X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Finvstat.c;h=3709ed02fcee2bfaeb5119896d9292277671ba2a;hb=eb8857cfa2ceaa2635f14888b95d6f71b8def045;hp=f5bc6aa9e31cc0b40d32dfb1aef17963b72b0dbf;hpb=38f321eafd606e8ac91da5c1450393fed73d37f5;p=idzebra-moved-to-github.git diff --git a/index/invstat.c b/index/invstat.c index f5bc6aa..3709ed0 100644 --- a/index/invstat.c +++ b/index/invstat.c @@ -1,10 +1,26 @@ -/* - * Copyright (C) 1994-1999, Index Data - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss, Heikki Levanto - * log at eof - * - */ +/* $Id: invstat.c,v 1.34 2002-08-05 14:08:08 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 @@ -30,7 +46,8 @@ struct inv_stat_info { #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, int count ) +static void print_dict_item (ZebraMaps zm, const char *s, int count, + int firstsys, int firstseq, int lastsys, int lastseq ) { int reg_type = s[1]; char keybuf[IT_MAX_WORD+1]; @@ -48,7 +65,8 @@ static void print_dict_item (ZebraMaps zm, const char *s, int count ) } *to = '\0'; /* yaz_log (LOG_LOG, "%s", keybuf); */ - printf("%10d %s\n",count, keybuf); + printf("%10d %s %d.%d - %d.%d\n",count, keybuf, + firstsys,firstseq, lastsys,lastseq); } static int inv_stat_handle (char *name, const char *info, int pos, @@ -58,13 +76,19 @@ static int inv_stat_handle (char *name, const char *info, int pos, int i = 0; struct inv_stat_info *stat_info = (struct inv_stat_info*) client; 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(ISAMS_P)); - memcpy (&isam_p, info+1, sizeof(ISAMS_P)); - + if (!stat_info->zh->reg->isamd) + { + assert (*info == sizeof(ISAMS_P)); + memcpy (&isam_p, info+1, sizeof(ISAMS_P)); + } if (stat_info->zh->reg->isams) { @@ -79,6 +103,13 @@ static int inv_stat_handle (char *name, const char *info, int pos, stat_info->cksum = stat_info->cksum * 65509 + key.sysno + 11 * key.seqno; occurx++; + if (-1==firstsys) + { + firstseq=key.seqno; + firstsys=key.sysno; + } + lastsys=key.sysno; + lastseq=key.seqno; } assert (occurx == occur); stat_info->no_isam_entries[0] += occur; @@ -106,6 +137,13 @@ static int inv_stat_handle (char *name, const char *info, int pos, stat_info->cksum = stat_info->cksum * 65509 + key.sysno + 11 * key.seqno; occurx++; + if (-1==firstsys) + { + firstseq=key.seqno; + firstsys=key.sysno; + } + lastsys=key.sysno; + lastseq=key.seqno; } assert (occurx == occur); stat_info->no_isam_entries[isc_type(isam_p)] += occur; @@ -116,8 +154,9 @@ static int inv_stat_handle (char *name, const char *info, int pos, ISAMD_PP pp; int occurx = 0; struct it_key key; - - pp = isamd_pp_open (stat_info->zh->reg->isamd, isam_p); + /* printf("[%d: %d %d %d %d %d %d] ", */ + /* info[0], info[1], info[2], info[3], info[4], info[5], info[7]);*/ + pp = isamd_pp_open (stat_info->zh->reg->isamd, info+1, info[0]); occur = isamd_pp_num (pp); while (isamd_pp_read(pp, &key)) @@ -125,23 +164,34 @@ static int inv_stat_handle (char *name, const char *info, int pos, stat_info->cksum = stat_info->cksum * 65509 + key.sysno + 11 * key.seqno; occurx++; + /* printf("%d.%d ", key.sysno, key.seqno); */ /*!*/ + if (-1==firstsys) + { + firstseq=key.seqno; + firstsys=key.sysno; + } + lastsys=key.sysno; + lastseq=key.seqno; if ( pp->is->method->debug >8 ) logf (LOG_LOG,"sysno=%d seqno=%d (%x/%x) oc=%d/%d ofs=%d ", key.sysno, key.seqno, key.sysno, key.seqno, occur,occurx, pp->offset); } + /* printf("\n"); */ /*!*/ +#ifdef SKIPTHIS if ( pp->is->method->debug >7 ) logf(LOG_LOG,"item %d=%d:%d says %d keys, counted %d", isam_p, isamd_type(isam_p), isamd_block(isam_p), occur, occurx); +#endif if (occurx != occur) logf(LOG_LOG,"Count error!!! read %d, counted %d", occur, occurx); assert (occurx == occur); - if ( is_singleton(isam_p) ) - stat_info->no_isam_entries[SINGLETON_TYPE] += occur; - else - stat_info->no_isam_entries[isamd_type(isam_p)] += occur; + i = pp->cat; + if (info[1]) + i=SINGLETON_TYPE; + stat_info->no_isam_entries[i] += occur; isamd_pp_close (pp); } if (stat_info->zh->reg->isamb) @@ -160,6 +210,13 @@ static int inv_stat_handle (char *name, const char *info, int pos, stat_info->cksum = stat_info->cksum * 65509 + key.sysno + 11 * key.seqno; occur++; + if (-1==firstsys) + { + firstseq=key.seqno; + firstsys=key.sysno; + } + lastsys=key.sysno; + lastseq=key.seqno; } isamb_pp_close_x (pp, &size, &blocks); stat_info->isamb_blocks[cat] += blocks; @@ -169,12 +226,13 @@ static int inv_stat_handle (char *name, const char *info, int pos, 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); + print_dict_item(stat_info->zh->reg->zebra_maps, name, occur, + firstsys,firstseq, lastsys, lastseq); return 0; } @@ -337,106 +395,3 @@ void zebra_register_statistics (ZebraHandle zh, int dumpdict) zebra_end_read (zh); } - -/* - * - * $Log: invstat.c,v $ - * Revision 1.30 2002-07-11 13:03:01 heikki - * Added dumpdict command line option to dump the - * dictionary before doing the usual stats - * - * Revision 1.29 2002/06/19 10:29:17 adam - * align block sizes for isam sys. Better plot for test - * - * Revision 1.28 2002/04/30 19:31:09 adam - * isamb delete; more statistics - * - * Revision 1.27 2002/04/30 08:28:37 adam - * isamb fixes for pp_read. Statistics - * - * Revision 1.26 2002/04/29 18:03:46 adam - * More isamb statistics - * - * Revision 1.25 2002/04/26 08:44:47 adam - * Index statistics working again - * - * Revision 1.24 2002/04/05 08:46:26 adam - * Zebra with full functionality - * - * Revision 1.23 2002/04/04 14:14:13 adam - * Multiple registers (alpha early) - * - * Revision 1.22 2002/02/20 17:30:01 adam - * Work on new API. Locking system re-implemented - * - * Revision 1.21 2000/07/13 10:14:20 heikki - * Removed compiler warnings when making zebra - * - * Revision 1.20 1999/12/01 13:30:30 adam - * Updated configure for Zmbol/Zebra dependent settings. - * - * Revision 1.19 1999/11/30 13:48:03 adam - * Improved installation. Updated for inclusion of YAZ header files. - * - * Revision 1.18 1999/10/06 11:46:36 heikki - * mproved statistics on isam-d - * - * Revision 1.17 1999/08/20 08:28:37 heikki - * Log levels - * - * Revision 1.16 1999/08/18 08:38:22 heikki - * Memory leak hunting - * - * Revision 1.15 1999/08/18 08:34:53 heikki - * isamd - * - * Revision 1.14 1999/07/14 10:59:26 adam - * Changed functions isc_getmethod, isams_getmethod. - * Improved fatal error handling (such as missing EXPLAIN schema). - * - * Revision 1.13 1999/07/08 14:23:27 heikki - * Fixed a bug in isamh_pp_read and cleaned up a bit - * - * Revision 1.12 1999/07/06 12:28:04 adam - * Updated record index structure. Format includes version ID. Compression - * algorithm ID is stored for each record block. - * - * Revision 1.11 1999/05/15 14:36:38 adam - * Updated dictionary. Implemented "compression" of dictionary. - * - * Revision 1.10 1999/05/12 13:08:06 adam - * First version of ISAMS. - * - * Revision 1.9 1999/02/12 13:29:23 adam - * Implemented position-flag for registers. - * - * Revision 1.8 1999/02/02 14:50:53 adam - * Updated WIN32 code specific sections. Changed header. - * - * Revision 1.7 1998/03/13 15:30:50 adam - * New functions isc_block_used and isc_block_size. Fixed 'leak' - * in isc_alloc_block. - * - * Revision 1.6 1998/03/06 13:54:02 adam - * Fixed two nasty bugs in isc_merge. - * - * 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. - */