lingucomponent/source/languageguessing/simpleguesser.hxx

   1 /***************************************************************************
   2  *
   3  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   4  *
   5  * Copyright 2008 by Sun Microsystems, Inc.
   6  *
   7  * OpenOffice.org - a multi-platform office productivity suite
   8  *
   9  * $RCSfile: simpleguesser.hxx,v $
  10  * $Revision: 1.6 $
  11  *
  12  * This file is part of OpenOffice.org.
  13  *
  14  * OpenOffice.org is free software: you can redistribute it and/or modify
  15  * it under the terms of the GNU Lesser General Public License version 3
  16  * only, as published by the Free Software Foundation.
  17  *
  18  * OpenOffice.org is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  21  * GNU Lesser General Public License version 3 for more details
  22  * (a copy is included in the LICENSE file that accompanied this code).
  23  *
  24  * You should have received a copy of the GNU Lesser General Public License
  25  * version 3 along with OpenOffice.org.  If not, see
  26  * <http://www.openoffice.org/license.html>
  27  * for a copy of the LGPLv3 License.
  28  *
  29  ************************************************************************/
  30 #ifndef SIMPLEGUESSER_H
  31 #define SIMPLEGUESSER_H
  32
  33 #include <string.h>
  34 #include <string>
  35 #include <cstdlib>
  36 #include <vector>
  37 #include <guess.hxx>
  38
  39 #define MAX_STRING_LENGTH_TO_ANALYSE 200
  40
  41 using namespace std;
  42
  43 /**
  44 @author Jocelyn Merand
  45 */
  46 class SimpleGuesser{
  47 public:
  48     /**inits the object with conf file "./conf.txt"*/
  49     SimpleGuesser();
  50
  51     /** Compares the current Simpleguesser with an other
  52      * @param SimpleGuesser& sg the other guesser to compare
  53      */
  54     void operator=(SimpleGuesser& sg);
  55
  56     /**
  57      * destroy the object
  58      */
  59     ~SimpleGuesser();
  60
  61     /**
  62      * Analyze a text and return the most probable languages of the text
  63      * @param char* text is the text to analyze
  64      * @return the list of guess
  65      */
  66     vector<Guess> GuessLanguage(char* text);
  67
  68     /**
  69      * Analyze a text and return the most probable language of the text
  70      * @param char* text is the text to analyze
  71      * @return the guess (containing language)
  72      */
  73     Guess GuessPrimaryLanguage(char* text);
  74
  75     /**
  76      * List all available languages (possibly to be in guesses)
  77      * @return the list of languages
  78      */
  79     vector<Guess> GetAvailableLanguages();
  80
  81     /**
  82      * List all languages (possibly in guesses or not)
  83      * @return the list of languages
  84      */
  85     vector<Guess> GetAllManagedLanguages();
  86
  87     /**
  88      * List all Unavailable languages (disable for any reason)
  89      * @return the list of languages
  90      */
  91     vector<Guess> GetUnavailableLanguages();
  92
  93     /**
  94      * Mark a language enabled
  95      * @param string lang the language to enable (build like language-COUNTRY-encoding)
  96      */
  97     void EnableLanguage(string lang);
  98
  99     /**
 100      * Mark a language disabled
 101      * @param string lang the language to disable (build like language-COUNTRY-encoding)
 102      */
 103     void DisableLanguage(string lang);
 104
 105     /**
 106      * Load a new DB of fingerprints
 107      * @param const char* thePathOfConfFile self explaining
 108      * @param const char* prefix is the path where the directory witch contains fingerprint files is stored
 109     */
 110     void SetDBPath(const char* thePathOfConfFile, const char* prefix);
 111
 112 protected:
 113
 114     //Where typical fingerprints (n-gram tables) are stored
 115     void* h;
 116
 117     //Is used to select languages into the fingerprints DB, the mask is used to indicate if we want enabled disabled or both
 118     vector<Guess> GetManagedLanguages(const char mask);
 119
 120     //Like getManagedLanguages, this function enable or disable a language and it depends of the mask
 121     void XableLanguage(string lang, char mask);
 122 };
 123
 124 #endif