_info { int dirty; /* window is dirty */ int hits; /* no of hits */ char *temp_path; + int (*cmp)(const void *p1, const void *p2); }; struct rset_temp_rfd { struct rset_temp_info *info; struct rset_temp_rfd *next; + int *countp; + void *buf; }; static void *r_create(RSET ct, const struct rset_control *sel, void *parms) { - rset_temp_parms *temp_parms = parms; + rset_temp_parms *temp_parms = (rset_temp_parms *) parms; struct rset_temp_info *info; - info = xmalloc (sizeof(struct rset_temp_info)); + info = (struct rset_temp_info *) xmalloc (sizeof(struct rset_temp_info)); info->fd = -1; info->fname = NULL; info->key_size = temp_parms->key_size; info->buf_size = 4096; - info->buf_mem = xmalloc (info->buf_size); + info->buf_mem = (char *) xmalloc (info->buf_size); info->pos_cur = 0; info->pos_end = 0; info->pos_buf = 0; info->dirty = 0; info->hits = -1; + info->cmp = temp_parms->cmp; if (!temp_parms->temp_path) info->temp_path = NULL; else { - info->temp_path = xmalloc (strlen(temp_parms->temp_path)+1); + info->temp_path = (char *) xmalloc (strlen(temp_parms->temp_path)+1); strcpy (info->temp_path, temp_parms->temp_path); } ct->no_rset_terms = 1; - ct->rset_terms = xmalloc (sizeof(*ct->rset_terms)); + ct->rset_terms = (RSET_TERM *) xmalloc (sizeof(*ct->rset_terms)); ct->rset_terms[0] = temp_parms->rset_term; + return info; } static RSFD r_open (RSET ct, int flag) { - struct rset_temp_info *info = ct->buf; + struct rset_temp_info *info = (struct rset_temp_info *) ct->buf; struct rset_temp_rfd *rfd; assert (info->fd == -1); @@ -194,9 +132,13 @@ static RSFD r_open (RSET ct, int flag) exit (1); } } - rfd = xmalloc (sizeof(*rfd)); + rfd = (struct rset_temp_rfd *) xmalloc (sizeof(*rfd)); rfd->info = info; r_rewind (rfd); + + rfd->countp = &ct->rset_terms[0]->count; + rfd->buf = xmalloc (info->key_size); + return rfd; } @@ -209,9 +151,26 @@ static void r_flush (RSFD rfd, int mk) if (!info->fname && mk) { - char *s = (char*) tempnam (info->temp_path, "zrs"); +#if HAVE_MKSTEMP + char template[1024]; + + if (info->temp_path) + sprintf (template, "%s/zrsXXXXXX", info->temp_path); + else + sprintf (template, "zrsXXXXXX"); - info->fname = xmalloc (strlen(s)+1); + info->fd = mkstemp (template); + + if (info->fd == -1) + { + logf (LOG_FATAL|LOG_ERRNO, "mkstemp %s", template); + exit (1); + } + info->fname = (char *) xmalloc (strlen(template)+1); + strcpy (info->fname, template); +#else + char *s = (char*) tempnam (info->temp_path, "zrs"); + info->fname = (char *) xmalloc (strlen(s)+1); strcpy (info->fname, s); logf (LOG_DEBUG, "creating tempfile %s", info->fname); @@ -221,10 +180,12 @@ static void r_flush (RSFD rfd, int mk) logf (LOG_FATAL|LOG_ERRNO, "open %s", info->fname); exit (1); } +#endif } if (info->fname && info->fd != -1 && info->dirty) { - size_t r, count; + size_t count; + int r; if (lseek (info->fd, info->pos_buf, SEEK_SET) == -1) { @@ -234,7 +195,7 @@ static void r_flush (RSFD rfd, int mk) count = info->buf_size; if (count > info->pos_end - info->pos_buf) count = info->pos_end - info->pos_buf; - if ((r = write (info->fd, info->buf_mem, count)) < count) + if ((r = write (info->fd, info->buf_mem, count)) < (int) count) { if (r == -1) logf (LOG_FATAL|LOG_ERRNO, "read %s", info->fname); @@ -257,12 +218,13 @@ static void r_close (RSFD rfd) close (info->fd); info->fd = -1; } + xfree (((struct rset_temp_rfd *)rfd)->buf); xfree (rfd); } static void r_delete (RSET ct) { - struct rset_temp_info *info = ct->buf; + struct rset_temp_info *info = (struct rset_temp_info*) ct->buf; if (info->fname) unlink (info->fname); @@ -291,7 +253,8 @@ static void r_reread (RSFD rfd) if (info->fname) { - size_t r, count; + size_t count; + int r; info->pos_border = info->pos_cur + info->buf_size; if (info->pos_border > info->pos_end) @@ -304,7 +267,7 @@ static void r_reread (RSFD rfd) logf (LOG_FATAL|LOG_ERRNO, "lseek %s", info->fname); exit (1); } - if ((r = read (info->fd, info->buf_mem, count)) < count) + if ((r = read (info->fd, info->buf_mem, count)) < (int) count) { if (r == -1) logf (LOG_FATAL|LOG_ERRNO, "read %s", info->fname); @@ -331,14 +294,15 @@ static void r_rewind (RSFD rfd) static int r_count (RSET ct) { - struct rset_temp_info *info = ct->buf; + struct rset_temp_info *info = (struct rset_temp_info *) ct->buf; return info->pos_end / info->key_size; } static int r_read (RSFD rfd, void *buf, int *term_index) { - struct rset_temp_info *info = ((struct rset_temp_rfd*)rfd)->info; + struct rset_temp_rfd *mrfd = (struct rset_temp_rfd*) rfd; + struct rset_temp_info *info = mrfd->info; size_t nc = info->pos_cur + info->key_size; @@ -354,6 +318,12 @@ static int r_read (RSFD rfd, void *buf, int *term_index) info->key_size); info->pos_cur = nc; *term_index = 0; + + if (*mrfd->countp == 0 || (*info->cmp)(buf, mrfd->buf) > 1) + { + memcpy (mrfd->buf, buf, mrfd->info->key_size); + (*mrfd->countp)++; + } return 1; }