Copybara bot | be50d49 | 2023-11-30 00:16:42 +0100 | [diff] [blame] | 1 | <?php |
| 2 | |
| 3 | /** |
| 4 | * FixedBitNotation |
| 5 | * |
| 6 | * @author Andre DeMarre |
| 7 | * @package FixedBitNotation |
| 8 | */ |
| 9 | |
| 10 | /** |
| 11 | * The FixedBitNotation class is for binary to text conversion. It |
| 12 | * can handle many encoding schemes, formally defined or not, that |
| 13 | * use a fixed number of bits to encode each character. |
| 14 | * |
| 15 | * @package FixedBitNotation |
| 16 | */ |
| 17 | class FixedBitNotation |
| 18 | { |
| 19 | protected $_chars; |
| 20 | protected $_bitsPerCharacter; |
| 21 | protected $_radix; |
| 22 | protected $_rightPadFinalBits; |
| 23 | protected $_padFinalGroup; |
| 24 | protected $_padCharacter; |
| 25 | protected $_charmap; |
| 26 | |
| 27 | /** |
| 28 | * Constructor |
| 29 | * |
| 30 | * @param integer $bitsPerCharacter Bits to use for each encoded |
| 31 | * character |
| 32 | * @param string $chars Base character alphabet |
| 33 | * @param boolean $rightPadFinalBits How to encode last character |
| 34 | * @param boolean $padFinalGroup Add padding to end of encoded |
| 35 | * output |
| 36 | * @param string $padCharacter Character to use for padding |
| 37 | */ |
| 38 | public function __construct( |
| 39 | $bitsPerCharacter, $chars = NULL, $rightPadFinalBits = FALSE, |
| 40 | $padFinalGroup = FALSE, $padCharacter = '=') |
| 41 | { |
| 42 | // Ensure validity of $chars |
| 43 | if (!is_string($chars) || ($charLength = strlen($chars)) < 2) { |
| 44 | $chars = |
| 45 | '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-,'; |
| 46 | $charLength = 64; |
| 47 | } |
| 48 | |
| 49 | // Ensure validity of $bitsPerCharacter |
| 50 | if ($bitsPerCharacter < 1) { |
| 51 | // $bitsPerCharacter must be at least 1 |
| 52 | $bitsPerCharacter = 1; |
| 53 | $radix = 2; |
| 54 | |
| 55 | } elseif ($charLength < 1 << $bitsPerCharacter) { |
| 56 | // Character length of $chars is too small for $bitsPerCharacter |
| 57 | // Set $bitsPerCharacter to greatest acceptable value |
| 58 | $bitsPerCharacter = 1; |
| 59 | $radix = 2; |
| 60 | |
| 61 | while ($charLength >= ($radix <<= 1) && $bitsPerCharacter < 8) { |
| 62 | $bitsPerCharacter++; |
| 63 | } |
| 64 | |
| 65 | $radix >>= 1; |
| 66 | |
| 67 | } elseif ($bitsPerCharacter > 8) { |
| 68 | // $bitsPerCharacter must not be greater than 8 |
| 69 | $bitsPerCharacter = 8; |
| 70 | $radix = 256; |
| 71 | |
| 72 | } else { |
| 73 | $radix = 1 << $bitsPerCharacter; |
| 74 | } |
| 75 | |
| 76 | $this->_chars = $chars; |
| 77 | $this->_bitsPerCharacter = $bitsPerCharacter; |
| 78 | $this->_radix = $radix; |
| 79 | $this->_rightPadFinalBits = $rightPadFinalBits; |
| 80 | $this->_padFinalGroup = $padFinalGroup; |
| 81 | $this->_padCharacter = $padCharacter[0]; |
| 82 | } |
| 83 | |
| 84 | /** |
| 85 | * Encode a string |
| 86 | * |
| 87 | * @param string $rawString Binary data to encode |
| 88 | * @return string |
| 89 | */ |
| 90 | public function encode($rawString) |
| 91 | { |
| 92 | // Unpack string into an array of bytes |
| 93 | $bytes = unpack('C*', $rawString); |
| 94 | $byteCount = count($bytes); |
| 95 | |
| 96 | $encodedString = ''; |
| 97 | $byte = array_shift($bytes); |
| 98 | $bitsRead = 0; |
| 99 | |
| 100 | $chars = $this->_chars; |
| 101 | $bitsPerCharacter = $this->_bitsPerCharacter; |
| 102 | $rightPadFinalBits = $this->_rightPadFinalBits; |
| 103 | $padFinalGroup = $this->_padFinalGroup; |
| 104 | $padCharacter = $this->_padCharacter; |
| 105 | |
| 106 | // Generate encoded output; |
| 107 | // each loop produces one encoded character |
| 108 | for ($c = 0; $c < $byteCount * 8 / $bitsPerCharacter; $c++) { |
| 109 | |
| 110 | // Get the bits needed for this encoded character |
| 111 | if ($bitsRead + $bitsPerCharacter > 8) { |
| 112 | // Not enough bits remain in this byte for the current |
| 113 | // character |
| 114 | // Save the remaining bits before getting the next byte |
| 115 | $oldBitCount = 8 - $bitsRead; |
| 116 | $oldBits = $byte ^ ($byte >> $oldBitCount << $oldBitCount); |
| 117 | $newBitCount = $bitsPerCharacter - $oldBitCount; |
| 118 | |
| 119 | if (!$bytes) { |
| 120 | // Last bits; match final character and exit loop |
| 121 | if ($rightPadFinalBits) $oldBits <<= $newBitCount; |
| 122 | $encodedString .= $chars[$oldBits]; |
| 123 | |
| 124 | if ($padFinalGroup) { |
| 125 | // Array of the lowest common multiples of |
| 126 | // $bitsPerCharacter and 8, divided by 8 |
| 127 | $lcmMap = array(1 => 1, 2 => 1, 3 => 3, 4 => 1, |
| 128 | 5 => 5, 6 => 3, 7 => 7, 8 => 1); |
| 129 | $bytesPerGroup = $lcmMap[$bitsPerCharacter]; |
| 130 | $pads = $bytesPerGroup * 8 / $bitsPerCharacter |
| 131 | - ceil((strlen($rawString) % $bytesPerGroup) |
| 132 | * 8 / $bitsPerCharacter); |
| 133 | $encodedString .= str_repeat($padCharacter[0], $pads); |
| 134 | } |
| 135 | |
| 136 | break; |
| 137 | } |
| 138 | |
| 139 | // Get next byte |
| 140 | $byte = array_shift($bytes); |
| 141 | $bitsRead = 0; |
| 142 | |
| 143 | } else { |
| 144 | $oldBitCount = 0; |
| 145 | $newBitCount = $bitsPerCharacter; |
| 146 | } |
| 147 | |
| 148 | // Read only the needed bits from this byte |
| 149 | $bits = $byte >> 8 - ($bitsRead + ($newBitCount)); |
| 150 | $bits ^= $bits >> $newBitCount << $newBitCount; |
| 151 | $bitsRead += $newBitCount; |
| 152 | |
| 153 | if ($oldBitCount) { |
| 154 | // Bits come from seperate bytes, add $oldBits to $bits |
| 155 | $bits = ($oldBits << $newBitCount) | $bits; |
| 156 | } |
| 157 | |
| 158 | $encodedString .= $chars[$bits]; |
| 159 | } |
| 160 | |
| 161 | return $encodedString; |
| 162 | } |
| 163 | |
| 164 | /** |
| 165 | * Decode a string |
| 166 | * |
| 167 | * @param string $encodedString Data to decode |
| 168 | * @param boolean $caseSensitive |
| 169 | * @param boolean $strict Returns NULL if $encodedString contains |
| 170 | * an undecodable character |
| 171 | * @return string|NULL |
| 172 | */ |
| 173 | public function decode($encodedString, $caseSensitive = TRUE, |
| 174 | $strict = FALSE) |
| 175 | { |
| 176 | if (!$encodedString || !is_string($encodedString)) { |
| 177 | // Empty string, nothing to decode |
| 178 | return ''; |
| 179 | } |
| 180 | |
| 181 | $chars = $this->_chars; |
| 182 | $bitsPerCharacter = $this->_bitsPerCharacter; |
| 183 | $radix = $this->_radix; |
| 184 | $rightPadFinalBits = $this->_rightPadFinalBits; |
| 185 | $padFinalGroup = $this->_padFinalGroup; |
| 186 | $padCharacter = $this->_padCharacter; |
| 187 | |
| 188 | // Get index of encoded characters |
| 189 | if ($this->_charmap) { |
| 190 | $charmap = $this->_charmap; |
| 191 | |
| 192 | } else { |
| 193 | $charmap = array(); |
| 194 | |
| 195 | for ($i = 0; $i < $radix; $i++) { |
| 196 | $charmap[$chars[$i]] = $i; |
| 197 | } |
| 198 | |
| 199 | $this->_charmap = $charmap; |
| 200 | } |
| 201 | |
| 202 | // The last encoded character is $encodedString[$lastNotatedIndex] |
| 203 | $lastNotatedIndex = strlen($encodedString) - 1; |
| 204 | |
| 205 | // Remove trailing padding characters |
| 206 | while ($encodedString[$lastNotatedIndex] == $padCharacter[0]) { |
| 207 | $encodedString = substr($encodedString, 0, $lastNotatedIndex); |
| 208 | $lastNotatedIndex--; |
| 209 | } |
| 210 | |
| 211 | $rawString = ''; |
| 212 | $byte = 0; |
| 213 | $bitsWritten = 0; |
| 214 | |
| 215 | // Convert each encoded character to a series of unencoded bits |
| 216 | for ($c = 0; $c <= $lastNotatedIndex; $c++) { |
| 217 | |
| 218 | if (!isset($charmap[$encodedString[$c]]) && !$caseSensitive) { |
| 219 | // Encoded character was not found; try other case |
| 220 | if (isset($charmap[$cUpper |
| 221 | = strtoupper($encodedString[$c])])) { |
| 222 | $charmap[$encodedString[$c]] = $charmap[$cUpper]; |
| 223 | |
| 224 | } elseif (isset($charmap[$cLower |
| 225 | = strtolower($encodedString[$c])])) { |
| 226 | $charmap[$encodedString[$c]] = $charmap[$cLower]; |
| 227 | } |
| 228 | } |
| 229 | |
| 230 | if (isset($charmap[$encodedString[$c]])) { |
| 231 | $bitsNeeded = 8 - $bitsWritten; |
| 232 | $unusedBitCount = $bitsPerCharacter - $bitsNeeded; |
| 233 | |
| 234 | // Get the new bits ready |
| 235 | if ($bitsNeeded > $bitsPerCharacter) { |
| 236 | // New bits aren't enough to complete a byte; shift them |
| 237 | // left into position |
| 238 | $newBits = $charmap[$encodedString[$c]] << $bitsNeeded |
| 239 | - $bitsPerCharacter; |
| 240 | $bitsWritten += $bitsPerCharacter; |
| 241 | |
| 242 | } elseif ($c != $lastNotatedIndex || $rightPadFinalBits) { |
| 243 | // Zero or more too many bits to complete a byte; |
| 244 | // shift right |
| 245 | $newBits = $charmap[$encodedString[$c]] >> $unusedBitCount; |
| 246 | $bitsWritten = 8; //$bitsWritten += $bitsNeeded; |
| 247 | |
| 248 | } else { |
| 249 | // Final bits don't need to be shifted |
| 250 | $newBits = $charmap[$encodedString[$c]]; |
| 251 | $bitsWritten = 8; |
| 252 | } |
| 253 | |
| 254 | $byte |= $newBits; |
| 255 | |
| 256 | if ($bitsWritten == 8 || $c == $lastNotatedIndex) { |
| 257 | // Byte is ready to be written |
| 258 | $rawString .= pack('C', $byte); |
| 259 | |
| 260 | if ($c != $lastNotatedIndex) { |
| 261 | // Start the next byte |
| 262 | $bitsWritten = $unusedBitCount; |
| 263 | $byte = ($charmap[$encodedString[$c]] |
| 264 | ^ ($newBits << $unusedBitCount)) << 8 - $bitsWritten; |
| 265 | } |
| 266 | } |
| 267 | |
| 268 | } elseif ($strict) { |
| 269 | // Unable to decode character; abort |
| 270 | return NULL; |
| 271 | } |
| 272 | } |
| 273 | |
| 274 | return $rawString; |
| 275 | } |
| 276 | } |