HfstSymbolDefs.h

Go to the documentation of this file.

// Copyright (c) 2016 University of Helsinki                          
//                                                                    
// This library is free software; you can redistribute it and/or      
// modify it under the terms of the GNU Lesser General Public         
// License as published by the Free Software Foundation; either       
// version 3 of the License, or (at your option) any later version.
// See the file COPYING included with this distribution for more      
// information.

#ifndef _SYMBOL_DEFS_H_
#define _SYMBOL_DEFS_H_
#include <vector>
#include <string>
#include <map>
#include <set>

#include "hfstdll.h"

/* Macros that can be used instead of hfsT::internal_epsilon etc 
   if there are problems with variable initialization. */
#define INTERNAL_EPSILON "@_EPSILON_SYMBOL_@"
#define INTERNAL_UNKNOWN "@_UNKNOWN_SYMBOL_@"
#define INTERNAL_IDENTITY "@_IDENTITY_SYMBOL_@"
#define INTERNAL_DEFAULT "@_DEFAULT_SYMBOL_@"


namespace hfst
{

  typedef std::string String;

  /* A set of strings. */
  typedef std::set<String> StringSet;

  typedef std::vector<String> StringVector; 

  typedef std::pair<String, String> StringPair;

  typedef std::vector<StringPair> StringPairVector;

  typedef std::set<StringPair> StringPairSet;

  typedef std::map<String, String> HfstSymbolSubstitutions;

  typedef std::map<StringPair, StringPair> HfstSymbolPairSubstitutions;

  typedef std::pair<float, StringPairVector> HfstTwoLevelPath;
  typedef std::set<HfstTwoLevelPath> HfstTwoLevelPaths;
  
/* The internal representations */
  const std::string internal_epsilon = "@_EPSILON_SYMBOL_@";
  const std::string internal_unknown = "@_UNKNOWN_SYMBOL_@";
  const std::string internal_identity = "@_IDENTITY_SYMBOL_@";
  const std::string internal_default = "@_DEFAULT_SYMBOL_@";
  
  /* Check whether a string is equal to reserved internal representation. */
  HFSTDLL bool is_epsilon(std::string const & str);
  HFSTDLL bool is_unknown(std::string const & str);
  HFSTDLL bool is_identity(std::string const & str);
  HFSTDLL bool is_default(std::string const & str);
  HFSTDLL bool is_epsilon(const char * str);
  HFSTDLL bool is_unknown(const char * str);
  HFSTDLL bool is_identity(const char * str);
  HFSTDLL bool is_default(const char * str);

  /* For internal use */
  typedef std::pair<unsigned int, unsigned int> NumberPair;
  typedef std::vector<NumberPair> NumberPairVector;
  typedef std::set<NumberPair> NumberPairSet;
  typedef std::map<String,unsigned int> StringNumberMap;
  typedef std::map<unsigned int,unsigned int> NumberNumberMap;

  namespace symbols {
    HFSTDLL void collect_unknown_sets(StringSet &s1, StringSet &unknown1,
                  StringSet &s2, StringSet &unknown2);
    HFSTDLL int longest_path_length(const hfst::HfstTwoLevelPaths & paths, bool equally_long=false);
    HFSTDLL hfst::HfstTwoLevelPaths get_longest_paths(const hfst::HfstTwoLevelPaths & paths);
    HFSTDLL StringVector to_string_vector(const hfst::HfstTwoLevelPath & path);
    HFSTDLL std::string to_string(const StringVector & sv, bool spaces=false);
    HFSTDLL std::string to_string(const StringPairVector & sv, bool spaces=false);
    HFSTDLL StringVector to_string_vector(const StringPairVector & spv, bool input_side);
    HFSTDLL hfst::HfstTwoLevelPaths remove_flags(const hfst::HfstTwoLevelPaths & paths);
    HFSTDLL hfst::HfstTwoLevelPath remove_flags(const hfst::HfstTwoLevelPath & path);
    HFSTDLL StringPairSet to_string_pair_set(const StringSet & ss);
    HFSTDLL StringPairVector remove_flags(const StringPairVector &v);
    HFSTDLL StringVector remove_flags(const StringVector &v);
  }
}
#endif