/*
CheckTopTen
~ This function goes through the lyrics of the song to see if
~ the current song's frequency of a word should be stored in the top 10
*/
void WordTable::CheckTopTen(Song* song){
for (size_t i = 0; i < song->lyrics.size(); i++){
uint32_t hash = hashWord(alphaOnly(song->lyrics[i]));
wordNode* node = table[hash];
if (node->numSongs == 11){ //if top 10 songs + current song
//currSong = the current song's info (song + word)
SongFreq curr = node->songs_and_freqs[node->numSongs-1];
for (size_t j = 0; j < node->numSongs; j++){
//find song with minimum frequency of word
size_t minIndex = j;
SongFreq minSong = node->songs_and_freqs[j];
for (size_t k = 0; k < node->numSongs; k++){
if (node->songs_and_freqs[k].freq <
minSong.freq){
minIndex = k;
minSong = node->songs_and_freqs[k];
}
}
if (minSong.freq < curr.freq){ //swap if needed
node->songs_and_freqs[minIndex] = curr;
break;
}
}
node->numSongs--;
}
}
}
// makeLocalHashTable: Loop through a song's lyrics, making an association
// list with information about how many times each word appears.
void LocalHashTable::makeLocalHashTable(Song *aSong, int inIndex) {
string lyricToAdd; uint32_t hashCode; uint32_t hashPosition;
for (int j = 0; j < aSong->getLyricsLength(); j++) { // for all lyrics,
// expanding if load factor exceeds ~.8
if (HTLength > (HTCapacity/1.2)) expandLocalHashTable();
lyricToAdd = alphaOnly(aSong->getLyrics(j)); //strip lyric of formatting
hashCode = hash_string(lyricToAdd); // find hashposition
hashPosition = hashCode % (uint32_t) HTCapacity;
uint32_t iter = hashPosition;
while ((iter % (uint32_t) HTCapacity) < (uint32_t) HTCapacity) {
iter = iter % (uint32_t) HTCapacity;
if (freqNodes[iter].getTheWord() == "") { //empty node, initializing
freqNodes[iter].setTheWord(lyricToAdd);
freqNodes[iter].setSongIndex(inIndex);
freqNodes[iter].addPosition(j);
HTLength++;
break;
}
if (freqNodes[iter].getTheWord() == lyricToAdd) {
freqNodes[iter].addPosition(j); // a match has been found
break;
}
iter++;
}
}
}
void LocalHashTable::expandLocalHashTable() {
int newCap = (HTCapacity*2)-1; // effectively doubling, but keeping odd
FreqNode* temp = new FreqNode[newCap];
for (int i = 0; i < HTCapacity; i++) {
// if you find a word node, rehash the word and set the new location
string iterword = freqNodes[i].getTheWord();
if (iterword != "") {
uint32_t hashPosition = hash_string(alphaOnly(iterword));
hashPosition = hashPosition % (uint32_t) newCap;
while (hashPosition < (uint32_t) newCap) {
if (temp[hashPosition].getTheWord() == "") {
temp[hashPosition] = freqNodes[i];
break;
}
hashPosition++;
}
}
}
delete[] freqNodes;
freqNodes = temp;
HTCapacity = newCap;
}
void Data::print(songwithfreq song, string theword){
size_t count = 0;
for (size_t i = 0; i < songlist[song.index].lyrics.size(); i++){
if (alphaOnly(songlist[song.index].lyrics[i]) == theword){
cout << "Title: " << songlist[song.index].title << endl;
cout << "Artist: " << songlist[song.index].artist << endl;
cout << "Context: ";
// print 5 before and after each instance, unless there aren't 5
if (i < 5) {
for (size_t j = 0; j < (i+6); j++){
cout << songlist[song.index].lyrics[j] << " ";
}
cout << endl;
count++;
}
else {
for (size_t j = (i-5); j < (i+6); j++){
cout << songlist[song.index].lyrics[j] << " ";
if (j + 1 == songlist[song.index].lyrics.size()){
break;
}
}
cout << endl;
count++;
}
cout << endl;
}
if (count == song.frequency) break;
}
}
/*
analyzeSong
~ Processes song lyrics into the hash table:
*/
void WordTable::analyzeSong(Song* song){
for (size_t i = 0; i < song->lyrics.size(); i++){
if (load_factor() > LOADMAX) expand(); //expand if necessary
uint32_t hash = hashWord(alphaOnly(song->lyrics[i]));
if (table[hash] == NULL){ //if word doesn't exist in table
//make new wordNode
wordNode *node = new wordNode;
node->word = alphaOnly(song->lyrics[i]);
//first songs_and_freqs index initialized
node->songs_and_freqs[0].song = song;
node->songs_and_freqs[0].freq = 1;
node->numSongs=1;
table[hash] = node;
numWords++;
}
else duplicateWord(hash, song); //word is in table already
}
}
void Data::search(string word){
// strip the word of any punctuation
string theword = alphaOnly(word);
// retrieve the topten vector
vector<songwithfreq> list = thetable.search(theword);
// print all in topten
if (list.empty()){
cout << endl;
}
else {
for (size_t i = 0; i < list.size(); i++){
print(list[i], theword);
}
}
cout << "<END-OF-REPORT>" << endl;
}
void Data::read_lyrics(char * filename, bool show_progress)
{
ifstream in(filename); // creates an input stream
int song_count = 0; // for progress indicator
string artist, title, word; // data from the file
vector<string> lyrics;
size_t index = 0;
// -- While more data to read...
while (!in.eof())
{
// -- First line is the artist
getline(in, artist);
if (in.fail()) break;
// -- Second line is the title
getline(in, title);
if (in.fail()) break;
if ( show_progress )
{
song_count++;
if (song_count % 10000 == 0) {
cout << "At " << song_count <<
" Artist: " << artist <<
" Title:" << title << endl;
}
}
// -- Then read all words and insert into HT until we hit the
// -- special <BREAK> token
while ( in >> word && word != "<BREAK>" ){
lyrics.push_back(word);
thetable.insert_word(alphaOnly(word), index);
}
// -- Important: skip the newline left behind
in.ignore();
process(title, artist, lyrics);
lyrics.clear();
index++;
}
}
/*
search
~ Takes in a word, hashes it to find the word in the hash table
~ Creates a descending list of the songs_and_freqs array for that word.
*/
void WordTable::search(){
string word;
//cout << "Put Word " << endl;
cin >> word;
uint32_t hash = hashWord(alphaOnly(word));
if (table[hash]!= NULL) {
wordNode* node = table[hash];
Sort(node);
//songs_and_freqs array now in descending order
//calls printContext
for (size_t i = 0; i < node->numSongs; i++){
printContext(node->songs_and_freqs[i].song, node->word);
}
}
cout << "<END OF REPORT>" << endl;
}
/*
printContext
~ prints the five words before and the five words after the
~ searched word for each song.
*/
void WordTable::printContext(Song* song, string word){
for (size_t i = 0; i < song->lyrics.size(); i++){
if(alphaOnly(song->lyrics[i]) == word){
size_t j;
//Check if word is in the first 5 words in the song
if (i < 5) j = 0; //if i is in the first 5 words
// start at index 0
else j = i - 5; //otherwise start 5 words
//before our word
cout << "Title: " << song->title << endl;
cout << "Artist: " << song->artist << endl;
cout << "Context: ";
while (j < i+6){
if (j >= song->lyrics.size()) break;
cout << song->lyrics[j] << " ";
j++;
}
cout << endl << endl;
}
}
}