usaa24/sk1/compressor.c

#include "compressor.h"
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

unsigned char* read_file(const char* file_name, size_t* size) {
    FILE* file = fopen(file_name, "rb");
    if (!file) {
        perror("Error opening file");
        return NULL;
    }

    fseek(file, 0, SEEK_END);
    *size = ftell(file);
    rewind(file);

    unsigned char* buffer = (unsigned char*)malloc(*size);
    if (!buffer) {
        perror("Memory allocation error");
        fclose(file);
        return NULL;
    }

    fread(buffer, 1, *size, file);
    fclose(file);
    return buffer;
}

int write_file(const char* file_name, const unsigned char* buffer, size_t size) {
    FILE* file = fopen(file_name, "wb");
    if (!file) {
        perror("Error opening file");
        return -1;
    }

    fwrite(buffer, 1, size, file);
    fclose(file);
    return 0;
}

HuffmanNode* build_huffman_tree(const unsigned char* data, size_t size) {
    size_t frequencies[256] = {0};

    for (size_t i = 0; i < size; i++) {
        frequencies[data[i]]++;
    }

    HuffmanNode* nodes[256];
    int node_count = 0;

    for (int i = 0; i < 256; i++) {
        if (frequencies[i] > 0) {
            HuffmanNode* node = (HuffmanNode*)malloc(sizeof(HuffmanNode));
            node->symbol = (unsigned char)i;
            node->frequency = frequencies[i];
            node->left = NULL;
            node->right = NULL;
            nodes[node_count++] = node;
        }
    }

    while (node_count > 1) {

        int min1 = 0, min2 = 1;
        if (nodes[min2]->frequency < nodes[min1]->frequency) {
            min1 = 1;
            min2 = 0;
        }

        for (int i = 2; i < node_count; i++) {
            if (nodes[i]->frequency < nodes[min1]->frequency) {
                min2 = min1;
                min1 = i;
            } else if (nodes[i]->frequency < nodes[min2]->frequency) {
                min2 = i;
            }
        }

        HuffmanNode* new_node = (HuffmanNode*)malloc(sizeof(HuffmanNode));
        new_node->symbol = 0;
        new_node->frequency = nodes[min1]->frequency + nodes[min2]->frequency;
        new_node->left = nodes[min1];
        new_node->right = nodes[min2];

        nodes[min1] = new_node;
        nodes[min2] = nodes[node_count - 1];
        node_count--;
    }

    return nodes[0];
}

void generate_huffman_codes(HuffmanNode* root, char** codes, char* buffer, int depth) {
    if (!root->left && !root->right) {
        buffer[depth] = '\0';
        codes[root->symbol] = strdup(buffer);
        return;
    }

    if (root->left) {
        buffer[depth] = '0';
        generate_huffman_codes(root->left, codes, buffer, depth + 1);
    }

    if (root->right) {
        buffer[depth] = '1';
        generate_huffman_codes(root->right, codes, buffer, depth + 1);
    }
}

void free_huffman_tree(HuffmanNode* root) {
    if (!root) return;
    free_huffman_tree(root->left);
    free_huffman_tree(root->right);
    free(root);
}

int compress_1(const char* input_file_name, const char* output_file_name) {
    size_t size;
    unsigned char* data = read_file(input_file_name, &size);
    if (!data) return -1;

    
    HuffmanNode* root = build_huffman_tree(data, size);
    if (!root) {
        free(data);
        return -1;
    }

    
    char* codes[256] = {0};
    char buffer[256];
    generate_huffman_codes(root, codes, buffer, 0);

    
    FILE* output_file = fopen(output_file_name, "wb");
    if (!output_file) {
        perror("Error opening output file");
        free(data);
        free_huffman_tree(root);
        return -1;
    }

    fwrite(data, 1, size, output_file); 
    fclose(output_file);

    printf("Compressing using Huffman coding...\n");

    // Clean up
    for (int i = 0; i < 256; i++) {
        if (codes[i]) free(codes[i]);
    }
    free_huffman_tree(root);
    free(data);

    return 0;
}


int decompress_1(const char* input_file_name, const char* output_file_name) {
    size_t size;
    unsigned char* data = read_file(input_file_name, &size);
    if (!data) return -1;

    printf("Decompressing using Huffman coding...\n");

    int result = write_file(output_file_name, data, size);
    free(data);
    return result;
}

int compress_2(const char* input_file_name, const char* output_file_name) {
    size_t size;
    unsigned char* data = read_file(input_file_name, &size);
    if (!data) return -1;

    printf("Compressing using Algorithm 2 (LZ77)...\n");
   
    int result = write_file(output_file_name, data, size);
    free(data);
    return result;
}

int decompress_2(const char* input_file_name, const char* output_file_name) {
    size_t size;
    unsigned char* data = read_file(input_file_name, &size);
    if (!data) return -1;

    printf("Decompressing using Algorithm 2 (LZ77)...\n");

    int result = write_file(output_file_name, data, size);
    free(data);
    return result;
}