The world's most popular open source database

MySQL 5.1: mysql/src/5.1-dbg/storage/myisam/myisam_ftdump.c Source File
mysql/src/5.1-dbg/storage/myisam/myisam_ftdump.c

00001 /* Copyright (C) 2000 MySQL AB & MySQL Finland AB & TCX DataKonsult AB
00002 
00003    This program is free software; you can redistribute it and/or modify
00004    it under the terms of the GNU General Public License as published by
00005    the Free Software Foundation; either version 2 of the License, or
00006    (at your option) any later version.
00007 
00008    This program is distributed in the hope that it will be useful,
00009    but WITHOUT ANY WARRANTY; without even the implied warranty of
00010    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
00011    GNU General Public License for more details.
00012 
00013    You should have received a copy of the GNU General Public License
00014    along with this program; if not, write to the Free Software
00015    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA */
00016 
00017 /* Written by Sergei A. Golubchik, who has a shared copyright to this code
00018    added support for long options (my_getopt) 22.5.2002 by Jani Tolonen */
00019 
00020 #include "ftdefs.h"
00021 #include <my_getopt.h>
00022 
00023 static void usage();
00024 static void complain(int val);
00025 static my_bool get_one_option(int, const struct my_option *, char *);
00026 
00027 static int count=0, stats=0, dump=0, lstats=0;
00028 static my_bool verbose;
00029 static char *query=NULL;
00030 static uint lengths[256];
00031 
00032 #define MAX_LEN (HA_FT_MAXBYTELEN+10)
00033 #define HOW_OFTEN_TO_WRITE 10000
00034 
00035 static struct my_option my_long_options[] =
00036 {
00037   {"help", 'h', "Display help and exit.",
00038    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00039   {"help", '?', "Synonym for -h.",
00040    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00041   {"count", 'c', "Calculate per-word stats (counts and global weights).",
00042    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00043   {"dump", 'd', "Dump index (incl. data offsets and word weights).",
00044    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00045   {"length", 'l', "Report length distribution.",
00046    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00047   {"stats", 's', "Report global stats.",
00048    0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0},
00049   {"verbose", 'v', "Be verbose.",
00050    (gptr*) &verbose, (gptr*) &verbose, 0, GET_BOOL, NO_ARG, 0, 0, 0, 0, 0, 0},
00051   { 0, 0, 0, 0, 0, 0, GET_NO_ARG, NO_ARG, 0, 0, 0, 0, 0, 0}
00052 };
00053 
00054 
00055 int main(int argc,char *argv[])
00056 {
00057   int error=0, subkeys;
00058   uint keylen, keylen2=0, inx, doc_cnt=0;
00059   float weight= 1.0;
00060   double gws, min_gws=0, avg_gws=0;
00061   MI_INFO *info;
00062   char buf[MAX_LEN], buf2[MAX_LEN], buf_maxlen[MAX_LEN], buf_min_gws[MAX_LEN];
00063   ulong total=0, maxlen=0, uniq=0, max_doc_cnt=0;
00064   struct { MI_INFO *info; } aio0, *aio=&aio0; /* for GWS_IN_USE */
00065 
00066   MY_INIT(argv[0]);
00067   if ((error= handle_options(&argc, &argv, my_long_options, get_one_option)))
00068     exit(error);
00069   if (count || dump)
00070     verbose=0;
00071   if (!count && !dump && !lstats && !query)
00072     stats=1;
00073 
00074   if (verbose)
00075     setbuf(stdout,NULL);
00076 
00077   if (argc < 2)
00078     usage();
00079 
00080   {
00081     char *end;
00082     inx= (uint) strtoll(argv[1], &end, 10);
00083     if (*end)
00084       usage();
00085   }
00086 
00087   init_key_cache(dflt_key_cache,MI_KEY_BLOCK_LENGTH,USE_BUFFER_INIT, 0, 0);
00088 
00089   if (!(info=mi_open(argv[0], O_RDONLY,
00090                      HA_OPEN_ABORT_IF_LOCKED|HA_OPEN_FROM_SQL_LAYER)))
00091   {
00092     error=my_errno;
00093     goto err;
00094   }
00095 
00096   *buf2=0;
00097   aio->info=info;
00098 
00099   if ((inx >= info->s->base.keys) ||
00100       !(info->s->keyinfo[inx].flag & HA_FULLTEXT))
00101   {
00102     printf("Key %d in table %s is not a FULLTEXT key\n", inx, info->filename);
00103     goto err;
00104   }
00105 
00106   mi_lock_database(info, F_EXTRA_LCK);
00107 
00108   info->lastpos= HA_OFFSET_ERROR;
00109   info->update|= HA_STATE_PREV_FOUND;
00110 
00111   while (!(error=mi_rnext(info,NULL,inx)))
00112   {
00113     keylen=*(info->lastkey);
00114 
00115     subkeys=ft_sintXkorr(info->lastkey+keylen+1);
00116     if (subkeys >= 0)
00117       weight=*(float*)&subkeys;
00118 
00119 #ifdef HAVE_SNPRINTF
00120     snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey+1);
00121 #else
00122     sprintf(buf,"%.*s",(int) keylen,info->lastkey+1);
00123 #endif
00124     my_casedn_str(default_charset_info,buf);
00125     total++;
00126     lengths[keylen]++;
00127 
00128     if (count || stats)
00129     {
00130       doc_cnt++;
00131       if (strcmp(buf, buf2))
00132       {
00133         if (*buf2)
00134         {
00135           uniq++;
00136           avg_gws+=gws=GWS_IN_USE;
00137           if (count)
00138             printf("%9u %20.7f %s\n",doc_cnt,gws,buf2);
00139           if (maxlen<keylen2)
00140           {
00141             maxlen=keylen2;
00142             strmov(buf_maxlen, buf2);
00143           }
00144           if (max_doc_cnt < doc_cnt)
00145           {
00146             max_doc_cnt=doc_cnt;
00147             strmov(buf_min_gws, buf2);
00148             min_gws=gws;
00149           }
00150         }
00151         strmov(buf2, buf);
00152         keylen2=keylen;
00153         doc_cnt=0;
00154       }
00155     }
00156     if (dump)
00157     {
00158       if (subkeys>=0)
00159         printf("%9lx %20.7f %s\n", (long) info->lastpos,weight,buf);
00160       else
00161         printf("%9lx => %17d %s\n",(long) info->lastpos,-subkeys,buf);
00162     }
00163     if (verbose && (total%HOW_OFTEN_TO_WRITE)==0)
00164       printf("%10ld\r",total);
00165   }
00166   mi_lock_database(info, F_UNLCK);
00167 
00168   if (count || stats)
00169   {
00170     doc_cnt++;
00171     if (*buf2)
00172     {
00173       uniq++;
00174       avg_gws+=gws=GWS_IN_USE;
00175       if (count)
00176         printf("%9u %20.7f %s\n",doc_cnt,gws,buf2);
00177       if (maxlen<keylen2)
00178       {
00179         maxlen=keylen2;
00180         strmov(buf_maxlen, buf2);
00181       }
00182       if (max_doc_cnt < doc_cnt)
00183       {
00184         max_doc_cnt=doc_cnt;
00185         strmov(buf_min_gws, buf2);
00186         min_gws=gws;
00187       }
00188     }
00189   }
00190 
00191   if (stats)
00192   {
00193     count=0;
00194     for (inx=0;inx<256;inx++)
00195     {
00196       count+=lengths[inx];
00197       if ((ulong) count >= total/2)
00198         break;
00199     }
00200     printf("Total rows: %lu\nTotal words: %lu\n"
00201            "Unique words: %lu\nLongest word: %lu chars (%s)\n"
00202            "Median length: %u\n"
00203            "Average global weight: %f\n"
00204            "Most common word: %lu times, weight: %f (%s)\n",
00205            (long) info->state->records, total, uniq, maxlen, buf_maxlen,
00206            inx, avg_gws/uniq, max_doc_cnt, min_gws, buf_min_gws);
00207   }
00208   if (lstats)
00209   {
00210     count=0;
00211     for (inx=0; inx<256; inx++)
00212     {
00213       count+=lengths[inx];
00214       if (count && lengths[inx])
00215         printf("%3u: %10lu %5.2f%% %20lu %4.1f%%\n", inx,
00216                (ulong) lengths[inx],100.0*lengths[inx]/total,(ulong) count,
00217                100.0*count/total);
00218     }
00219   }
00220 
00221 err:
00222   if (error && error != HA_ERR_END_OF_FILE)
00223     printf("got error %d\n",my_errno);
00224   if (info)
00225     mi_close(info);
00226   return 0;
00227 }
00228 
00229 
00230 static my_bool
00231 get_one_option(int optid, const struct my_option *opt __attribute__((unused)),
00232                char *argument __attribute__((unused)))
00233 {
00234   switch(optid) {
00235   case 'd':
00236     dump=1;
00237     complain(count || query);
00238     break;
00239   case 's':
00240     stats=1;
00241     complain(query!=0);
00242     break;
00243   case 'c':
00244     count= 1;
00245     complain(dump || query);
00246     break;
00247   case 'l':
00248     lstats=1;
00249     complain(query!=0);
00250     break;
00251   case '?':
00252   case 'h':
00253     usage();
00254   }
00255   return 0;
00256 }
00257 
00258 #include <help_start.h>
00259 
00260 static void usage()
00261 {
00262   printf("Use: myisam_ftdump <table_name> <index_num>\n");
00263   my_print_help(my_long_options);
00264   my_print_variables(my_long_options);
00265   NETWARE_SET_SCREEN_MODE(1);
00266   exit(1);
00267 }
00268 
00269 #include <help_end.h>
00270 
00271 static void complain(int val) /* Kinda assert :-)  */
00272 {
00273   if (val)
00274   {
00275     printf("You cannot use these options together!\n");
00276     exit(1);
00277   }
00278 }
Generated on Sat Sep 9 00:50:32 2006 for MySQL 5.1 by
1.4.7