config_file.cc

   1 /* config_file.cc: configuration load for trec experiments
   2  *
   3  * ----START-LICENCE----
   4  * Copyright 2003 Andy MacFarlane, City University
   5  *
   6  * This program is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU General Public License as
   8  * published by the Free Software Foundation; either version 2 of the
   9  * License, or (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
  19  * USA
  20  * -----END-LICENCE-----
  21  */
  22
  23 #include <iostream>
  24 #include <fstream>
  25 #include <algorithm>
  26 #include <string>
  27 #include <xapian.h>
  28 #include "config_file.h"
  29 #include "split.h"
  30
  31 using namespace Xapian;
  32 using namespace std;
  33
  34 void CONFIG_TREC::record_tag( string config_tag, string config_value ) {
  35
  36   int found=0;
  37
  38   if( config_tag == "textfile" ) {
  39     textfile = config_value;
  40     found = 1;
  41   } // END if
  42   if( config_tag == "stopsfile" ) {
  43     stopsfile = config_value;
  44     found = 1;
  45   } // END if
  46   if( config_tag == "language" ) {
  47     language = config_value;
  48     found = 1;
  49   } // END if
  50   if( config_tag == "db" ) {
  51     db = config_value;
  52     found = 1;
  53   } // END if
  54   if( config_tag == "querytype" ) {
  55     querytype = config_value;
  56     found = 1;
  57   } // END if
  58   if( config_tag == "queryfile" ) {
  59     queryfile = config_value;
  60     found = 1;
  61   } // END if
  62   if( config_tag == "resultsfile" ) {
  63     resultsfile = config_value;
  64     found = 1;
  65   } // END if
  66   if( config_tag == "transfile" ) {
  67     transfile = config_value;
  68     found = 1;
  69   } // END if
  70   if( config_tag == "noresults" ) {
  71     noresults = atoi(config_value.c_str());
  72     found = 1;
  73   } // END if
  74   if( config_tag == "const_k1" ) {
  75     const_k1 = atof(config_value.c_str());
  76     found = 1;
  77   } // END if
  78   if( config_tag == "const_b" ) {
  79     const_b = atof(config_value.c_str());
  80     found = 1;
  81   } // END if
  82   if( config_tag == "topicfile" ) {
  83     topicfile = config_value;
  84     found = 1;
  85   } // END if
  86   if( config_tag == "topicfields" ) {
  87     topicfields = config_value;
  88     found = 1;
  89   } // END if
  90   if( config_tag == "relfile" ) {
  91     relfile = config_value;
  92     found = 1;
  93   } // END if
  94   if( config_tag == "runname" ) {
  95     runname = config_value;
  96     found = 1;
  97   } // END if
  98   if( config_tag == "nterms" ) {
  99     nterms = atoi(config_value.c_str());
 100     found = 1;
 101   } // END if
 102
 103   if( !found ) {
 104     cout << "ERROR: could not locate tag [" << config_tag << "] for value [" << config_value
 105          << "]" << endl;
 106   } // END if
 107
 108 } // END record_tag
 109
 110 void CONFIG_TREC::setup_config( string filename ) {
 111
 112   // set defaults
 113   textfile = "noneassigned";     // must enter a file/dir for text
 114   language = "english";          // corpus language
 115   db = "noneassigned";           // must enter path of database
 116   querytype = "n";               // type of query: default is n=normal
 117   queryfile = "noneassigned";    // must enter path/filename of query file
 118   resultsfile = "trec.log";      // path/filename of results file
 119   transfile = "transaction.log";  // transaction log file (timings etc)
 120   noresults = 1000;              // no of results to save in results log file
 121   const_k1 = 1.2;                // value for K1 constant (BM25)
 122   const_b = 0.75;                // value for B constant (BM25)
 123   topicfile = "noneassigned";    // path/filename of topic file
 124   topicfields = "t";             // fields of topic to use from topic file: default title
 125   relfile= "noneassigned";       // path/filename of relevance judgements file
 126   runname = "xapiantrec";         // name of the run
 127   nterms = 100;                  // no of terms to pick from the topic
 128
 129   std::ifstream configfile( filename.c_str() );
 130
 131   if( !configfile ) {
 132     cerr << "ERROR: you must specify a valid configuration file name" << endl;
 133     exit(0);
 134   } //else cout << "CONFIG) Opened configuration file: " << filename << endl;
 135
 136   while( !configfile.eof() ) {
 137
 138     // read in lines from the configuration file
 139     string data;
 140     //  the tag
 141     string config_tag;
 142     // get the value
 143     string config_value;
 144
 145     // identify and save information from the configuration file
 146     if( !configfile.eof() ) {
 147       configfile >> data;
 148       config_tag = data;
 149       configfile >> data;
 150       config_value = data;
 151       //cout << "GOT) values [" << config_tag << "] and [" << config_value << "]" << endl;
 152
 153       // record the tag
 154       if( !configfile.eof() ) record_tag( config_tag, config_value );
 155     } // END if
 156
 157   } // END while
 158
 159 } // END setup_config
 160
 161 int CONFIG_TREC::check_query_config() {
 162 // ensure that all the information required by query generator has been entered in config file
 163
 164         if( queryfile == "noneassigned" ) {
 165                 cerr << "ERROR: you must specify the query file" << endl;
 166                 return 0;
 167         } // END if
 168         if( stopsfile == "noneassigned" ) {
 169                 cerr << "ERROR: you must specify the stop word file" << endl;
 170                 return 0;
 171         } // END if
 172         if( topicfile == "noneassigned" ) {
 173                 cerr << "ERROR: you must specify the topic file" << endl;
 174                 return 0;
 175         } // END if
 176   if( db == "noneassigned" ) {
 177                 cerr << "ERROR: you must specify the db path" << endl;
 178                 return 0;
 179         } // END if
 180
 181         return 1;
 182
 183 } // END check_query_config
 184
 185 int CONFIG_TREC::check_index_config( ) {
 186 // ensure that all the information required by indexer has been entered in config file
 187
 188         if( stopsfile == "noneassigned" ) {
 189                 cerr << "ERROR: you must specify the stops file" << endl;
 190                 return 0;
 191         } // END if
 192
 193  if( db == "noneassigned" ) {
 194                 cerr << "ERROR: you must specify the db path" << endl;
 195                 return 0;
 196         } // END if
 197
 198  if( textfile == "noneassigned" ) {
 199                 cerr << "ERROR: you must specify the db path" << endl;
 200                 return 0;
 201         } // END if
 202
 203         return 1;
 204
 205 } // END check_index_config
 206
 207 int CONFIG_TREC::check_search_config( ) {
 208 // ensure that all the information required by search program has been entered in config file
 209
 210         if( queryfile == "noneassigned" ) {
 211                 cerr << "ERROR: you must specify the " << endl;
 212                 return 0;
 213         } // END if
 214         if( stopsfile == "noneassigned" ) {
 215                 cerr << "ERROR: you must specify the " << endl;
 216                 return 0;
 217         } // END if
 218
 219         return 1;
 220
 221 } // END check_search_config