usaa21/sk2a/compressor.c

#include <assert.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "compressor.h"
#define BUFSIZE 1024
#define SIZE 256
#define MAX_WORDS 100000

// One codeword is previous index - new index -character

// Trie (a.k.a prefix tree)
struct dict {
        int index;
        // Character is encoded as non zero value in characters
        struct dict* characters[SIZE];
};

// Free trie
void delete(struct dict* tree){
   if (tree == NULL){
           return;
   }
   for (int i = 0 ;i < SIZE; i++){
           delete(tree-> characters[i]);
   }

}

void get_words (struct dict* tree,char* words[]){
    if (tree == NULL){
        return;
    }
    for (int i = 0 ;i < SIZE; i++){
        if (tree-> characters[i] > 0){
            int need_size = snprintf(NULL,0,"-%d-%c\n",tree->index,i);
            words[tree->characters[i]->index] = calloc(need_size + 1,sizeof(char));
            memset(words[tree->characters[i]->index],0,need_size + 1);
            sprintf(words[tree->characters[i]->index],"-%d-%c\n",tree->index,i);

            get_words(tree-> characters[i],words);
        }
    }
}

int mylz78compress(char* buff,int size,char** outbuf){
    // https://medium.com/@dbudhrani/how-data-compression-works-exploring-lz78-e97e539138
    // The trie represents the current phrase
    // The first node represents the empty string with index 0
    struct dict* root = calloc(1,sizeof(struct dict));
    struct dict* current = root;
    int index = 1;
    int c = 0;
    char *last = NULL;
    for(int i = 0; i < size; i++){
        c = buff[i];
        // Current character is in dictionary, continue searching
        if (current->characters[c] > 0){
            // ak este nieco ostalo a toto je koniec
            if (i == size - 1){
                int sz_needed = snprintf(NULL,0,"-%d-%c\n",current->index,c);
                last = calloc(sz_needed + 1,sizeof(char));
                sprintf(last,"-%d-%c\n",current->index,c);
            }
            
            current = current->characters[c];
        }
        else {
            current->characters[c] = calloc(1,sizeof(struct dict));
            current->characters[c]->index = index;
            index += 1;
            current = root;
        }
    }

    char **words = calloc(index,sizeof(char*));
    get_words(root,words);

    int size_needed = 0;
    for(int i=1; i < index; i++){
        if(words[i] != NULL){
            size_needed += strlen(words[i]);
        }
    }

    if (last != NULL){
        size_needed += strlen(last);
    }

    char *temp = calloc(size_needed + 10,sizeof(char));
    int pos = 0;
    for(int i=0; i < index; i++){
        if (words[i] != NULL){
            strcpy(temp + pos,words[i]);
            pos += strlen(words[i]);
        }
    }

    if (last != NULL){
        strcpy(temp + pos,last);
        pos += strlen(last);
    }

    delete(root);


    *outbuf = temp;
    return strlen(temp);
}


int myrlcompress(char* buff,int size,char** outbuf){
    // 2 * size je urcite dost, kedze v najhorsom pripade bude kazdy znak osamote
    *outbuf = calloc(2 * size + 1,sizeof(char));
    memset(*outbuf,0,size + 1);

    int head = -1;
    int length = 0;
    int pos = 0;
    for(int i = 0; i < size; i++){
        int c = buff[i];

        if (c == head){
            length += 1;
            if (i == size - 1){
                (*outbuf)[pos] = head;
                (*outbuf)[pos + 1] = length;
                pos += 2;
                break;
            }
        }
        else {
            if (head != -1){
                (*outbuf)[pos] = head;
                (*outbuf)[pos + 1] = length;
                pos += 2;
            }
            if (i == size - 1){
                (*outbuf)[pos] = c;
                (*outbuf)[pos + 1] = 1;
                pos += 2;
                break;
            }

            head = c;
            length = 1;
        }

        // ak sme blizko preteceniu, vypiseme aktualny znak
        if (length == 255){
            (*outbuf)[pos] = head;
            (*outbuf)[pos + 1] = length;
            pos += 2;
            head = -1;
            length = 0;
        }
    }

    return pos;
}


void compress(FILE* infile,FILE* outfile){
    fseek(infile, 0, SEEK_END);
    int insize = ftell(infile) + 1;
    rewind(infile);
    char *buffer = calloc(insize,sizeof(char));
    
    memset(buffer,0,insize);
    insize = fread(buffer,sizeof(char),insize - 1,infile);
    
    if (insize == 0){
        assert(!ferror(infile));
    }
    
    char *tempbuf = NULL;
    int tempsize = myrlcompress(buffer,insize,&tempbuf);
    
    char *outbuf = NULL;
    int outsize = mylz78compress(tempbuf,tempsize,&outbuf);
    
    if (outsize   > 0){
        fwrite(outbuf,sizeof(char),outsize,outfile);
    }


}


int myrldecompress(char* buff,int size,char** outbuf){
    *outbuf = calloc(size + 1,sizeof(char));
    memset(*outbuf,0,size + 1);
    
    int pos = 0;
    for(int i = 0; i < size; i+=2){
        char c = buff[i];
        char length = buff[i + 1];

        // ak nam nebude stacit miesto v *outbuf, treba allocovat viac
        //zistil som ze pri velkosti 10 MB to nebude nutne
        /*while((length + pos) > (size - 1)){
            size *= 2;
            char *tempbuf = calloc(size,sizeof(char));
            memset(tempbuf,'\0',size);
            strcpy(tempbuf,*outbuf);
            *outbuf = tempbuf;
        }*/

        for (int j = 0; j < length; j++){
            (*outbuf)[pos] = c;
            pos += 1;
        }
    }

    return pos;
}


int mylz78decompress(char* buff,int size,char** outbuf){
    *outbuf = calloc(size,sizeof(char));
    memset(*outbuf,0,size);
    char *words[MAX_WORDS];
    words[0] = "";
    char c;
    int index = 1;
    int number;
    int pos = 0;
    int bytes_read, bytes_read_all = 0;

    while (sscanf(buff + bytes_read_all,"-%d-%c\n%n",&number,&c,&bytes_read) == 2){
        bytes_read_all += bytes_read;

        words[index] = calloc(strlen(words[number]) + 2,sizeof(char*));
        strcpy(words[index],words[number]);
        words[index][strlen(words[number])] = c;

        // ak nam nebude stacit miesto v *outbuf, treba allocovat viac
        while((strlen(words[index]) + pos) > (size - 1)){
            size *= 2;
            char *tempbuf = calloc(size,sizeof(char));
            memset(tempbuf,'\0',size);
            strcpy(tempbuf,*outbuf);
            *outbuf = tempbuf;
        }

        strcpy(*outbuf + pos,words[index]);
        pos += strlen(words[index]);
        index += 1;
    }

    return strlen(*outbuf);
}


void decompress(FILE* infile,FILE* outfile){
    fseek(infile, 0, SEEK_END);
    int insize = ftell(infile) + 1;
    rewind(infile);
    char *buffer = calloc(insize,sizeof(char));
    
    memset(buffer,0,insize);
    insize = fread(buffer,sizeof(char),insize - 1,infile);
    
    if (insize == 0){
        assert(!ferror(infile));
    }
    
    char *tempbuf = NULL;
    int tempsize = mylz78decompress(buffer,insize,&tempbuf);
    
    char *outbuf = NULL;
    int outsize = myrldecompress(tempbuf,tempsize,&outbuf);
    
    if (outsize > 0){
        fwrite(outbuf,sizeof(char),outsize,outfile);
    }

}