X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Ftrav.c;h=267fedf14bc8060927bb26c898ef26d128bce1a2;hb=4a617d48ef8d8516c83bf49fac1206b580ba537e;hp=b3ffb458f476cc90921d5c0330b4adb4f0f4e365;hpb=9167923869bcb07faf7e357c883f070be098c66e;p=idzebra-moved-to-github.git diff --git a/index/trav.c b/index/trav.c index b3ffb45..267fedf 100644 --- a/index/trav.c +++ b/index/trav.c @@ -1,120 +1,36 @@ -/* - * Copyright (C) 1994-1996, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: trav.c,v $ - * Revision 1.29 1997-02-12 20:39:47 adam - * Implemented options -f that limits the log to the first - * records. - * Changed some log messages also. - * - * Revision 1.28 1996/11/01 08:58:44 adam - * Interface to isamc system now includes update and delete. - * - * Revision 1.27 1996/10/29 14:06:56 adam - * Include zebrautl.h instead of alexutil.h. - * - * Revision 1.26 1996/06/04 10:19:01 adam - * Minor changes - removed include of ctype.h. - * - * Revision 1.25 1996/05/01 13:46:37 adam - * First work on multiple records in one file. - * New option, -offset, to the "unread" command in the filter module. - * - * Revision 1.24 1996/04/26 10:00:23 adam - * Added option -V to zebraidx to display version information. - * Removed stupid warnings from file update. - * - * Revision 1.23 1996/04/12 07:02:25 adam - * File update of single files. - * - * Revision 1.22 1996/04/09 06:50:50 adam - * Bug fix: bad reference in function fileUpdateR. - * - * Revision 1.21 1996/03/22 15:34:18 quinn - * Fixed bad reference - * - * Revision 1.20 1996/03/21 14:50:10 adam - * File update uses modify-time instead of change-time. - * - * Revision 1.19 1996/03/20 16:16:55 quinn - * Added diagnostic output - * - * Revision 1.18 1996/03/19 12:43:27 adam - * Bug fix: File update traversal didn't handle trailing slashes correctly. - * Bug fix: Update of sub directory groups wasn't handled correctly. - * - * Revision 1.17 1996/02/12 18:45:17 adam - * Changed naming of some functions. - * - * Revision 1.16 1996/02/05 12:30:02 adam - * Logging reduced a bit. - * The remaining running time is estimated during register merge. - * - * Revision 1.15 1995/12/07 17:38:48 adam - * Work locking mechanisms for concurrent updates/commit. - * - * Revision 1.14 1995/12/06 12:41:26 adam - * New command 'stat' for the index program. - * Filenames can be read from stdin by specifying '-'. - * Bug fix/enhancement of the transformation from terms to regular - * expressons in the search engine. - * - * Revision 1.13 1995/11/28 09:09:46 adam - * Zebra config renamed. - * Use setting 'recordId' to identify record now. - * Bug fix in recindex.c: rec_release_blocks was invokeded even - * though the blocks were already released. - * File traversal properly deletes records when needed. - * - * Revision 1.12 1995/11/24 11:31:37 adam - * Commands add & del read filenames from stdin if source directory is - * empty. - * Match criteria supports 'constant' strings. - * - * Revision 1.11 1995/11/22 17:19:19 adam - * Record management uses the bfile system. - * - * Revision 1.10 1995/11/21 15:01:16 adam - * New general match criteria implemented. - * New feature: document groups. - * - * Revision 1.9 1995/11/21 09:20:32 adam - * Yet more work on record match. - * - * Revision 1.8 1995/11/20 16:59:46 adam - * New update method: the 'old' keys are saved for each records. - * - * Revision 1.7 1995/11/20 11:56:28 adam - * Work on new traversal. - * - * Revision 1.6 1995/11/17 15:54:42 adam - * Started work on virtual directory structure. - * - * Revision 1.5 1995/10/17 18:02:09 adam - * New feature: databases. Implemented as prefix to words in dictionary. - * - * Revision 1.4 1995/09/28 09:19:46 adam - * xfree/xmalloc used everywhere. - * Extract/retrieve method seems to work for text records. - * - * Revision 1.3 1995/09/06 16:11:18 adam - * Option: only one word key per file. - * - * Revision 1.2 1995/09/04 12:33:43 adam - * Various cleanup. YAZ util used instead. - * - * Revision 1.1 1995/09/01 14:06:36 adam - * Split of work into more files. - * - */ +/* $Id: trav.c,v 1.41 2002-09-03 11:44:54 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 -#include -#include -#include #include +#ifdef WIN32 +#include +#define S_ISREG(x) (x & _S_IFREG) +#define S_ISDIR(x) (x & _S_IFDIR) +#else +#include +#endif +#include #include #include @@ -127,14 +43,15 @@ static int repComp (const char *a, const char *b, size_t len) return memcmp (a, b, len); } -static void repositoryExtractR (int deleteFlag, char *rep, - struct recordGroup *rGroup) +static void repositoryExtractR (ZebraHandle zh, int deleteFlag, char *rep, + struct recordGroup *rGroup, + int level) { struct dir_entry *e; int i; size_t rep_len = strlen (rep); - e = dir_open (rep); + e = dir_open (rep, zh->path_reg, rGroup->followLinks); if (!e) return; logf (LOG_LOG, "dir %s", rep); @@ -142,16 +59,23 @@ static void repositoryExtractR (int deleteFlag, char *rep, rep[rep_len] = '/'; else --rep_len; + for (i=0; e[i].name; i++) { + char *ecp; strcpy (rep +rep_len+1, e[i].name); + if ((ecp = strrchr (e[i].name, '/'))) + *ecp = '\0'; + if (level == 0 && rGroup->databaseNamePath) + rGroup->databaseName = e[i].name; + switch (e[i].kind) { case dirs_file: - fileExtract (NULL, rep, rGroup, deleteFlag); + fileExtract (zh, NULL, rep, rGroup, deleteFlag); break; case dirs_dir: - repositoryExtractR (deleteFlag, rep, rGroup); + repositoryExtractR (zh, deleteFlag, rep, rGroup, level+1); break; } } @@ -159,9 +83,10 @@ static void repositoryExtractR (int deleteFlag, char *rep, } -static void fileDeleteR (struct dirs_info *di, struct dirs_entry *dst, - const char *base, char *src, - struct recordGroup *rGroup) +static void fileDeleteR (ZebraHandle zh, + struct dirs_info *di, struct dirs_entry *dst, + const char *base, char *src, + struct recordGroup *rGroup) { char tmppath[1024]; size_t src_len = strlen (src); @@ -172,7 +97,7 @@ static void fileDeleteR (struct dirs_info *di, struct dirs_entry *dst, { case dirs_file: sprintf (tmppath, "%s%s", base, dst->path); - fileExtract (&dst->sysno, tmppath, rGroup, 1); + fileExtract (zh, &dst->sysno, tmppath, rGroup, 1); strcpy (tmppath, dst->path); dst = dirs_read (di); @@ -189,9 +114,11 @@ static void fileDeleteR (struct dirs_info *di, struct dirs_entry *dst, } } -static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, - const char *base, char *src, - struct recordGroup *rGroup) +static void fileUpdateR (ZebraHandle zh, + struct dirs_info *di, struct dirs_entry *dst, + const char *base, char *src, + struct recordGroup *rGroup, + int level) { struct dir_entry *e_src; int i_src = 0; @@ -199,7 +126,7 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, size_t src_len = strlen (src); sprintf (tmppath, "%s%s", base, src); - e_src = dir_open (tmppath); + e_src = dir_open (tmppath, zh->path_reg, rGroup->followLinks); logf (LOG_LOG, "dir %s", tmppath); #if 0 @@ -223,7 +150,7 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, else if (!e_src) { strcpy (src, dst->path); - fileDeleteR (di, dst, base, src, rGroup); + fileDeleteR (zh, di, dst, base, src, rGroup); return; } else @@ -257,6 +184,9 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, else break; logf (LOG_DEBUG, "trav sd=%d", sd); + + if (level == 0 && rGroup->databaseNamePath) + rGroup->databaseName = e_src[i_src].name; if (sd == 0) { strcpy (src + src_len, e_src[i_src].name); @@ -267,7 +197,7 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, case dirs_file: if (e_src[i_src].mtime > dst->mtime) { - if (fileExtract (&dst->sysno, tmppath, rGroup, 0)) + if (fileExtract (zh, &dst->sysno, tmppath, rGroup, 0)) { dirs_add (di, src, dst->sysno, e_src[i_src].mtime); } @@ -277,7 +207,7 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, dst = dirs_read (di); break; case dirs_dir: - fileUpdateR (di, dst, base, src, rGroup); + fileUpdateR (zh, di, dst, base, src, rGroup, level+1); dst = dirs_last (di); logf (LOG_DEBUG, "last is %s", dst ? dst->path : "null"); break; @@ -295,11 +225,11 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, switch (e_src[i_src].kind) { case dirs_file: - if (fileExtract (&sysno, tmppath, rGroup, 0)) + if (fileExtract (zh, &sysno, tmppath, rGroup, 0)) dirs_add (di, src, sysno, e_src[i_src].mtime); break; case dirs_dir: - fileUpdateR (di, dst, base, src, rGroup); + fileUpdateR (zh, di, dst, base, src, rGroup, level+1); if (dst) dst = dirs_last (di); break; @@ -314,12 +244,12 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, switch (dst->kind) { case dirs_file: - fileExtract (&dst->sysno, tmppath, rGroup, 1); + fileExtract (zh, &dst->sysno, tmppath, rGroup, 1); dirs_del (di, dst->path); dst = dirs_read (di); break; case dirs_dir: - fileDeleteR (di, dst, base, src, rGroup); + fileDeleteR (zh, di, dst, base, src, rGroup); dst = dirs_last (di); } } @@ -327,7 +257,7 @@ static void fileUpdateR (struct dirs_info *di, struct dirs_entry *dst, dir_free (&e_src); } -static void groupRes (struct recordGroup *rGroup) +static void groupRes (ZebraHandle zh, struct recordGroup *rGroup) { char resStr[256]; char gPrefix[256]; @@ -338,21 +268,36 @@ static void groupRes (struct recordGroup *rGroup) sprintf (gPrefix, "%s.", rGroup->groupName); sprintf (resStr, "%srecordId", gPrefix); - rGroup->recordId = res_get (common_resource, resStr); + rGroup->recordId = res_get (zh->res, resStr); + sprintf (resStr, "%sdatabasePath", gPrefix); + rGroup->databaseNamePath = + atoi (res_get_def (zh->res, resStr, "0")); + + rGroup->databaseNamePath = + atoi (res_get_def (zh->res, resStr, "0")); + + if (rGroup->followLinks == -1) + { + sprintf (resStr, "%sfollowLinks", gPrefix); + rGroup->followLinks = + atoi (res_get_def (zh->res, resStr, "1")); + } } -void repositoryShow (struct recordGroup *rGroup) +void repositoryShow (ZebraHandle zh) + { + struct recordGroup *rGroup = &zh->rGroup; char src[1024]; int src_len; struct dirs_entry *dst; Dict dict; struct dirs_info *di; - if (!(dict = dict_open (FMATCH_DICT, 50, 1))) + if (!(dict = dict_open (zh->reg->bfs, FMATCH_DICT, 50, 0, 0))) { logf (LOG_FATAL, "dict_open fail of %s", FMATCH_DICT); - exit (1); + return; } assert (rGroup->path); @@ -365,7 +310,7 @@ void repositoryShow (struct recordGroup *rGroup) src[++src_len] = '\0'; } - di = dirs_open (dict, src); + di = dirs_open (dict, src, rGroup->flagRw); while ( (dst = dirs_read (di)) ) logf (LOG_LOG, "%s", dst->path); @@ -373,7 +318,8 @@ void repositoryShow (struct recordGroup *rGroup) dict_close (dict); } -static void fileUpdate (Dict dict, struct recordGroup *rGroup, +static void fileUpdate (ZebraHandle zh, + Dict dict, struct recordGroup *rGroup, const char *path) { struct dirs_info *di; @@ -383,10 +329,20 @@ static void fileUpdate (Dict dict, struct recordGroup *rGroup, int src_len; assert (path); + + if (zh->path_reg && !yaz_is_abspath(path)) + { + strcpy (src, zh->path_reg); + strcat (src, "/"); + } + else + *src = '\0'; + strcat (src, path); + zebra_file_stat (src, &sbuf, rGroup->followLinks); + strcpy (src, path); src_len = strlen (src); - stat (src, &sbuf); if (S_ISREG(sbuf.st_mode)) { struct dirs_entry *e_dst; @@ -396,13 +352,13 @@ static void fileUpdate (Dict dict, struct recordGroup *rGroup, if (e_dst) { if (sbuf.st_mtime > e_dst->mtime) - if (fileExtract (&e_dst->sysno, src, rGroup, 0)) + if (fileExtract (zh, &e_dst->sysno, src, rGroup, 0)) dirs_add (di, src, e_dst->sysno, sbuf.st_mtime); } else { SYSNO sysno = 0; - if (fileExtract (&sysno, src, rGroup, 0)) + if (fileExtract (zh, &sysno, src, rGroup, 0)) dirs_add (di, src, sysno, sbuf.st_mtime); } dirs_free (&di); @@ -414,77 +370,91 @@ static void fileUpdate (Dict dict, struct recordGroup *rGroup, src[src_len] = '/'; src[++src_len] = '\0'; } - di = dirs_open (dict, src); + di = dirs_open (dict, src, rGroup->flagRw); *dst = '\0'; - fileUpdateR (di, dirs_read (di), src, dst, rGroup); + fileUpdateR (zh, di, dirs_read (di), src, dst, rGroup, 0); dirs_free (&di); } else { - logf (LOG_WARN, "Cannot handle file %s", src); + logf (LOG_WARN, "Ignoring path %s", src); } } -static void repositoryExtract (int deleteFlag, struct recordGroup *rGroup, +static void repositoryExtract (ZebraHandle zh, + int deleteFlag, struct recordGroup *rGroup, const char *path) { struct stat sbuf; char src[1024]; assert (path); + + if (zh->path_reg && !yaz_is_abspath(path)) + { + strcpy (src, zh->path_reg); + strcat (src, "/"); + } + else + *src = '\0'; + strcat (src, path); + zebra_file_stat (src, &sbuf, rGroup->followLinks); + strcpy (src, path); - stat (src, &sbuf); if (S_ISREG(sbuf.st_mode)) - fileExtract (NULL, src, rGroup, deleteFlag); + fileExtract (zh, NULL, src, rGroup, deleteFlag); else if (S_ISDIR(sbuf.st_mode)) - repositoryExtractR (deleteFlag, src, rGroup); + repositoryExtractR (zh, deleteFlag, src, rGroup, 0); else - logf (LOG_WARN, "Cannot handle file %s", src); + logf (LOG_WARN, "Ignoring path %s", src); } -static void repositoryExtractG (int deleteFlag, struct recordGroup *rGroup) +static void repositoryExtractG (ZebraHandle zh, + int deleteFlag, struct recordGroup *rGroup) { if (*rGroup->path == '\0' || !strcmp(rGroup->path, "-")) { char src[1024]; - while (scanf ("%s", src) == 1) - repositoryExtract (deleteFlag, rGroup, src); + while (scanf ("%1020s", src) == 1) + repositoryExtract (zh, deleteFlag, rGroup, src); } else - repositoryExtract (deleteFlag, rGroup, rGroup->path); + repositoryExtract (zh, deleteFlag, rGroup, rGroup->path); } -void repositoryUpdate (struct recordGroup *rGroup) +void repositoryUpdate (ZebraHandle zh) { - groupRes (rGroup); + struct recordGroup *rGroup = &zh->rGroup; + groupRes (zh, rGroup); assert (rGroup->path); if (rGroup->recordId && !strcmp (rGroup->recordId, "file")) { Dict dict; - if (!(dict = dict_open (FMATCH_DICT, 50, 1))) + if (!(dict = dict_open (zh->reg->bfs, FMATCH_DICT, 50, + rGroup->flagRw, 0))) { logf (LOG_FATAL, "dict_open fail of %s", FMATCH_DICT); - exit (1); + return ; } if (*rGroup->path == '\0' || !strcmp(rGroup->path, "-")) { char src[1024]; while (scanf ("%s", src) == 1) - fileUpdate (dict, rGroup, src); + fileUpdate (zh, dict, rGroup, src); } else - fileUpdate (dict, rGroup, rGroup->path); + fileUpdate (zh, dict, rGroup, rGroup->path); dict_close (dict); } else - repositoryExtractG (0, rGroup); + repositoryExtractG (zh, 0, rGroup); } -void repositoryDelete (struct recordGroup *rGroup) +void repositoryDelete (ZebraHandle zh) { - repositoryExtractG (1, rGroup); + repositoryExtractG (zh, 1, &zh->rGroup); }