Bachelors_Thesis_Code/simple_string_matching.cpp

/* #define NDEBUG */
/* #include "Rabin_fingerprint.hpp" */
/* #include "general_library.hpp" */
#include "processes.hpp"

#include <iostream>
#include <stdint.h>
#include <math.h>
#include <string>
#include <fstream>

void print_match (size_t index, size_t length, std::string &T) {
    std::cout << "Match found at index " << index << " with the text \"";
    for (size_t i = 0; i < length; i++)
        std::cout << T[index + i];
    std::cout << "\"" << std::endl;
}

int main() {
    /* std::ifstream ifs("books/the_complete_works_of_william_shakespeare.txt"); */
    std::ifstream ifs("books/genji_monogatari_english.txt");
    std::string T( (std::istreambuf_iterator<char>(ifs) ),
                    (std::istreambuf_iterator<char>() ) );

    std::string P = "word";

    std::cout << "Searching for pattern:" << std::endl;
    std::cout << "    " << P << std::endl;
    /* std::cout << "in text:" << std::endl; */
    /* std::cout << "    " << T << std::endl; */
    std::cout << std::endl;

    uint32_t irreducible_polynomial = get_random_irreducible_polynomial_in_Z2(31);
    size_t window_size_in_bits = P.length()*8;

    // Hash the pattern
    Rabin_fingerprint_process phiP(irreducible_polynomial, (size_t)window_size_in_bits);
    for (char c : P)
        phiP.stream_char(c);

    // Hash the text
    Rabin_fingerprint_process phiT(irreducible_polynomial, window_size_in_bits);
    for (size_t i = 0; i < P.length(); i++)
        phiT.stream_char(T[i]);
    if (phiT.get_fingerprint() == phiP.get_fingerprint())
        print_match(0, P.length(), T);

    for (size_t i = P.length(); i < T.length(); i++) {
        phiT.stream_char(T[i]);
        if (phiT.get_fingerprint() == phiP.get_fingerprint())
            print_match(i-P.length()+1, P.length(), T);
    }

    std::cout << std::endl;
    std::cout << "Done!" << std::endl;

    return EXIT_SUCCESS;
}