CG MIME Type

#include <iostream>
#include <string>
#include <vector>
#include <sstream>

using namespace std;

/*!!! Just before pasting this code to pastebin it failed the last test once randomly */

//Slightly modified djb2 hashing function.
unsigned long int get_hash(const string &extension, const vector<vector<vector<string>>> &table){
    unsigned long int index = 5381;
    for(int i = 0; i < extension.length(); i++){
        index = (((index << 7) + index) + extension[i]);
    }
    index = index % table.size();

    return index;
}

//Inserts the extension and media type into the hash table.
void insert_hashed(const string &extension, const string &MT, vector<vector<vector<string>>> &table){
    unsigned long int index = get_hash(extension, table);

    table[index].push_back({extension, MT});
}

//Looks for a media type using the extension as key. If found returns the media type, if not returns "UNKNOWN".
string find_hashed(const string &extension, const vector<vector<vector<string>>> &table){
    unsigned long int index = get_hash(extension, table);

    int i = 0;
    int size = table[index].size();
    while(i != size && extension != table[index][i][0]){
            i++;
    }
    if(i == size){
        return "UNKNOWN";
    }
    return table[index][i][1];
}

//Converts a string to lower case.
void to_lower(string &str){
    for(int i = 0; i < str.length(); i++){
        str[i] = (char)tolower((unsigned char)str[i]);
    }
}

//Debug function. Used to test fitness of the hashing function. COMMENT it's use in main() for full performance
void test(vector<vector<vector<string>>> &table){
    int size = table.size();
    int buckets = 0;
    int min_bucket = table[0].size();
    int max_bucket = 0;

    int bucket_size = 0;
    for(int i = 0; i < size; i++){
        bucket_size = table[i].size();

        if(bucket_size > 0){
            buckets++;
            if(bucket_size > max_bucket){
                max_bucket = bucket_size;
            }
        }
        if(bucket_size < min_bucket){
            min_bucket = bucket_size;
        }
    }
    cerr << "Buckets: " << buckets << "\n" <<
            "Empty indices: " << size - buckets << "\n" <<
            "Max size of bucket: " << max_bucket << "\n" <<
            "Min size of bucket: " << min_bucket << "\n" <<
            "Average fill of buckets(load): " << float((float)buckets/(float)size) << endl;
}

int main()
{
    int N; // Number of elements which make up the association table.
    cin >> N; cin.ignore();
    int Q; // Number Q of file names to be analyzed.
    cin >> Q; cin.ignore();

    vector<vector<vector<string>>> table (N, vector<vector<string>>()); // The hash table.
    string EXT; // File extension.
    string MT; // MIME type.

    for (int i = 0; i < N; i++) {
        cin >> EXT >> MT; cin.ignore();

        to_lower(EXT);
        insert_hashed(EXT, MT, table);
    }

    test(table); // COMMENT this line for full performance!

    stringstream out_stream; // Stream to store all the output messages.
    string FNAME;
    string extension;
    for (int i = 0; i < Q; i++) {
        getline(cin, FNAME); // One file name per line.
        extension = FNAME.substr(FNAME.find_last_of('.') + 1); // Extraction of the extension

        if(!FNAME.compare(extension)){ // If FNAME and extension are the same then there was no '.' so the FNAME has no extension.
            //cout << "UNKNOWN\n"; // Swapped for the stringstream method.
            out_stream << "UNKNOWN\n";
        } else{
            to_lower(extension);
            //cout << find_hashed(extension, table) << "\n"; // Swapped for the stringstream method.
            out_stream << find_hashed(extension, table) << "\n";
        }
        /*You can swap around the commenting of stringstream and cout in the above section
        and comment the cout below to see the Large dataset fail while outputting the whole answer.*/
    }

    cout << out_stream.str() << endl; // Prints all output messages at once saving some time.
    return 0;
}