bowTokenIterator.h

Go to the documentation of this file.

// Copyright 2002-2013 CEA LIST
// SPDX-FileCopyrightText: 2022 CEA LIST <gael.de-chalendar@cea.fr>
//
// SPDX-License-Identifier: MIT
 
/************************************************************************
 *
 * @file       bowTokenIterator.h
 * @author     Besancon Romaric (romaric.besancon@cea.fr)
 * @date       Thu Feb  9 2006
 * copyright   Copyright (C) 2006-2012 by CEA LIST
 * Project     BagOfWords
 * 
 * @brief      an iterator on BoWText that returns BoWTokens (including partial complex tokens)
 * 
 * 
 ***********************************************************************/
 
#ifndef BOWTOKENITERATOR_H
#define BOWTOKENITERATOR_H
 
#include "linguisticProcessing/LinguisticProcessingCommon.h"
 
#include <deque>
 
#include <boost/shared_ptr.hpp>
 
namespace Lima {
namespace Common {
namespace BagOfWords {
 
enum BoWTokenIteratorNamedEntitiesPartsPolicy {
  ITERATE_THROUGH_NAMEDENTITIES_PARTS,
  DO_NOT_ITERATE_THROUGH_NAMEDENTITIES_PARTS
};
 
class BoWText;
class AbstractBoWElement;
 
class BoWTokenIteratorPrivate;
class LIMA_BOW_EXPORT BoWTokenIterator {
 public:
  BoWTokenIterator(const BoWText& bowText,
                   const uint64_t nbMaxPartialCompounds=1000,
                   BoWTokenIteratorNamedEntitiesPartsPolicy iterateThroughNamedEntitiesParts = ITERATE_THROUGH_NAMEDENTITIES_PARTS);
  ~BoWTokenIterator();
      
  bool isAtEnd() const;
  boost::shared_ptr< Lima::Common::BagOfWords::AbstractBoWElement > getElement();
  BoWTokenIterator& operator++();   //prefix ++
  BoWTokenIterator operator++(int); //postfix++
 
  void deletePartialComplexTokens();
 
private:
  BoWTokenIterator(const BoWTokenIterator&);
  BoWTokenIterator& operator=(const BoWTokenIterator&);
  BoWTokenIteratorPrivate* m_d;
};
 
} // end namespace
} // end namespace
} // end namespace
 
#endif