queue *HuffmanTree(FILE *fin, uint32_t *numberOfChars){
unsigned char c;
int v[256] = {0};
int i;
struct huffman Huffman[256];
/* Read all characters from file */
while(fscanf(fin, "%c", &c) && !feof(fin)){
v[c]++;
(*numberOfChars)++;
}
/* Exception
* NULL counted more than once */
if(v[0] > 0)
v[0]--;
/* Create node for each character */
int countNodes = 0;
for(i = 0; i < 256; i++){
if(v[i] != 0){
Huffman[countNodes].caracter = i;
Huffman[countNodes].frequency = (float)(v[i])/(float)(*numberOfChars);
countNodes++;
}
}
/* Priority queue */
queue *head = createQueue(Huffman, countNodes);
/* Create HuffmanTree */
queue *root = createHuffmanTree(head);
return root;
}
// encoder
void Huffman::encode()
{
cout << "Begin encoding..." << endl;
cout << "Open file..." << endl;
inFile_.open(inFileName_.c_str(),ios::in);
cout << "Calculating frequency of all ascii chars..." << endl;
createNodeArray();
cout << "Done!" << endl;
inFile_.close();
cout << "Creating Priority-Queue..." << endl;
createPq();
cout << "Done!" << endl;
cout << "Creating Huffman-Tree..." << endl;
createHuffmanTree();
cout << "Done!" << endl;
cout << "Calculating Huffman-Code..." << endl;
calculateHuffmanCode();
cout << "Done!" << endl;
cout << "Saving to outputfile..." << endl;
saveToFile();
cout << "Done!" << endl;
cout << "Ecoding finished!" << endl;
}
Vocabulary::Vocabulary(const char * train_file, int min_count, bool doctag) :
m_vocab(NULL), m_vocab_size(0), m_train_words(0), m_vocab_capacity(1000),
m_vocab_hash(NULL), m_min_reduce(1), m_min_count(min_count), m_doctag(doctag)
{
if(m_doctag) m_min_count = 1;
m_vocab = (struct vocab_word_t *)calloc(m_vocab_capacity, sizeof(struct vocab_word_t));
m_vocab_hash = (int *)calloc(vocab_hash_size, sizeof(int));
loadFromTrainFile(train_file);
if(!m_doctag) createHuffmanTree();
}
/*
* Given a table of frequencies as an input, this function will generate the compression table.
* Caller must pass in an empty compression table to be filled by this function, and a table
* of character frequencies.
*
* Table will have 256 entries and look like:
* 01001
* 001
* 0111
* ...
* 00001
*/
eFileCode GetTableForGeneric(unsigned long long pFrequencies[ENTRIES], huffResult resultArray[256])
{
createHuffmanTree(pFrequencies, resultArray);
return FILE_SUCCESS;
}
/*
* Given a table of frequencies as an input, this function will generate the compression table.
* Caller must pass in an empty compression table to be filled by this function, and a table
* of character frequencies.
*
* Table will have 256 entries and look like:
* 01001
* 001
* 0111
* ...
* 00001
*/
eFileCode GetTableForGeneric(unsigned* pFrequencies, huffResult** resultArray)
{
*resultArray = createHuffmanTree(pFrequencies);
return FILE_SUCCESS;
}