#ifndef DATAPROV_H
#define DATAPROV_H

#include<Eigen/Dense>
#include "../tools/easylogging++.h"
#include<vector>

// TODO decide how to construct training data
// One possible way for training data:
//  Matrix composed of an array of feature vectors
//  Labels are composed of linked list, such as
//      6,3,4,0,5,0,0
//      =>  0->6 | 1->3 | 2->4->5
//  How to compensate for non exhaustive labeling?
//      Use -1 to indicate not yet labeled data
//      -1s will be excluded from training

typedef struct DataEntry{
    std::string qid;
    double rank;
    Eigen::VectorXd feature;
} DataEntry;

class DataList{
private:
    int n;
    std::vector<DataEntry*> data;
public:
    int getSize(){return data.size();}
    void addEntry(DataEntry* d){data.push_back(d);}
    void setfSize(int fsize){n=fsize;}
    int getfSize(){return n;}
    int clear(){
        for (int i=0;i<data.size();++i)
            delete data[i];
        data.clear();
    }
    inline std::vector<DataEntry*>& getData(){
        return data;
    }
    ~DataList(){
        clear();
    }
};

class DataProvider  //Virtual base class for data input
{
protected:
    bool eof;
public:
    DataProvider():eof(false){};

    bool EOFile(){return eof;}
    int getAllData(DataList &out){\
        out.clear();
        DataList buf;
        while (!EOFile())
        {
            getDataSet(buf);
            out.getData().insert(out.getData().end(),buf.getData().begin(),buf.getData().end());
        }
    }
    virtual int getDataSet(DataList &out) = 0;
    virtual int open()=0;
    virtual int close()=0;
};

#endif