StreamingStringFeatures.h

Go to the documentation of this file.

/*
 * 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 3 of the License, or
 * (at your option) any later version.
 *
 * Written (W) 2011 Shashwat Lal Das
 * Copyright (C) 2011 Berlin Institute of Technology and Max-Planck-Society
 */
#ifndef _STREAMING_STRINGFEATURES__H__
#define _STREAMING_STRINGFEATURES__H__

#include <shogun/lib/common.h>
#include <shogun/mathematics/Math.h>
#include <shogun/base/Parameter.h>
#include <shogun/lib/DataType.h>
#include <shogun/io/InputParser.h>

#include <shogun/features/StreamingFeatures.h>
#include <shogun/features/Alphabet.h>

namespace shogun
{
template <class T> class CStreamingStringFeatures : public CStreamingFeatures
{
public:

    CStreamingStringFeatures();

    CStreamingStringFeatures(CStreamingFile* file,
                 bool is_labelled,
                 int32_t size);

    virtual ~CStreamingStringFeatures();

    virtual void set_vector_reader();

    virtual void set_vector_and_label_reader();

    void use_alphabet(EAlphabet alpha);

    void use_alphabet(CAlphabet* alpha);

    void set_remap(CAlphabet* ascii_alphabet, CAlphabet* binary_alphabet);

    void set_remap(EAlphabet ascii_alphabet=DNA, EAlphabet binary_alphabet=RAWDNA);

    CAlphabet* get_alphabet();
    
    floatmax_t get_num_symbols();

    virtual void start_parser();

    virtual void end_parser();

    virtual bool get_next_example();

    SGString<T> get_vector();

    virtual float64_t get_label();

    virtual void release_example();

    virtual int32_t get_vector_length();

    virtual EFeatureType get_feature_type();

    virtual EFeatureClass get_feature_class();

    virtual CFeatures* duplicate() const;

    inline virtual const char* get_name() const { return "StreamingStringFeatures"; }

    virtual int32_t get_num_vectors() const;

    virtual int32_t get_size();

    virtual int32_t get_num_features();

private:

    void init();
    
    void init(CStreamingFile *file, bool is_labelled, int32_t size);

protected:

    CInputParser<T> parser;

    CAlphabet* alphabet;

    CAlphabet* alpha_ascii;

    CAlphabet* alpha_bin;

    CStreamingFile* working_file;

    SGString<T> current_sgstring;

    T* current_string;

    int32_t current_length;

    float64_t current_label;

    bool has_labels;

    bool remap_to_bin;

    int32_t num_symbols;
};

}
#endif // _STREAMING_STRINGFEATURES__H__