source: main/trunk/greenstone2/common-src/indexers/mg/src/text/mgdictlist.c@ 25147

/**************************************************************************
 *
 * mgdictlist.c -- Program to list a dictionary
 * Copyright (C) 1994  Neil Sharman
 *
 * This program 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 of the License, or
 * (at your option) any later version.
 *
 * This program 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 this program; if not, write to the Free Software
 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 *
 * $Id: mgdictlist.c 25147 2012-02-28 00:59:00Z kjdon $
 *
 **************************************************************************/

#include "sysfuncs.h"

#include "messages.h"
#include "memlib.h"
#include "local_strings.h"
#include "netorder.h"  /* [RPAP - Jan 97: Endian Ordering] */

#include "mg_files.h"
#include "text.h"
#include "invf.h"
#include "locallib.h"
#include "words.h"

/*
   $Log$
   Revision 1.1  2003/02/20 21:18:24  mdewsnip
   Addition of MG package for search and retrieval

   Revision 1.1  1999/08/10 21:18:17  sjboddie
   renamed mg-1.3d directory mg

   Revision 1.2  1998/11/25 07:55:49  rjmcnab

   Modified mg to that you can specify the stemmer you want
   to use via a command line option. You specify it to
   mg_passes during the build process. The number of the
   stemmer that you used is stored within the inverted
   dictionary header and the stemmed dictionary header so
   the correct stemmer is used in later stages of building
   and querying.

   Revision 1.1  1998/11/17 09:35:24  rjmcnab
   *** empty log message ***

   * Revision 1.4  1994/11/29  00:32:07  tes
   * Committing the new merged files and changes.
   *
   * Revision 1.3  1994/10/20  03:57:01  tes
   * I have rewritten the boolean query optimiser and abstracted out the
   * components of the boolean query.
   *
   * Revision 1.2  1994/09/20  04:41:56  tes
   * For version 1.1
   *
 */

static char *RCSID = "$Id: mgdictlist.c 25147 2012-02-28 00:59:00Z kjdon $";


int quick = 0;
int no_of_words[2];
mg_u_long maxcodelen[2];

char *dictname = "";




void 
DumpStemDict (FILE * f)
{
  struct invf_dict_header idh;
  int i;
  u_char prev[MAXSTEMLEN + 1];

  fread (&idh, sizeof (idh), 1, f);

  /* [RPAP - Jan 97: Endian Ordering] */
  NTOHUL(idh.lookback);
  NTOHUL(idh.dict_size);
  NTOHUL(idh.total_bytes);
  NTOHUL(idh.index_string_bytes);
  NTOHD(idh.input_bytes); /* [RJM 07/97: 4G limit] */
  NTOHUL(idh.num_of_docs);
  NTOHUL(idh.static_num_of_docs);
  NTOHUL(idh.num_of_words);
  NTOHUL(idh.stemmer_num);
  NTOHUL(idh.stem_method);

  if (quick)
    printf ("%d\n", idh.dict_size);
  else
    {
      printf ("# lookback           = %u\n", idh.lookback);
      printf ("# dict size          = %u\n", idh.dict_size);
      printf ("# total bytes        = %u\n", idh.total_bytes);
      printf ("# index string bytes = %u\n", idh.index_string_bytes);
      printf ("# input bytes        = %.0f\n", idh.input_bytes); /* [RJM 07/97: 4G limit] */
      printf ("# num of docs        = %u\n", idh.num_of_docs);
      printf ("# static num of docs = %u\n", idh.static_num_of_docs);
      printf ("# num of words       = %u\n", idh.num_of_words);
      printf ("#\n");
    }

  for (i = 0; i < idh.dict_size; i++)
    {
      register mg_u_long copy, suff;
      mg_u_long wcnt, fcnt;

      /* build a new word on top of prev */
      copy = getc (f);
      suff = getc (f);
      *prev = copy + suff;
      fread (prev + copy + 1, sizeof (u_char), suff, f);

      /* read other data, but no need to store it */
      fread (&fcnt, sizeof (fcnt), 1, f);
      fread (&wcnt, sizeof (wcnt), 1, f);

      /* [RPAP - Jan 97: Endian Ordering] */
      NTOHUL(fcnt);
      NTOHUL(wcnt);

      if (!quick)
    {
      printf ("%d: %8d ", i, wcnt);
      printf ("/ %5d ", fcnt);
      printf ("%2d %2d\t\"", *prev, copy);
    }
      printf ("%s", word2str (prev));
      if (quick)
    printf (" %d %d\n", wcnt, fcnt);
      else
    {
      putchar ('"');
      putchar ('\n');
    }
    }
}




void 
ReadInWords (FILE * f)
{
  comp_frags_header cfh;
  mg_u_long *codes;
  u_char prev[MAXSTEMLEN + 1];
  int i;

  if (Read_cfh (f, &cfh, NULL, NULL) == -1)
    FatalError (1, "Unable to read in the dictionary");

  printf ("#\n");
  printf ("#   max code len       = %u\n", cfh.hd.maxcodelen);
  printf ("#   total bytes        = %d\n", cfh.uncompressed_size);
  printf ("#\n");

  if (!(codes = Generate_Huffman_Codes (&cfh.hd, NULL)))
    FatalError (1, "no memory for huffman codes\n");

  for (i = 0; i < cfh.hd.num_codes; i++)
    {
      register int val, copy, j, k;
      char code[33];
      val = fgetc (f);
      copy = (val >> 4) & 0xf;
      val &= 0xf;

      fread (prev + copy + 1, sizeof (u_char), val, f);
      *prev = val + copy;

      for (k = 0, j = cfh.hd.clens[i] - 1; j >= 0; j--, k++)
    code[k] = '0' + ((codes[i] >> j) & 1);
      code[k] = '\0';

      printf ("%d: %2d : %*s : \"%s\"\n", i, cfh.hd.clens[i],
          cfh.hd.maxcodelen, code, word2str (prev));
    }
  Xfree (codes);
  Xfree (cfh.hd.clens);
}


void 
ReadCharHuffman (FILE * f, char *title)
{
  int i;
  huff_data hd;
  mg_u_long *codes;

  if (Read_Huffman_Data (f, &hd, NULL, NULL) == -1)
    FatalError (1, "Unable to read huffman data");

  if (!(codes = Generate_Huffman_Codes (&hd, NULL)))
    FatalError (1, "no memory for huffman codes\n");

  printf ("#\n# %s\n#\n", title);
  for (i = 0; i < hd.num_codes; i++)
    if (hd.clens[i])
      {
    int j, k;
    char code[33];
    for (k = 0, j = hd.clens[i] - 1; j >= 0; j--, k++)
      code[k] = '0' + ((codes[i] >> j) & 1);
    code[k] = '\0';
    printf ("%2d : %*s : \"%s\"\n", hd.clens[i],
        hd.maxcodelen, code, char2str (i));
      }
  Xfree (codes);
  Xfree (hd.clens);
}


void 
ReadLenHuffman (FILE * f, char *title)
{
  int i;
  huff_data hd;
  mg_u_long *codes;

  if (Read_Huffman_Data (f, &hd, NULL, NULL) == -1)
    FatalError (1, "Unable to read huffman data");

  if (!(codes = Generate_Huffman_Codes (&hd, NULL)))
    FatalError (1, "no memory for huffman codes\n");

  printf ("#\n# %s\n#\n", title);
  for (i = 0; i < hd.num_codes; i++)
    if (hd.clens[i])
      {
    int j, k;
    char code[33];
    for (k = 0, j = hd.clens[i] - 1; j >= 0; j--, k++)
      code[k] = '0' + ((codes[i] >> j) & 1);
    code[k] = '\0';
    printf ("%2d : %*s : %d\n", hd.clens[i],
        hd.maxcodelen, code, i);
      }
  Xfree (codes);
  Xfree (hd.clens);
}





void 
DumpTextDict (FILE * f)
{
  struct compression_dict_header cdh;
  int which;

  if (Read_cdh (f, &cdh, NULL, NULL) == -1)
    FatalError (1, "Unable to read dictionary header");
  switch (cdh.dict_type)
    {
    case MG_COMPLETE_DICTIONARY:
      printf ("# COMPLETE DICTIONARY\n");
      break;
    case MG_PARTIAL_DICTIONARY:
      printf ("# PARTIAL DICTIONARY\n");
      break;
    case MG_SEED_DICTIONARY:
      printf ("# SEED DICTIONARY\n");
      break;
    }
  printf ("# num words          = %d\n", cdh.num_words[1]);
  printf ("# num word chars     = %d\n", cdh.num_word_chars[1]);
  printf ("# num non-words      = %d\n", cdh.num_words[0]);
  printf ("# num non-word chars = %d\n", cdh.num_word_chars[0]);
  printf ("# lookback           = %d\n", cdh.lookback);

  for (which = 0; which < 2; which++)
    switch (cdh.dict_type)
      {
      case MG_COMPLETE_DICTIONARY:
    {
      ReadInWords (f);
    }
    break;
      case MG_PARTIAL_DICTIONARY:
    {
      if (cdh.num_words[which])
        ReadInWords (f);

      ReadCharHuffman (f, "Characters");
      ReadLenHuffman (f, "Lengths");
    }
    break;
      case MG_SEED_DICTIONARY:
    {
      if (cdh.num_words[which])
        ReadInWords (f);

      ReadCharHuffman (f, "Characters");
      ReadLenHuffman (f, "Lengths");
    }
    break;
      }
}




void 
DumpStatsDict (FILE * f)
{
  int i;
  compression_stats_header csh;

  fread (&csh, sizeof (csh), 1, f);

  for (i = 0; i < 2; i++)
    {
      int j;
      frags_stats_header fsh;

      fread (&fsh, sizeof (fsh), 1, f);

      /* [RPAP - Jan 97: Endian Ordering] */
      NTOHUL(fsh.num_frags);
      NTOHUL(fsh.mem_for_frags);

      if (!quick)
    printf ("#\n# num %9s      = %u\n#\n", i ? "words" : "non-words",
        fsh.num_frags);

      for (j = 0; j < fsh.num_frags; j++)
    {
      u_char Word[16];
      mg_u_long freq, occur_num;

      fread (&freq, sizeof (freq), 1, f);
      fread (&occur_num, sizeof (occur_num), 1, f);

      /* [RPAP - Jan 97: Endian Ordering] */
      NTOHUL(freq);
      NTOHUL(occur_num);

      Word[0] = fgetc (f);
      fread (Word + 1, Word[0], 1, f);
      printf ("%d: %7d : %7d : \"%s\"\n", j, freq,
          occur_num, word2str (Word));
    }
    }
}


int main (int argc, char **argv)
{
  FILE *fp;
  mg_u_long magic = 0;

  if (argc < 2)
    FatalError (1, "A file name must be specified");
  dictname = argv[1];
  if (strcmp (dictname, "-q") == 0)
    {
      quick = 1;
      if (argc < 3)
    FatalError (1, "A file name must be specified");
      dictname = argv[2];
    }
  if (!(fp = fopen (dictname, "rb")))  /* [RPAP - Feb 97: WIN32 Port] */
    FatalError (1, "Unable to open \"%s\"", dictname);

  fread (&magic, sizeof (magic), 1, fp);

  NTOHUL(magic);  /* [RPAP - Jan 97: Endian Ordering] */

  switch (magic)
    {
    case MAGIC_STEM_BUILD:
      if (!quick)
    printf ("# Contents of STEM file \"%s\"\n#\n", dictname);
      DumpStemDict (fp);
      break;
    case MAGIC_DICT:
      if (!quick)
    printf ("# Contents of DICT file \"%s\"\n#\n", dictname);
      DumpTextDict (fp);
      break;
    case MAGIC_STATS_DICT:
      if (!quick)
    printf ("# Contents of STATS file \"%s\"\n#\n", dictname);
      DumpStatsDict (fp);
      break;
    default:
      FatalError (1, "Bad magic number. \"%s\" cannot be dumped", dictname);
    }
  fclose (fp);
  return 0;
}