//http://www.cplusplus.com/doc/tutorial/files/

// reading a text file
#include <iostream>
#include <fstream>
#include <string>
#include <map>
#include <iomanip>

#include <cstdlib>
using namespace std;

map<string,int> words;// Words found
void put(string word) {
    ++words[word];
}

void print() {


    for (auto iter = words.begin(); iter != words.end(); iter++) {
        cout << "[" << iter->first << "] = " << iter->second << endl;
    }
}

void print(map<string, int>& M) {

    // Declare a multimap
    multimap<int, string, std::greater<int>> MM;

    // Insert every (key-value) pairs from
    // map M to multimap MM as (value-key)
    // pairs
    for (auto& it : M) {
        MM.insert({ it.second, it.first });
    }

    // Print the multimap
    for (auto& it : MM) {
        cout<< std::setw(15) << it.second << ' '
             << it.first << endl;
    }

    cout<< "size: "<< M.size();
}

void tokenizer(string text) {
    const std::string separators {
        " ,;:.\"!?'\n(){}*"
    }; // Word delimiters
    
    size_t start { text.find_first_not_of(separators) }; // First word start index
    while (start != std::string::npos) { // Find the words
        size_t end = text.find_first_of(separators, start + 1); // Find end of word
        if (end == std::string::npos) // Found a separator?
            end = text.length(); // No, so set to end of text
        //words.push_back(text.substr(start, end - start)); // Store the word
        put(text.substr(start, end - start));
        start = text.find_first_not_of(separators, end + 1); // Find first character of next word
    }
}

string h1=R"(D:\HarryPotter\harry potter7.txt)";

int main () {
    string line;
    string text {};
    ifstream myfile (h1);
    if (myfile.is_open()) {
        while ( getline (myfile,line) ) {
            //cout << line << '\n';
            text+=line;
        }
        myfile.close();
    }

    else cout << "Unable to open file";
    tokenizer(text);
    print(words);
    return 0;
}