-/*
- * Copyright (C) 1994-1997, Index Data I/S
- * All rights reserved.
- * Sebastian Hammer, Adam Dickmeiss
- *
- * $Log: invstat.c,v $
- * Revision 1.5 1997-09-17 12:19:13 adam
- * Zebra version corresponds to YAZ version 1.4.
- * Changed Zebra server so that it doesn't depend on global common_resource.
- *
- * Revision 1.4 1996/11/08 11:10:21 adam
- * Buffers used during file match got bigger.
- * Compressed ISAM support everywhere.
- * Bug fixes regarding masking characters in queries.
- * Redesigned Regexp-2 queries.
- *
- * Revision 1.3 1996/06/04 10:18:58 adam
- * Minor changes - removed include of ctype.h.
- *
- * Revision 1.2 1996/05/22 08:25:56 adam
- * Minor change.
- *
- * Revision 1.1 1996/05/14 14:04:34 adam
- * In zebraidx, the 'stat' command is improved. Statistics about ISAM/DICT
- * is collected.
- *
- */
+/* $Id: invstat.c,v 1.43 2005-01-15 19:38:25 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 <stdio.h>
#include <assert.h>
#include <string.h>
#include "index.h"
-#include "recindex.h"
struct inv_stat_info {
- ISAM isam;
- ISAMC isamc;
+ ZebraHandle zh;
+ zint no_isam_entries[9];
int no_dict_entries;
int no_dict_bytes;
int isam_bounds[20];
int isam_occurrences[20];
char tmp[128];
+ int isamb_levels[10][5];
+ int isamb_sizes[10];
+ int isamb_blocks[10];
+ unsigned long cksum;
+ int dumpwords;
};
+#define SINGLETON_TYPE 8 /* the type to use for singletons that */
+ /* have no block and no block type */
+
+static void print_dict_item (ZebraMaps zm, const char *s, zint count,
+ int firstsys, int firstseq, int lastsys, int lastseq )
+{
+ int reg_type = s[1];
+ char keybuf[IT_MAX_WORD+1];
+ char *to = keybuf;
+ const char *from = s + 2;
+
+ while (*from)
+ {
+ const char *res = zebra_maps_output (zm, reg_type, &from);
+ if (!res)
+ *to++ = *from++;
+ else
+ while (*res)
+ *to++ = *res++;
+ }
+ *to = '\0';
+ /* yaz_log (YLOG_LOG, "%s", keybuf); */
+ printf("%10" ZINT_FORMAT0 " %s %d.%d - %d.%d\n", count, keybuf,
+ firstsys,firstseq, lastsys,lastseq);
+}
+
static int inv_stat_handle (char *name, const char *info, int pos,
void *client)
{
- int occur;
+ zint occur = 0;
int i = 0;
struct inv_stat_info *stat_info = (struct inv_stat_info*) client;
- ISAM_P isam_p;
+ ISAMS_P isam_p;
+ int firstsys=-1;
+ int firstseq=-1;
+ int lastsys=-1;
+ int lastseq=-1;
stat_info->no_dict_entries++;
stat_info->no_dict_bytes += strlen(name);
- assert (*info == sizeof(ISAM_P));
- memcpy (&isam_p, info+1, sizeof(ISAM_P));
+ assert (*info == sizeof(ISAMS_P));
+ memcpy (&isam_p, info+1, sizeof(ISAMS_P));
- if (stat_info->isam)
+ if (stat_info->zh->reg->isams)
{
- ISPT ispt;
+ ISAMS_PP pp;
+ int occurx = 0;
+ struct it_key key;
- ispt = is_position (stat_info->isam, isam_p);
- occur = is_numkeys (ispt);
- is_pt_free (ispt);
+ pp = isams_pp_open (stat_info->zh->reg->isams, isam_p);
+ occur = isams_pp_num (pp);
+ while (isams_pp_read(pp, &key))
+ {
+ occurx++;
+ }
+ assert (occurx == occur);
+ stat_info->no_isam_entries[0] += occur;
+ isams_pp_close (pp);
}
- if (stat_info->isamc)
+ if (stat_info->zh->reg->isamc)
{
ISAMC_PP pp;
- int occurx = 0;
- char buf[128];
+ zint occurx = 0;
+ struct it_key key;
- pp = isc_pp_open (stat_info->isamc, isam_p);
+ pp = isc_pp_open (stat_info->zh->reg->isamc, isam_p);
occur = isc_pp_num (pp);
-#if 1
- while (isc_pp_read(pp, buf))
+ while (isc_pp_read(pp, &key))
+ {
occurx++;
+ }
assert (occurx == occur);
-#endif
+ stat_info->no_isam_entries[isc_type(isam_p)] += occur;
isc_pp_close (pp);
}
+ if (stat_info->zh->reg->isamb)
+ {
+ ISAMB_PP pp;
+ struct it_key key;
+ int cat = (int) (isam_p & 3);
+ int level;
+ int size;
+ int blocks;
+
+ pp = isamb_pp_open_x(stat_info->zh->reg->isamb, isam_p, &level, 0);
+ while (isamb_pp_read(pp, &key))
+ {
+ occur++;
+ }
+ isamb_pp_close_x (pp, &size, &blocks);
+ stat_info->isamb_blocks[cat] += blocks;
+ stat_info->isamb_sizes[cat] += size;
+ if (level > 4)
+ level = 4;
+ stat_info->isamb_levels[cat][level] ++;
+ stat_info->no_isam_entries[cat] += occur;
+ }
+ i=0;
while (occur > stat_info->isam_bounds[i] && stat_info->isam_bounds[i])
i++;
++(stat_info->isam_occurrences[i]);
-
+ if (stat_info->dumpwords)
+ print_dict_item(stat_info->zh->reg->zebra_maps, name, occur,
+ firstsys,firstseq, lastsys, lastseq);
return 0;
}
-void inv_prstat (BFiles bfs)
+int zebra_register_statistics (ZebraHandle zh, int dumpdict)
{
- Dict dict;
- ISAM isam = NULL;
- ISAMC isamc = NULL;
- Records records;
int i, prev;
int before = 0;
+ zint occur;
int after = 1000000000;
struct inv_stat_info stat_info;
char term_dict[2*IT_MAX_WORD+2];
+ if (zebra_begin_read (zh))
+ return 1;
+
+ stat_info.zh = zh;
+ stat_info.dumpwords=dumpdict;
+
term_dict[0] = 1;
term_dict[1] = 0;
- dict = dict_open (bfs, FNAME_DICT, 100, 0);
- if (!dict)
- {
- logf (LOG_FATAL, "dict_open fail");
- exit (1);
- }
- if (res_get_match (common_resource, "isam", "c", NULL))
- {
- isamc = isc_open (bfs, FNAME_ISAMC, 0, key_isamc_m (common_resource));
- if (!isamc)
- {
- logf (LOG_FATAL, "isc_open fail");
- exit (1);
- }
- }
- else
- {
- isam = is_open (bfs, FNAME_ISAM, key_compare, 0,
- sizeof(struct it_key), common_resource);
- if (!isam)
- {
- logf (LOG_FATAL, "is_open fail");
- exit (1);
- }
- }
- records = rec_open (bfs, 0);
-
+ for (i = 0; i<=SINGLETON_TYPE; i++)
+ stat_info.no_isam_entries[i] = 0;
stat_info.no_dict_entries = 0;
stat_info.no_dict_bytes = 0;
- stat_info.isam = isam;
- stat_info.isamc = isamc;
stat_info.isam_bounds[0] = 1;
stat_info.isam_bounds[1] = 2;
stat_info.isam_bounds[2] = 3;
- stat_info.isam_bounds[3] = 5;
+ stat_info.isam_bounds[3] = 6;
stat_info.isam_bounds[4] = 10;
stat_info.isam_bounds[5] = 20;
stat_info.isam_bounds[6] = 30;
stat_info.isam_bounds[17] = 1000000;
stat_info.isam_bounds[18] = 0;
+ stat_info.cksum = 0;
+
for (i = 0; i<20; i++)
stat_info.isam_occurrences[i] = 0;
- dict_scan (dict, term_dict, &before, &after, &stat_info, inv_stat_handle);
+ for (i = 0; i<10; i++)
+ {
+ int j;
+ for (j = 0; j<5; j++)
+ stat_info.isamb_levels[i][j] = 0;
+ stat_info.isamb_sizes[i] = 0;
+ stat_info.isamb_blocks[i] = 0;
+ }
+
+ dict_scan (zh->reg->dict, term_dict, &before, &after, &stat_info,
+ inv_stat_handle);
+
+ if (zh->reg->isamc)
+ {
+ fprintf (stdout, " Blocks Occur Size KB Bytes/Entry\n");
+ for (i = 0; isc_block_used (zh->reg->isamc, i) >= 0; i++)
+ {
+ fprintf (stdout, " %8" ZINT_FORMAT0 " %8" ZINT_FORMAT0,
+ isc_block_used (zh->reg->isamc, i),
+ stat_info.no_isam_entries[i]);
+
+ if (stat_info.no_isam_entries[i])
+ fprintf (stdout, " %8d %f",
+ (int) ((1023.0 + (double)
+ isc_block_used(zh->reg->isamc, i) *
+ isc_block_size(zh->reg->isamc,i))/1024),
+ ((double) isc_block_used(zh->reg->isamc, i) *
+ isc_block_size(zh->reg->isamc,i))/
+ stat_info.no_isam_entries[i]);
+ fprintf (stdout, "\n");
+ }
+ }
- rec_close (&records);
- dict_close (dict);
- if (isam)
- is_close (isam);
- if (isamc)
- isc_close (isamc);
+ if (zh->reg->isamb)
+ {
+ for (i = 0; i<4; i++)
+ {
+ int j;
+ int bsize = isamb_block_info(zh->reg->isamb, i);
+ if (bsize < 0)
+ break;
+ fprintf (stdout, "Category %d\n", i);
+ fprintf (stdout, "Block size %d\n", bsize);
+ fprintf (stdout, "Blocks: %d\n", stat_info.isamb_blocks[i]);
+ fprintf (stdout, "Size: %d\n", stat_info.isamb_sizes[i]);
+ fprintf (stdout, "Entries: " ZINT_FORMAT "\n",
+ stat_info.no_isam_entries[i]);
+ fprintf (stdout, "Total %d\n", stat_info.isamb_blocks[i]*
+ bsize);
+ for (j = 0; j<5; j++)
+ if (stat_info.isamb_levels[i][j])
+ fprintf (stdout, "Level%d %d\n", j,
+ stat_info.isamb_levels[i][j]);
+ fprintf (stdout, "\n");
+ }
+ }
+ fprintf (stdout, "Checksum %08lX\n", stat_info.cksum);
- fprintf (stderr, "%d dictionary entries. %d bytes for strings\n",
- stat_info.no_dict_entries, stat_info.no_dict_bytes);
- fprintf (stderr, " size occurrences\n");
+ fprintf (stdout, "Distinct words %d\n", stat_info.no_dict_entries);
+ occur = 0;
+ for (i = 0; i<9; i++)
+ occur += stat_info.no_isam_entries[i];
+ fprintf (stdout, "Word pos " ZINT_FORMAT "\n", occur);
+ fprintf (stdout, " Occurrences Words\n");
prev = 1;
for (i = 0; stat_info.isam_bounds[i]; i++)
{
int here = stat_info.isam_bounds[i];
- fprintf (stderr, "%7d-%-7d %7d\n",
+ fprintf (stdout, "%7d-%-7d %7d\n",
prev, here, stat_info.isam_occurrences[i]);
prev = here+1;
}
- fprintf (stderr, "%7d- %7d\n",
+ fprintf (stdout, "%7d- %7d\n",
prev, stat_info.isam_occurrences[i]);
+ xmalloc_trav("unfreed"); /*! while hunting memory leaks */
+ zebra_end_read (zh);
+ return 0;
}
+