|
| 1 | +#include <ctype.h> |
| 2 | +#include <stdint.h> |
| 3 | +#include <stdio.h> |
| 4 | +#include <stdlib.h> |
| 5 | +#include <string.h> |
| 6 | + |
| 7 | +static const unsigned char base64_alphabet[65] = |
| 8 | +"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; |
| 9 | + |
| 10 | +static unsigned char* base64_encode(const unsigned char* input, size_t length, size_t* output_length); |
| 11 | +static unsigned char* base64_decode(const unsigned char* input, size_t length, size_t* output_length); |
| 12 | +static int is_valid_base64(const char* string); |
| 13 | +static void show_usage(void); |
| 14 | + |
| 15 | +int main(int argc, char* argv[]) { |
| 16 | + if (argc != 3) { |
| 17 | + show_usage(); |
| 18 | + } |
| 19 | + |
| 20 | + const char* mode = argv[1]; |
| 21 | + const char* input_text = argv[2]; |
| 22 | + |
| 23 | + if (strcmp(mode, "encode") == 0) { |
| 24 | + if (strlen(input_text) == 0) show_usage(); |
| 25 | + |
| 26 | + size_t input_length = strlen(input_text); |
| 27 | + unsigned char* data = base64_encode((const unsigned char*)input_text, input_length, NULL); |
| 28 | + if (!data) { |
| 29 | + show_usage(); |
| 30 | + } |
| 31 | + |
| 32 | + printf("%s\n", data); |
| 33 | + free(data); |
| 34 | + } |
| 35 | + else if (strcmp(mode, "decode") == 0) { |
| 36 | + if (strlen(input_text) == 0 || !is_valid_base64(input_text)) { |
| 37 | + show_usage(); |
| 38 | + } |
| 39 | + |
| 40 | + size_t input_length = strlen(input_text); |
| 41 | + size_t decoded_length; |
| 42 | + unsigned char* data = base64_decode((const unsigned char*)input_text, input_length, &decoded_length); |
| 43 | + if (!data || decoded_length == 0) { |
| 44 | + show_usage(); |
| 45 | + } |
| 46 | + |
| 47 | + fwrite(data, 1, decoded_length, stdout); |
| 48 | + fflush(stdout); |
| 49 | + free(data); |
| 50 | + } |
| 51 | + else { |
| 52 | + show_usage(); |
| 53 | + } |
| 54 | + |
| 55 | + return 0; |
| 56 | +} |
| 57 | + |
| 58 | +static void show_usage(void) { |
| 59 | + fprintf(stderr, "Usage: please provide a mode and a string to encode/decode\n"); |
| 60 | + exit(EXIT_FAILURE); |
| 61 | +} |
| 62 | + |
| 63 | +static int is_valid_base64(const char* str) { |
| 64 | + if (!str) return 0; |
| 65 | + |
| 66 | + size_t length = strlen(str); |
| 67 | + |
| 68 | + // Base64 strings must have length multiple of 4 |
| 69 | + if (length == 0 || length % 4 != 0) { |
| 70 | + return 0; |
| 71 | + } |
| 72 | + |
| 73 | + // Each character must be alphanumeric, '+' or '/', or '=' for padding |
| 74 | + for (size_t i = 0; i < length; i++) { |
| 75 | + unsigned char c = (unsigned char)str[i]; |
| 76 | + if (!isalnum(c) && c != '+' && c != '/' && c != '=') { |
| 77 | + return 0; |
| 78 | + } |
| 79 | + } |
| 80 | + |
| 81 | + // Padding '=' only allowed at the end (max two) |
| 82 | + size_t padding_count = 0; |
| 83 | + if (length >= 1 && str[length - 1] == '=') { |
| 84 | + padding_count++; |
| 85 | + } |
| 86 | + if (length >= 2 && str[length - 2] == '=') { |
| 87 | + padding_count++; |
| 88 | + } |
| 89 | + // '=' should not appear before the padding section |
| 90 | + for (size_t i = 0; i < length - padding_count; i++) { |
| 91 | + if (str[i] == '=') { |
| 92 | + return 0; |
| 93 | + } |
| 94 | + } |
| 95 | + |
| 96 | + return 1; |
| 97 | +} |
| 98 | + |
| 99 | +static unsigned char* base64_encode(const unsigned char* input, size_t length, size_t* output_length) { |
| 100 | + if (!input) return NULL; |
| 101 | + |
| 102 | + // 4 chars for every 3 bytes, rounded up |
| 103 | + size_t encoded_length = 4 * ((length + 2) / 3); |
| 104 | + unsigned char* data = malloc(encoded_length + 1); // +1 for NUL terminator |
| 105 | + if (!data) return NULL; |
| 106 | + |
| 107 | + size_t input_index = 0, encoded_index = 0; |
| 108 | + while (input_index < length) { |
| 109 | + uint32_t buffer = 0; |
| 110 | + int remaining = (int)(length - input_index); |
| 111 | + |
| 112 | + // Pack up to 3 bytes into 24 bits buffer |
| 113 | + buffer |= ((uint32_t)input[input_index++]) << 16; |
| 114 | + if (remaining > 1) { |
| 115 | + buffer |= ((uint32_t)input[input_index++]) << 8; |
| 116 | + } |
| 117 | + if (remaining > 2) { |
| 118 | + buffer |= (uint32_t)input[input_index++]; |
| 119 | + } |
| 120 | + |
| 121 | + // Extract 6-bit groups and map to Base64 chars |
| 122 | + data[encoded_index++] = base64_alphabet[(buffer >> 18) & 0x3F]; |
| 123 | + data[encoded_index++] = base64_alphabet[(buffer >> 12) & 0x3F]; |
| 124 | + data[encoded_index++] = (remaining > 1) ? base64_alphabet[(buffer >> 6) & 0x3F] : '='; |
| 125 | + data[encoded_index++] = (remaining > 2) ? base64_alphabet[buffer & 0x3F] : '='; |
| 126 | + } |
| 127 | + |
| 128 | + data[encoded_index] = '\0'; // Null-terminate output string |
| 129 | + |
| 130 | + if (output_length) { |
| 131 | + *output_length = encoded_index; |
| 132 | + } |
| 133 | + return data; |
| 134 | +} |
| 135 | + |
| 136 | +static unsigned char* base64_decode(const unsigned char* input, size_t length, size_t* output_length) { |
| 137 | + if (!input) return NULL; |
| 138 | + |
| 139 | + unsigned char decoding_table[256]; |
| 140 | + memset(decoding_table, 0x80, 256); // 0x80 marks invalid chars |
| 141 | + |
| 142 | + for (int i = 0; i < 64; i++) { |
| 143 | + decoding_table[base64_alphabet[i]] = (unsigned char)i; |
| 144 | + } |
| 145 | + decoding_table['='] = 0; |
| 146 | + |
| 147 | + // Count valid base64 characters (ignore whitespace or invalid chars) |
| 148 | + size_t valid_char_count = 0; |
| 149 | + for (size_t i = 0; i < length; i++) { |
| 150 | + if (decoding_table[input[i]] != 0x80) { |
| 151 | + valid_char_count++; |
| 152 | + } |
| 153 | + } |
| 154 | + |
| 155 | + if (valid_char_count == 0 || valid_char_count % 4 != 0) { |
| 156 | + return NULL; |
| 157 | + } |
| 158 | + |
| 159 | + size_t decoded_length = valid_char_count / 4 * 3; |
| 160 | + unsigned char* data = malloc(decoded_length); |
| 161 | + if (!data) return NULL; |
| 162 | + |
| 163 | + size_t input_index = 0, decoded_index = 0; |
| 164 | + unsigned char block[4]; |
| 165 | + int padding_count = 0; |
| 166 | + size_t block_char_count = 0; |
| 167 | + |
| 168 | + while (input_index < length) { |
| 169 | + unsigned char val = decoding_table[input[input_index++]]; |
| 170 | + if (val == 0x80) { |
| 171 | + continue; // skip invalid chars (including whitespace) |
| 172 | + } |
| 173 | + |
| 174 | + if (input[input_index - 1] == '=') { |
| 175 | + padding_count++; |
| 176 | + } |
| 177 | + block[block_char_count++] = val; |
| 178 | + |
| 179 | + if (block_char_count == 4) { |
| 180 | + data[decoded_index++] = (block[0] << 2) | (block[1] >> 4); |
| 181 | + data[decoded_index++] = (block[1] << 4) | (block[2] >> 2); |
| 182 | + data[decoded_index++] = (block[2] << 6) | block[3]; |
| 183 | + block_char_count = 0; |
| 184 | + |
| 185 | + if (padding_count) { |
| 186 | + if (padding_count == 1) { |
| 187 | + decoded_index--; |
| 188 | + } |
| 189 | + else if (padding_count == 2) { |
| 190 | + decoded_index -= 2; |
| 191 | + } |
| 192 | + else { |
| 193 | + free(data); |
| 194 | + return NULL; // invalid padding length |
| 195 | + } |
| 196 | + break; |
| 197 | + } |
| 198 | + } |
| 199 | + } |
| 200 | + |
| 201 | + if (output_length) { |
| 202 | + *output_length = decoded_index; |
| 203 | + } |
| 204 | + return data; |
| 205 | +} |
0 commit comments