/*
- * Copyright (C) 1994-1995, Index Data I/S
+ * Copyright (C) 1994-1996, Index Data I/S
* All rights reserved.
* Sebastian Hammer, Adam Dickmeiss
*
* $Log: main.c,v $
- * Revision 1.11 1995-09-29 14:01:45 adam
+ * Revision 1.47 1997-02-12 20:39:46 adam
+ * Implemented options -f <n> that limits the log to the first <n>
+ * records.
+ * Changed some log messages also.
+ *
+ * Revision 1.46 1997/02/10 10:20:13 adam
+ * Flag fileVerboseFlag set to 0 (default).
+ *
+ * Revision 1.45 1996/11/08 11:10:26 adam
+ * Buffers used during file match got bigger.
+ * Compressed ISAM support everywhere.
+ * Bug fixes regarding masking characters in queries.
+ * Redesigned Regexp-2 queries.
+ *
+ * Revision 1.44 1996/10/29 14:09:48 adam
+ * Use of cisam system - enabled if setting isamc is 1.
+ *
+ * Revision 1.43 1996/06/06 12:08:42 quinn
+ * Added showRecord function
+ *
+ * Revision 1.42 1996/05/31 09:07:01 quinn
+ * Work on character-set handling
+ *
+ * Revision 1.41 1996/05/14 14:04:34 adam
+ * In zebraidx, the 'stat' command is improved. Statistics about ISAM/DICT
+ * is collected.
+ *
+ * Revision 1.40 1996/04/26 10:00:23 adam
+ * Added option -V to zebraidx to display version information.
+ * Removed stupid warnings from file update.
+ *
+ * Revision 1.39 1996/04/09 10:05:21 adam
+ * Bug fix: prev_name buffer possibly too small; allocated in key_file_init.
+ *
+ * Revision 1.38 1996/03/26 16:01:14 adam
+ * New setting lockPath: directory of various lock files.
+ *
+ * Revision 1.37 1996/03/19 12:43:26 adam
+ * Bug fix: File update traversal didn't handle trailing slashes correctly.
+ * Bug fix: Update of sub directory groups wasn't handled correctly.
+ *
+ * Revision 1.36 1996/02/12 18:45:37 adam
+ * New fileVerboseFlag in record group control.
+ *
+ * Revision 1.35 1996/02/12 15:56:11 adam
+ * New code command: unread.
+ *
+ * Revision 1.34 1996/02/07 14:06:39 adam
+ * Better progress report during register merge.
+ * New command: clean - removes temporary shadow files.
+ *
+ * Revision 1.33 1996/02/06 17:11:18 adam
+ * Minor changes.
+ *
+ * Revision 1.32 1996/02/01 20:50:04 adam
+ * Bug fix: zebraIndexUnlock was always called even though zebraIndexLock
+ * was never called - happens when no commands are specified.
+ *
+ * Revision 1.31 1996/01/08 19:15:46 adam
+ * New input filter that works!
+ *
+ * Revision 1.30 1995/12/12 16:00:59 adam
+ * System call sync(2) used after update/commit.
+ * Locking (based on fcntl) uses F_EXLCK and F_SHLCK instead of F_WRLCK
+ * and F_RDLCK.
+ *
+ * Revision 1.29 1995/12/11 11:43:30 adam
+ * Locking based on fcntl instead of flock.
+ * Setting commitEnable removed. Command line option -n can be used to
+ * prevent commit if commit setting is defined in the configuration file.
+ *
+ * Revision 1.28 1995/12/08 16:22:56 adam
+ * Work on update while servers are running. Three lock files introduced.
+ * The servers reload their registers when necessary, but they don't
+ * reestablish result sets yet.
+ *
+ * Revision 1.27 1995/12/07 17:38:47 adam
+ * Work locking mechanisms for concurrent updates/commit.
+ *
+ * Revision 1.26 1995/12/06 12:41:23 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 1.25 1995/12/01 16:24:39 adam
+ * Commit files use separate meta file area.
+ *
+ * Revision 1.24 1995/11/30 17:01:38 adam
+ * New setting commitCache: points to commit directories/files.
+ * New command commit: commits at the end of a zebraidx run.
+ *
+ * Revision 1.23 1995/11/30 08:34:31 adam
+ * Started work on commit facility.
+ * Changed a few malloc/free to xmalloc/xfree.
+ *
+ * Revision 1.22 1995/11/28 09:09:42 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 1.21 1995/11/27 14:27:39 adam
+ * Renamed 'update' command to 'dir'.
+ *
+ * Revision 1.20 1995/11/27 13:58:53 adam
+ * New option -t. storeStore data implemented in server.
+ *
+ * Revision 1.19 1995/11/25 10:24:06 adam
+ * More record fields - they are enumerated now.
+ * New options: flagStoreData flagStoreKey.
+ *
+ * Revision 1.18 1995/11/22 17:19:17 adam
+ * Record management uses the bfile system.
+ *
+ * Revision 1.17 1995/11/21 15:01:16 adam
+ * New general match criteria implemented.
+ * New feature: document groups.
+ *
+ * Revision 1.16 1995/11/20 11:56:27 adam
+ * Work on new traversal.
+ *
+ * Revision 1.15 1995/11/01 16:25:51 quinn
+ * *** empty log message ***
+ *
+ * Revision 1.14 1995/10/17 18:02:09 adam
+ * New feature: databases. Implemented as prefix to words in dictionary.
+ *
+ * Revision 1.13 1995/10/10 12:24:39 adam
+ * Temporary sort files are compressed.
+ *
+ * Revision 1.12 1995/10/04 16:57:20 adam
+ * Key input and merge sort in one pass.
+ *
+ * Revision 1.11 1995/09/29 14:01:45 adam
* Bug fixes.
*
* Revision 1.10 1995/09/28 14:22:57 adam
#include <assert.h>
#include <unistd.h>
-#include <alexutil.h>
+#include <data1.h>
#include "index.h"
char *prog;
-size_t mem_max = 4*1024*1024;
+size_t mem_max = 0;
+
+static void abort_func (int level, const char *msg, void *info)
+{
+ if (level & LOG_FATAL)
+ abort ();
+}
int main (int argc, char **argv)
{
int ret;
int cmd = 0;
char *arg;
- char *base_name = NULL;
- char *base_path = NULL;
+ char *configName = NULL;
int nsections;
- char **mbuf;
+ int disableCommit = 0;
+
+ struct recordGroup rGroupDef;
+
+ rGroupDef.groupName = NULL;
+ rGroupDef.databaseName = NULL;
+ rGroupDef.path = NULL;
+ rGroupDef.recordId = NULL;
+ rGroupDef.recordType = NULL;
+ rGroupDef.flagStoreData = -1;
+ rGroupDef.flagStoreKeys = -1;
+ rGroupDef.flagShowRecords = 0;
+ rGroupDef.fileVerboseLimit = 100000;
prog = *argv;
- while ((ret = options ("r:v:m:", argv, argc, &arg)) != -2)
+ if (argc < 2)
+ {
+ fprintf (stderr, "zebraidx [options] command <dir> ...\n"
+ "Commands:\n"
+ " update <dir> Update index with files below <dir>.\n"
+ " If <dir> is empty filenames are read from stdin.\n"
+ " delete <dir> Delete index with files below <dir>.\n"
+ " commit Commit changes\n"
+ " clean Clean shadow files\n"
+ "Options:\n"
+ " -t <type> Index files as <type> (grs or text).\n"
+ " -c <config> Read configuration file <config>.\n"
+ " -g <group> Index files according to group settings.\n"
+ " -d <database> Records belong to Z39.50 database <database>.\n"
+ " -m <mbytes> Use <mbytes> before flushing keys to disk.\n"
+ " -n Don't use shadow system.\n"
+ " -s Show analysis on stdout, but do no work.\n"
+ " -v <level> Set logging to <level>.\n"
+ " -f <n> Display information for the first <n> records.\n"
+ " -V Show version.\n"
+ );
+ exit (1);
+ }
+ while ((ret = options ("sVt:c:g:d:m:v:nf:", argv, argc, &arg)) != -2)
{
if (ret == 0)
{
- if (!base_name)
+ const char *rval;
+ if(cmd == 0) /* command */
{
- base_name = arg;
-
- common_resource = res_open (base_name);
if (!common_resource)
{
- logf (LOG_FATAL, "Cannot open resource `%s'", base_name);
- exit (1);
+ logf (LOG_LOG, "zebra version %s %s",
+ ZEBRAVER, ZEBRADATE);
+ common_resource = res_open (configName ?
+ configName : FNAME_CONFIG);
+ if (!common_resource)
+ {
+ logf (LOG_FATAL, "cannot open resource `%s'",
+ configName);
+ exit (1);
+ }
+ data1_set_tabpath (res_get (common_resource,
+ "profilePath"));
+ bf_lockDir (res_get (common_resource, "lockDir"));
+ init_charmap();
}
- }
- else if(cmd == 0) /* command */
- {
- if (!strcmp (arg, "add"))
+ if (!strcmp (arg, "update"))
+ cmd = 'u';
+ else if (!strcmp (arg, "update1"))
+ cmd = 'U';
+ else if (!strcmp (arg, "update2"))
+ cmd = 'm';
+ else if (!strcmp (arg, "dump"))
+ cmd = 's';
+ else if (!strcmp (arg, "del") || !strcmp(arg, "delete"))
+ cmd = 'd';
+ else if (!strcmp (arg, "commit"))
{
- cmd = 'a';
+ zebraIndexLock (1);
+ rval = res_get (common_resource, "shadow");
+ if (rval && *rval)
+ bf_cache (1);
+ else
+ {
+ logf (LOG_FATAL, "Cannot perform commit");
+ logf (LOG_FATAL, "No shadow area defined");
+ exit (1);
+ }
+ if (bf_commitExists ())
+ {
+ logf (LOG_LOG, "commit start");
+ zebraIndexLockMsg ("c");
+ zebraIndexWait (1);
+ logf (LOG_LOG, "commit execute");
+ bf_commitExec ();
+ sync ();
+ zebraIndexLockMsg ("d");
+ zebraIndexWait (0);
+ logf (LOG_LOG, "commit clean");
+ bf_commitClean ();
+ }
+ else
+ logf (LOG_LOG, "cothing to commit");
}
- else if (!strcmp (arg, "del"))
+ else if (!strcmp (arg, "clean"))
{
- cmd = 'd';
+ zebraIndexLock (1);
+ if (bf_commitExists ())
+ {
+ zebraIndexLockMsg ("d");
+ zebraIndexWait (0);
+ logf (LOG_LOG, "commit clean");
+ bf_commitClean ();
+ }
+ else
+ logf (LOG_LOG, "nothing to clean");
+ }
+ else if (!strcmp (arg, "stat") || !strcmp (arg, "status"))
+ {
+ zebraIndexLock (0);
+ rval = res_get (common_resource, "shadow");
+ if (rval && *rval)
+ {
+ bf_cache (1);
+ zebraIndexLockMsg ("r");
+ }
+ rec_prstat ();
+ inv_prstat ();
}
else
{
- logf (LOG_FATAL, "Unknown command: %s", arg);
+ logf (LOG_FATAL, "unknown command: %s", arg);
exit (1);
}
}
else
{
- unlink ("keys.tmp");
- key_open (mem_max);
- repository (cmd, arg, base_path);
+ struct recordGroup rGroup;
+
+ log_event_end (abort_func, NULL);
+ zebraIndexLock (0);
+ rval = res_get (common_resource, "shadow");
+ if (rval && *rval && !disableCommit)
+ {
+ bf_cache (1);
+ zebraIndexLockMsg ("r");
+ }
+ else
+ {
+ bf_cache (0);
+ zebraIndexLockMsg ("w");
+ }
+ zebraIndexWait (0);
+
+ memcpy (&rGroup, &rGroupDef, sizeof(rGroup));
+ rGroup.path = arg;
+ switch (cmd)
+ {
+ case 'u':
+ key_open (mem_max);
+ logf (LOG_LOG, "updating %s", rGroup.path);
+ repositoryUpdate (&rGroup);
+ nsections = key_close ();
+ break;
+ case 'U':
+ key_open (mem_max);
+ logf (LOG_LOG, "updating (pass 1) %s", rGroup.path);
+ repositoryUpdate (&rGroup);
+ key_close ();
+ nsections = 0;
+ break;
+ case 'd':
+ key_open (mem_max);
+ logf (LOG_LOG, "deleting %s", rGroup.path);
+ repositoryDelete (&rGroup);
+ nsections = key_close ();
+ break;
+ case 's':
+ logf (LOG_LOG, "dumping %s", rGroup.path);
+ repositoryShow (&rGroup);
+ nsections = 0;
+ break;
+ case 'm':
+ nsections = -1;
+ break;
+ default:
+ nsections = 0;
+ }
cmd = 0;
+ if (nsections)
+ {
+ logf (LOG_LOG, "merging with index");
+ key_input (nsections, 60);
+ sync ();
+ }
+ log_event_end (NULL, NULL);
}
}
- else if (ret == 'v')
- {
- log_init (log_mask_str(arg), prog, NULL);
- }
- else if (ret == 'r')
+ else if (ret == 'V')
{
- base_path = arg;
+ fprintf (stderr, "Zebra %s %s\n", ZEBRAVER, ZEBRADATE);
}
+ else if (ret == 'v')
+ log_init (log_mask_str(arg), prog, NULL);
else if (ret == 'm')
- {
mem_max = 1024*1024*atoi(arg);
- }
+ else if (ret == 'd')
+ rGroupDef.databaseName = arg;
+ else if (ret == 's')
+ rGroupDef.flagShowRecords = 1;
+ else if (ret == 'g')
+ rGroupDef.groupName = arg;
+ else if (ret == 'f')
+ rGroupDef.fileVerboseLimit = atoi(arg);
+ else if (ret == 'c')
+ configName = arg;
+ else if (ret == 't')
+ rGroupDef.recordType = arg;
+ else if (ret == 'n')
+ disableCommit = 1;
else
{
- logf (LOG_FATAL, "Unknown option '-%s'", arg);
+ logf (LOG_FATAL, "unknown option '-%s'", arg);
exit (1);
}
}
- if (!base_name)
- {
- fprintf (stderr, "index [-v log] [-r repository] "
- "base cmd1 dir1 cmd2 dir2 ...\n");
- exit (1);
- }
- nsections = key_close ();
- if (!nsections)
- exit (0);
- logf (LOG_LOG, "Merge sorting");
- mbuf = xmalloc (100000);
- merge_sort (mbuf, 1, nsections+1);
- xfree (mbuf);
- logf (LOG_LOG, "Input");
- key_input (FNAME_WORD_DICT, FNAME_WORD_ISAM, "keys1.tmp", 60);
+ if (common_resource)
+ zebraIndexUnlock ();
exit (0);
}