| 1 |
Binary-to-Text Utilities for PHP |
| 2 |
================================= |
| 3 |
|
| 4 |
For now, the only class in this repository is **Base2n**. |
| 5 |
|
| 6 |
Base2n is for binary-to-text conversion with arbitrary encoding schemes that represent binary data in a base 2<sup>n</sup> notation. It can handle non-standard variants of many standard encoding schemes such as [Base64][rfc4648base64] and [Base32][rfc4648base32]. Many binary-to-text encoding schemes use a fixed number of bits of binary data to generate each encoded character. Such schemes generalize to a single algorithm, implemented here. |
| 7 |
|
| 8 |
[rfc4648base64]: http://tools.ietf.org/html/rfc4648#section-4 "RFC 4648 Base64 Specification" |
| 9 |
[rfc4648base32]: http://tools.ietf.org/html/rfc4648#section-6 "RFC 4648 Base32 Specification" |
| 10 |
|
| 11 |
Binary-to-text encoding is usually used to represent data in a notation that is safe for transport over text-based protocols, and there are several other practical uses. See the examples below. |
| 12 |
|
| 13 |
|
| 14 |
|
| 15 |
Basic Base2n Usage |
| 16 |
------------------ |
| 17 |
|
| 18 |
With Base2n, you define your encoding scheme parametrically. Let's instantiate a [Base32][rfc4648base32] encoder: |
| 19 |
|
| 20 |
```php |
| 21 |
// RFC 4648 base32 alphabet; case-insensitive |
| 22 |
$base32 = new Base2n(5, 'ABCDEFGHIJKLMNOPQRSTUVWXYZ234567', FALSE, TRUE, TRUE); |
| 23 |
$encoded = $base32->encode('encode this'); |
| 24 |
// MVXGG33EMUQHI2DJOM====== |
| 25 |
``` |
| 26 |
|
| 27 |
|
| 28 |
### Constructor Parameters |
| 29 |
|
| 30 |
- <code>integer $bitsPerCharacter</code> **Required**. The number of bits to use for each encoded character; 1–8. The most practical range is 1–6. The encoding's radix is a power of 2: <code>2^$bitsPerCharacter</code>. |
| 31 |
1. [base-2, binary][binary] |
| 32 |
2. [base-4, quaternary][quaternary] |
| 33 |
3. [base-8, octal][octal] |
| 34 |
4. [base-16, hexadecimal][hexadecimal] |
| 35 |
5. [base-32][base32] |
| 36 |
6. [base-64][base64] |
| 37 |
7. base-128 |
| 38 |
8. base-256 |
| 39 |
|
| 40 |
[binary]: http://en.wikipedia.org/wiki/Binary_numeral_system "Binary Notation" |
| 41 |
[quaternary]: http://en.wikipedia.org/wiki/Quaternary_numeral_system "Base-2 Notation" |
| 42 |
[octal]: http://en.wikipedia.org/wiki/Octal "Octal Notation" |
| 43 |
[hexadecimal]: http://en.wikipedia.org/wiki/Base16 "Hexadecimal Notation" |
| 44 |
[base32]: http://en.wikipedia.org/wiki/Base32 "Base32 Encoding" |
| 45 |
[base64]: http://en.wikipedia.org/wiki/Base64 "Base64 Encoding" |
| 46 |
|
| 47 |
- <code>string $chars</code> This string specifies the base alphabet. Must be <code>2^$bitsPerCharacter</code> long. Default: <code>0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-_</code> |
| 48 |
|
| 49 |
- <code>boolean $caseSensitive</code> To decode in a case-sensitive manner. Default: <code>FALSE</code> |
| 50 |
|
| 51 |
- <code>boolean $rightPadFinalBits</code> How to encode the last character when the bits remaining are fewer than <code>$bitsPerCharacter</code>. When <code>TRUE</code>, the bits to encode are placed in the most significant position of the final group of bits, with the lower bits set to <code>0</code>. When <code>FALSE</code>, the final bits are placed in the least significant position. For [RFC 4648][rfc4648] encodings, <code>$rightPadFinalBits</code>should be <code>TRUE</code>. Default: <code>FALSE</code> |
| 52 |
|
| 53 |
[rfc4648]: http://tools.ietf.org/html/rfc4648 "RFC 4648: Base16, Base32, Base64" |
| 54 |
|
| 55 |
- <code>boolean $padFinalGroup</code> It's common to encode characters in groups. For example, Base64 (which is based on 6 bits per character) converts 3 raw bytes into 4 encoded characters. If insufficient bytes remain at the end, the final group will be padded with <code>=</code> to complete a group of 4 characters, and the encoded length is always a multiple of 4. Although the information provided by the padding is redundant, some programs rely on it for decoding; Base2n does not. Default: <code>FALSE</code> |
| 56 |
|
| 57 |
- <code>string $padCharacter</code> When <code>$padFinalGroup</code> is <code>TRUE</code>, this is the pad character used. Default: <code>=</code> |
| 58 |
|
| 59 |
|
| 60 |
### <code>encode()</code> Parameters |
| 61 |
|
| 62 |
- <code>string $rawString</code> **Required**. The data to be encoded. |
| 63 |
|
| 64 |
|
| 65 |
### <code>decode()</code> Parameters |
| 66 |
|
| 67 |
- <code>string $encodedString</code> **Required**. The string to be decoded. |
| 68 |
- <code>boolean $strict</code> When <code>TRUE</code>, <code>NULL</code> will be returned if <code>$encodedString</code> contains an undecodable character. When <code>FALSE</code>, unknown characters are simply ignored. Default: <code>FALSE</code> |
| 69 |
|
| 70 |
|
| 71 |
|
| 72 |
Examples |
| 73 |
-------- |
| 74 |
|
| 75 |
PHP does not provide any Base32 encoding functions. By setting <code>$bitsPerCharacter</code> to 5 and specifying your desired alphabet in <code>$chars</code>, you can handle any variant of Base32: |
| 76 |
|
| 77 |
```php |
| 78 |
// RFC 4648 base32 alphabet; case-insensitive |
| 79 |
$base32 = new Base2n(5, 'ABCDEFGHIJKLMNOPQRSTUVWXYZ234567', FALSE, TRUE, TRUE); |
| 80 |
$encoded = $base32->encode('encode this'); |
| 81 |
// MVXGG33EMUQHI2DJOM====== |
| 82 |
``` |
| 83 |
|
| 84 |
```php |
| 85 |
// RFC 4648 base32hex alphabet |
| 86 |
$base32hex = new Base2n(5, '0123456789ABCDEFGHIJKLMNOPQRSTUV', FALSE, TRUE, TRUE); |
| 87 |
$encoded = $base32hex->encode('encode this'); |
| 88 |
// CLN66RR4CKG78Q39EC====== |
| 89 |
``` |
| 90 |
|
| 91 |
|
| 92 |
Octal notation: |
| 93 |
|
| 94 |
```php |
| 95 |
$octal = new Base2n(3); |
| 96 |
$encoded = $octal->encode('encode this'); |
| 97 |
// 312671433366214510072150322711 |
| 98 |
``` |
| 99 |
|
| 100 |
|
| 101 |
A convenient way to go back and forth between binary notation and its real binary representation: |
| 102 |
|
| 103 |
```php |
| 104 |
$binary = new Base2n(1); |
| 105 |
$encoded = $binary->encode('encode this'); |
| 106 |
// 0110010101101110011000110110111101100100011001010010000001110100011010000110100101110011 |
| 107 |
$decoded = $binary->decode($encoded); |
| 108 |
// encode this |
| 109 |
``` |
| 110 |
|
| 111 |
|
| 112 |
PHP uses a proprietary binary-to-text encoding scheme to generate session identifiers from random hash digests. The most efficient way to store these session IDs in a database is to decode them back to their raw hash digests. PHP's encoding scheme is configured with the <code>[session.hash_bits_per_character][phphashbits]</code> php.ini setting. The decoded size depends on the hash function, set with <code>[session.hash_function][phphash]</code> in php.ini. |
| 113 |
|
| 114 |
```php |
| 115 |
// session.hash_function = 0 |
| 116 |
// session.hash_bits_per_character = 5 |
| 117 |
// 128-bit session ID |
| 118 |
$sessionId = 'q3c8n4vqpq11i0vr6ucmafg1h3'; |
| 119 |
// Decodes to 16 bytes |
| 120 |
$phpBase32 = new Base2n(5, '0123456789abcdefghijklmnopqrstuv'); |
| 121 |
$rawSessionId = $phpBase32->decode($sessionId); |
| 122 |
``` |
| 123 |
|
| 124 |
```php |
| 125 |
// session.hash_function = 1 |
| 126 |
// session.hash_bits_per_character = 6 |
| 127 |
// 160-bit session ID |
| 128 |
$sessionId = '7Hf91mVc,q-9W1VndNNh3evVN83'; |
| 129 |
// Decodes to 20 bytes |
| 130 |
$phpBase64 = new Base2n(6, '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-,'); |
| 131 |
$rawSessionId = $phpBase64->decode($sessionId); |
| 132 |
``` |
| 133 |
|
| 134 |
[phphashbits]: http://php.net/manual/en/session.configuration.php#ini.session.hash-bits-per-character "PHP session.hash_bits_per_character" |
| 135 |
[phphash]: http://php.net/manual/en/session.configuration.php#ini.session.hash-function "PHP session.hash_function" |
| 136 |
|
| 137 |
|
| 138 |
Generate random security tokens: |
| 139 |
```php |
| 140 |
$tokenEncoder = new Base2n(6, '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-,'); |
| 141 |
$binaryToken = openssl_random_pseudo_bytes(32); // PHP >= 5.3 |
| 142 |
$token = $tokenEncoder->encode($binaryToken); |
| 143 |
// Example: U6M132v9FG-AHhBVaQWOg1gjyUi1IogNxuen0i3u3ep |
| 144 |
``` |
| 145 |
|
| 146 |
|
| 147 |
The rest of these examples are probably more fun than they are practical. |
| 148 |
|
| 149 |
|
| 150 |
We can encode arbitrary data with a 7-bit encoding. (Note that this is not the same as the [7bit MIME content-transfer-encoding][7bit].) |
| 151 |
```php |
| 152 |
// This uses all 7-bit ASCII characters |
| 153 |
$base128chars = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D\x0E\x0F" |
| 154 |
. "\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C\x1D\x1E\x1F" |
| 155 |
. "\x20\x21\x22\x23\x24\x25\x26\x27\x28\x29\x2A\x2B\x2C\x2D\x2E\x2F" |
| 156 |
. "\x30\x31\x32\x33\x34\x35\x36\x37\x38\x39\x3A\x3B\x3C\x3D\x3E\x3F" |
| 157 |
. "\x40\x41\x42\x43\x44\x45\x46\x47\x48\x49\x4A\x4B\x4C\x4D\x4E\x4F" |
| 158 |
. "\x50\x51\x52\x53\x54\x55\x56\x57\x58\x59\x5A\x5B\x5C\x5D\x5E\x5F" |
| 159 |
. "\x60\x61\x62\x63\x64\x65\x66\x67\x68\x69\x6A\x6B\x6C\x6D\x6E\x6F" |
| 160 |
. "\x70\x71\x72\x73\x74\x75\x76\x77\x78\x69\x7A\x7B\x7C\x7D\x7E\x7F"; |
| 161 |
|
| 162 |
$base128 = new Base2n(7, $base128chars); |
| 163 |
$encoded = $base128->encode('encode this'); |
| 164 |
``` |
| 165 |
[7bit]: http://msdn.microsoft.com/en-us/library/ms526290(v=exchg.10).aspx "7bit MIME Content-Transfer-Encoding" |
| 166 |
|
| 167 |
|
| 168 |
The following encoding guarantees that the most significant bit is set for every byte: |
| 169 |
```php |
| 170 |
// "High" base-128 encoding |
| 171 |
$high128chars = "\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8A\x8B\x8C\x8D\x8E\x8F" |
| 172 |
. "\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9A\x9B\x9C\x9D\x9E\x9F" |
| 173 |
. "\xA0\xA1\xA2\xA3\xA4\xA5\xA6\xA7\xA8\xA9\xAA\xAB\xAC\xAD\xAE\xAF" |
| 174 |
. "\xB0\xB1\xB2\xB3\xB4\xB5\xB6\xB7\xB8\xB9\xBA\xBB\xBC\xBD\xBE\xBF" |
| 175 |
. "\xC0\xC1\xC2\xC3\xC4\xC5\xC6\xC7\xC8\xC9\xCA\xCB\xCC\xCD\xCE\xCF" |
| 176 |
. "\xD0\xD1\xD2\xD3\xD4\xD5\xD6\xD7\xD8\xD9\xDA\xDB\xDC\xDD\xDE\xDF" |
| 177 |
. "\xE0\xE1\xE2\xE3\xE4\xE5\xE6\xE7\xE8\xE9\xEA\xEB\xEC\xED\xEE\xEF" |
| 178 |
. "\xF0\xF1\xF2\xF3\xF4\xF5\xF6\xF7\xF8\xF9\xFA\xFB\xFC\xFD\xFE\xFF"; |
| 179 |
|
| 180 |
$high128 = new Base2n(7, $high128chars); |
| 181 |
$encoded = $high128->encode('encode this'); |
| 182 |
``` |
| 183 |
|
| 184 |
|
| 185 |
Let's create an encoding using exclusively non-printable control characters! |
| 186 |
```php |
| 187 |
// Base-32 non-printable character encoding |
| 188 |
$noPrintChars = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D\x0E\x0F" |
| 189 |
. "\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C\x1D\x1E\x1F"; |
| 190 |
|
| 191 |
$nonPrintable32 = new Base2n(5, $noPrintChars); |
| 192 |
$encoded = $nonPrintable32->encode('encode this'); |
| 193 |
``` |
| 194 |
|
| 195 |
|
| 196 |
Why not encode data using only whitespace? Here's a base-4 encoding using space, tab, new line, and carriage return: |
| 197 |
```php |
| 198 |
// Base-4 whitespace encoding |
| 199 |
$whitespaceChars = " \t\n\r"; |
| 200 |
|
| 201 |
$whitespace = new Base2n(2, $whitespaceChars); |
| 202 |
$encoded = $whitespace->encode('encode this'); |
| 203 |
// "\t\n\t\t\t\n\r\n\t\n \r\t\n\r\r\t\n\t \t\n\t\t \n \t\r\t \t\n\n \t\n\n\t\t\r \r" |
| 204 |
|
| 205 |
$decoded = $whitespace->decode( |
| 206 |
"\t\n\t\t\t\n\r\n\t\n \r\t\n\r\r\t\n\t \t\n\t\t \n \t\r\t \t\n\n \t\n\n\t\t\r \r" |
| 207 |
); |
| 208 |
// encode this |
| 209 |
``` |
| 210 |
|
| 211 |
|
| 212 |
|
| 213 |
Counterexamples |
| 214 |
---------------- |
| 215 |
|
| 216 |
Base2n is not slow, but it will never outperform an encoding function implemented in C. When one exists, use it instead. |
| 217 |
|
| 218 |
|
| 219 |
PHP provides the <code>[base64_encode()][base64_encode]</code> and <code>[base64_decode()][base64_decode]</code> functions, and you should always use them for standard Base64. When you need to use a modified alphabet, you can translate the encoded output with <code>[strtr()][strtr]</code> or <code>[str_replace()][str_replace]</code>. |
| 220 |
|
| 221 |
[base64_encode]: http://php.net/base64_encode "PHP base64_encode() Function" |
| 222 |
[base64_decode]: http://php.net/base64_decode "PHP base64_decode() Function" |
| 223 |
[strtr]: http://php.net/strtr "PHP strtr() Function" |
| 224 |
[str_replace]: http://php.net/str_replace "PHP str_replace() Function" |
| 225 |
|
| 226 |
A common variant of Base64 is [modified for URLs and filenames][rfc4648base64url], where <code>+</code> and <code>/</code> are replaced with <code>-</code> and <code>_</code>, and the <code>=</code> padding is omitted. It's better to handle this variant with native PHP functions: |
| 227 |
|
| 228 |
```php |
| 229 |
// RFC 4648 base64url with Base2n... |
| 230 |
$base64url = new Base2n(6, 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_', TRUE, TRUE, FALSE); |
| 231 |
$encoded = $base64url->encode("encode this \xBF\xC2\xBF"); |
| 232 |
// ZW5jb2RlIHRoaXMgv8K_ |
| 233 |
|
| 234 |
// RFC 4648 base64url with native functions... |
| 235 |
$encoded = str_replace(array('+', '/', '='), array('-', '_', ''), base64_encode("encode this \xBF\xC2\xBF")); |
| 236 |
// ZW5jb2RlIHRoaXMgv8K_ |
| 237 |
``` |
| 238 |
|
| 239 |
[rfc4648base64url]: http://tools.ietf.org/html/rfc4648#page-7 "Modified Base64 for URLs" |
| 240 |
|
| 241 |
|
| 242 |
Native functions get slightly more cumbersome when every position in the alphabet has changed, as seen in this example of [decoding a Bcrypt hash][bmcf]: |
| 243 |
```php |
| 244 |
// Decode the salt and digest from a Bcrypt hash |
| 245 |
|
| 246 |
$hash = '$2y$14$i5btSOiulHhaPHPbgNUGdObga/GC.AVG/y5HHY1ra7L0C9dpCaw8u'; |
| 247 |
$encodedSalt = substr($hash, 7, 22); |
| 248 |
$encodedDigest = substr($hash, 29, 31); |
| 249 |
|
| 250 |
// Using Base2n... |
| 251 |
$bcrypt64 = new Base2n(6, './ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789', TRUE, TRUE); |
| 252 |
$rawSalt = $bcrypt64->decode($encodedSalt); // 16 bytes |
| 253 |
$rawDigest = $bcrypt64->decode($encodedDigest); // 23 bytes |
| 254 |
|
| 255 |
// Using native functions... |
| 256 |
$bcrypt64alphabet = './ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789'; |
| 257 |
$base64alphabet = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'; |
| 258 |
$rawSalt = base64_decode(strtr($encodedSalt, $bcrypt64alphabet, $base64alphabet)); // 16 bytes |
| 259 |
$rawDigest = base64_decode(strtr($encodedDigest, $bcrypt64alphabet, $base64alphabet)); // 23 bytes |
| 260 |
``` |
| 261 |
|
| 262 |
[bmcf]: https://github.com/ademarre/binary-mcf "Binary Modular Crypt Format (BMCF)" |
| 263 |
|
| 264 |
You can encode and decode hexadecimal with <code>[bin2hex()][bin2hex]</code> and <code>[pack()][pack]</code>: |
| 265 |
|
| 266 |
```php |
| 267 |
// Hexadecimal with Base2n... |
| 268 |
$hexadecimal = new Base2n(4); |
| 269 |
$encoded = $hexadecimal->encode('encode this'); // 656e636f64652074686973 |
| 270 |
$decoded = $hexadecimal->decode($encoded); // encode this |
| 271 |
|
| 272 |
// It's better to use native functions... |
| 273 |
$encoded = bin2hex('encode this'); // 656e636f64652074686973 |
| 274 |
$decoded = pack('H*', $encoded); // encode this |
| 275 |
// As of PHP 5.4 you can use hex2bin() instead of pack() |
| 276 |
``` |
| 277 |
|
| 278 |
[bin2hex]: http://php.net/bin2hex "PHP bin2hex() Function" |
| 279 |
[pack]: http://php.net/pack "PHP pack() Function" |
| 280 |
|
