| 1 |
<?php |
| 2 |
/* 9 April 2008. version 1.1 |
| 3 |
* |
| 4 |
* This is the php version of the Dean Edwards JavaScript's Packer, |
| 5 |
* Based on : |
| 6 |
* |
| 7 |
* ParseMaster, version 1.0.2 (2005-08-19) Copyright 2005, Dean Edwards |
| 8 |
* a multi-pattern parser. |
| 9 |
* KNOWN BUG: erroneous behavior when using escapeChar with a replacement |
| 10 |
* value that is a function |
| 11 |
* |
| 12 |
* packer, version 2.0.2 (2005-08-19) Copyright 2004-2005, Dean Edwards |
| 13 |
* |
| 14 |
* License: http://creativecommons.org/licenses/LGPL/2.1/ |
| 15 |
* |
| 16 |
* Ported to PHP by Nicolas Martin. |
| 17 |
* |
| 18 |
* ---------------------------------------------------------------------- |
| 19 |
* changelog: |
| 20 |
* 1.1 : correct a bug, '\0' packed then unpacked becomes '\'. |
| 21 |
* ---------------------------------------------------------------------- |
| 22 |
* |
| 23 |
* examples of usage : |
| 24 |
* $myPacker = new JavaScriptPacker($script, 62, true, false); |
| 25 |
* $packed = $myPacker->pack(); |
| 26 |
* |
| 27 |
* or |
| 28 |
* |
| 29 |
* $myPacker = new JavaScriptPacker($script, 'Normal', true, false); |
| 30 |
* $packed = $myPacker->pack(); |
| 31 |
* |
| 32 |
* or (default values) |
| 33 |
* |
| 34 |
* $myPacker = new JavaScriptPacker($script); |
| 35 |
* $packed = $myPacker->pack(); |
| 36 |
* |
| 37 |
* |
| 38 |
* params of the constructor : |
| 39 |
* $script: the JavaScript to pack, string. |
| 40 |
* $encoding: level of encoding, int or string : |
| 41 |
* 0,10,62,95 or 'None', 'Numeric', 'Normal', 'High ASCII'. |
| 42 |
* default: 62. |
| 43 |
* $fastDecode: include the fast decoder in the packed result, boolean. |
| 44 |
* default : true. |
| 45 |
* $specialChars: if you are flagged your private and local variables |
| 46 |
* in the script, boolean. |
| 47 |
* default: false. |
| 48 |
* |
| 49 |
* The pack() method return the compressed JavasScript, as a string. |
| 50 |
* |
| 51 |
* see http://dean.edwards.name/packer/usage/ for more information. |
| 52 |
* |
| 53 |
* Notes : |
| 54 |
* # need PHP 5 . Tested with PHP 5.1.2, 5.1.3, 5.1.4, 5.2.3 |
| 55 |
* |
| 56 |
* # The packed result may be different than with the Dean Edwards |
| 57 |
* version, but with the same length. The reason is that the PHP |
| 58 |
* function usort to sort array don't necessarily preserve the |
| 59 |
* original order of two equal member. The Javascript sort function |
| 60 |
* in fact preserve this order (but that's not require by the |
| 61 |
* ECMAScript standard). So the encoded keywords order can be |
| 62 |
* different in the two results. |
| 63 |
* |
| 64 |
* # Be careful with the 'High ASCII' Level encoding if you use |
| 65 |
* UTF-8 in your files... |
| 66 |
*/ |
| 67 |
class JavaScriptPacker { |
| 68 |
// constants |
| 69 |
const IGNORE = '$1'; |
| 70 |
// validate parameters |
| 71 |
private $_script = ''; |
| 72 |
private $_encoding = 62; |
| 73 |
private $_fastDecode = true; |
| 74 |
private $_specialChars = false; |
| 75 |
private $LITERAL_ENCODING = array( |
| 76 |
'None' => 0, |
| 77 |
'Numeric' => 10, |
| 78 |
'Normal' => 62, |
| 79 |
'High ASCII' => 95 |
| 80 |
); |
| 81 |
public function __construct($_script, $_encoding = 62, $_fastDecode = true, $_specialChars = false) |
| 82 |
{ |
| 83 |
$this->_script = $_script . "\n"; |
| 84 |
if (array_key_exists($_encoding, $this->LITERAL_ENCODING)) |
| 85 |
$_encoding = $this->LITERAL_ENCODING[$_encoding]; |
| 86 |
$this->_encoding = min((int)$_encoding, 95); |
| 87 |
$this->_fastDecode = $_fastDecode; |
| 88 |
$this->_specialChars = $_specialChars; |
| 89 |
} |
| 90 |
public function pack() { |
| 91 |
$this->_addParser('_basicCompression'); |
| 92 |
if ($this->_specialChars) |
| 93 |
$this->_addParser('_encodeSpecialChars'); |
| 94 |
if ($this->_encoding) |
| 95 |
$this->_addParser('_encodeKeywords'); |
| 96 |
// go! |
| 97 |
return $this->_pack($this->_script); |
| 98 |
} |
| 99 |
// apply all parsing routines |
| 100 |
private function _pack($script) { |
| 101 |
for ($i = 0; isset($this->_parsers[$i]); $i++) { |
| 102 |
$script = call_user_func(array(&$this,$this->_parsers[$i]), $script); |
| 103 |
} |
| 104 |
return $script; |
| 105 |
} |
| 106 |
// keep a list of parsing functions, they'll be executed all at once |
| 107 |
private $_parsers = array(); |
| 108 |
private function _addParser($parser) { |
| 109 |
$this->_parsers[] = $parser; |
| 110 |
} |
| 111 |
// zero encoding - just removal of white space and comments |
| 112 |
private function _basicCompression($script) { |
| 113 |
$parser = new ParseMaster(); |
| 114 |
// make safe |
| 115 |
$parser->escapeChar = '\\'; |
| 116 |
// protect strings |
| 117 |
$parser->add('/\'[^\'\\n\\r]*\'/', self::IGNORE); |
| 118 |
$parser->add('/"[^"\\n\\r]*"/', self::IGNORE); |
| 119 |
// remove comments |
| 120 |
$parser->add('/\\/\\/[^\\n\\r]*[\\n\\r]/', ' '); |
| 121 |
$parser->add('/\\/\\*[^*]*\\*+([^\\/][^*]*\\*+)*\\//', ' '); |
| 122 |
// protect regular expressions |
| 123 |
$parser->add('/\\s+(\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?)/', '$2'); // IGNORE |
| 124 |
$parser->add('/[^\\w\\x24\\/\'"*)\\?:]\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?/', self::IGNORE); |
| 125 |
// remove: ;;; doSomething(); |
| 126 |
if ($this->_specialChars) $parser->add('/;;;[^\\n\\r]+[\\n\\r]/'); |
| 127 |
// remove redundant semi-colons |
| 128 |
$parser->add('/\\(;;\\)/', self::IGNORE); // protect for (;;) loops |
| 129 |
$parser->add('/;+\\s*([};])/', '$2'); |
| 130 |
// apply the above |
| 131 |
$script = $parser->exec($script); |
| 132 |
// remove white-space |
| 133 |
$parser->add('/(\\b|\\x24)\\s+(\\b|\\x24)/', '$2 $3'); |
| 134 |
$parser->add('/([+\\-])\\s+([+\\-])/', '$2 $3'); |
| 135 |
$parser->add('/\\s+/', ''); |
| 136 |
// done |
| 137 |
return $parser->exec($script); |
| 138 |
} |
| 139 |
private function _encodeSpecialChars($script) { |
| 140 |
$parser = new ParseMaster(); |
| 141 |
// replace: $name -> n, $$name -> na |
| 142 |
$parser->add('/((\\x24+)([a-zA-Z$_]+))(\\d*)/', |
| 143 |
array('fn' => '_replace_name') |
| 144 |
); |
| 145 |
// replace: _name -> _0, double-underscore (__name) is ignored |
| 146 |
$regexp = '/\\b_[A-Za-z\\d]\\w*/'; |
| 147 |
// build the word list |
| 148 |
$keywords = $this->_analyze($script, $regexp, '_encodePrivate'); |
| 149 |
// quick ref |
| 150 |
$encoded = $keywords['encoded']; |
| 151 |
$parser->add($regexp, |
| 152 |
array( |
| 153 |
'fn' => '_replace_encoded', |
| 154 |
'data' => $encoded |
| 155 |
) |
| 156 |
); |
| 157 |
return $parser->exec($script); |
| 158 |
} |
| 159 |
private function _encodeKeywords($script) { |
| 160 |
// escape high-ascii values already in the script (i.e. in strings) |
| 161 |
if ($this->_encoding > 62) |
| 162 |
$script = $this->_escape95($script); |
| 163 |
// create the parser |
| 164 |
$parser = new ParseMaster(); |
| 165 |
$encode = $this->_getEncoder($this->_encoding); |
| 166 |
// for high-ascii, don't encode single character low-ascii |
| 167 |
$regexp = ($this->_encoding > 62) ? '/\\w\\w+/' : '/\\w+/'; |
| 168 |
// build the word list |
| 169 |
$keywords = $this->_analyze($script, $regexp, $encode); |
| 170 |
$encoded = $keywords['encoded']; |
| 171 |
// encode |
| 172 |
$parser->add($regexp, |
| 173 |
array( |
| 174 |
'fn' => '_replace_encoded', |
| 175 |
'data' => $encoded |
| 176 |
) |
| 177 |
); |
| 178 |
if (empty($script)) return $script; |
| 179 |
else { |
| 180 |
//$res = $parser->exec($script); |
| 181 |
//$res = $this->_bootStrap($res, $keywords); |
| 182 |
//return $res; |
| 183 |
return $this->_bootStrap($parser->exec($script), $keywords); |
| 184 |
} |
| 185 |
} |
| 186 |
private function _analyze($script, $regexp, $encode) { |
| 187 |
// analyse |
| 188 |
// retreive all words in the script |
| 189 |
$all = array(); |
| 190 |
preg_match_all($regexp, $script, $all); |
| 191 |
$_sorted = array(); // list of words sorted by frequency |
| 192 |
$_encoded = array(); // dictionary of word->encoding |
| 193 |
$_protected = array(); // instances of "protected" words |
| 194 |
$all = $all[0]; // simulate the javascript comportement of global match |
| 195 |
if (!empty($all)) { |
| 196 |
$unsorted = array(); // same list, not sorted |
| 197 |
$protected = array(); // "protected" words (dictionary of word->"word") |
| 198 |
$value = array(); // dictionary of charCode->encoding (eg. 256->ff) |
| 199 |
$this->_count = array(); // word->count |
| 200 |
$i = count($all); $j = 0; //$word = null; |
| 201 |
// count the occurrences - used for sorting later |
| 202 |
do { |
| 203 |
--$i; |
| 204 |
$word = '$' . $all[$i]; |
| 205 |
if (!isset($this->_count[$word])) { |
| 206 |
$this->_count[$word] = 0; |
| 207 |
$unsorted[$j] = $word; |
| 208 |
// make a dictionary of all of the protected words in this script |
| 209 |
// these are words that might be mistaken for encoding |
| 210 |
//if (is_string($encode) && method_exists($this, $encode)) |
| 211 |
$values[$j] = call_user_func(array(&$this, $encode), $j); |
| 212 |
$protected['$' . $values[$j]] = $j++; |
| 213 |
} |
| 214 |
// increment the word counter |
| 215 |
$this->_count[$word]++; |
| 216 |
} while ($i > 0); |
| 217 |
// prepare to sort the word list, first we must protect |
| 218 |
// words that are also used as codes. we assign them a code |
| 219 |
// equivalent to the word itself. |
| 220 |
// e.g. if "do" falls within our encoding range |
| 221 |
// then we store keywords["do"] = "do"; |
| 222 |
// this avoids problems when decoding |
| 223 |
$i = count($unsorted); |
| 224 |
do { |
| 225 |
$word = $unsorted[--$i]; |
| 226 |
if (isset($protected[$word]) /*!= null*/) { |
| 227 |
$_sorted[$protected[$word]] = substr($word, 1); |
| 228 |
$_protected[$protected[$word]] = true; |
| 229 |
$this->_count[$word] = 0; |
| 230 |
} |
| 231 |
} while ($i); |
| 232 |
// sort the words by frequency |
| 233 |
// Note: the javascript and php version of sort can be different : |
| 234 |
// in php manual, usort : |
| 235 |
// " If two members compare as equal, |
| 236 |
// their order in the sorted array is undefined." |
| 237 |
// so the final packed script is different of the Dean's javascript version |
| 238 |
// but equivalent. |
| 239 |
// the ECMAscript standard does not guarantee this behaviour, |
| 240 |
// and thus not all browsers (e.g. Mozilla versions dating back to at |
| 241 |
// least 2003) respect this. |
| 242 |
usort($unsorted, array(&$this, '_sortWords')); |
| 243 |
$j = 0; |
| 244 |
// because there are "protected" words in the list |
| 245 |
// we must add the sorted words around them |
| 246 |
do { |
| 247 |
if (!isset($_sorted[$i])) |
| 248 |
$_sorted[$i] = substr($unsorted[$j++], 1); |
| 249 |
$_encoded[$_sorted[$i]] = $values[$i]; |
| 250 |
} while (++$i < count($unsorted)); |
| 251 |
} |
| 252 |
return array( |
| 253 |
'sorted' => $_sorted, |
| 254 |
'encoded' => $_encoded, |
| 255 |
'protected' => $_protected); |
| 256 |
} |
| 257 |
private $_count = array(); |
| 258 |
private function _sortWords($match1, $match2) { |
| 259 |
return $this->_count[$match2] - $this->_count[$match1]; |
| 260 |
} |
| 261 |
// build the boot function used for loading and decoding |
| 262 |
private function _bootStrap($packed, $keywords) { |
| 263 |
$ENCODE = $this->_safeRegExp('$encode\\($count\\)'); |
| 264 |
// $packed: the packed script |
| 265 |
$packed = "'" . $this->_escape($packed) . "'"; |
| 266 |
// $ascii: base for encoding |
| 267 |
$ascii = min(count($keywords['sorted']), $this->_encoding); |
| 268 |
if ($ascii == 0) $ascii = 1; |
| 269 |
// $count: number of words contained in the script |
| 270 |
$count = count($keywords['sorted']); |
| 271 |
// $keywords: list of words contained in the script |
| 272 |
foreach ($keywords['protected'] as $i=>$value) { |
| 273 |
$keywords['sorted'][$i] = ''; |
| 274 |
} |
| 275 |
// convert from a string to an array |
| 276 |
ksort($keywords['sorted']); |
| 277 |
$keywords = "'" . implode('|',$keywords['sorted']) . "'.split('|')"; |
| 278 |
$encode = ($this->_encoding > 62) ? '_encode95' : $this->_getEncoder($ascii); |
| 279 |
$encode = $this->_getJSFunction($encode); |
| 280 |
$encode = preg_replace('/_encoding/','$ascii', $encode); |
| 281 |
$encode = preg_replace('/arguments\\.callee/','$encode', $encode); |
| 282 |
$inline = '\\$count' . ($ascii > 10 ? '.toString(\\$ascii)' : ''); |
| 283 |
// $decode: code snippet to speed up decoding |
| 284 |
if ($this->_fastDecode) { |
| 285 |
// create the decoder |
| 286 |
$decode = $this->_getJSFunction('_decodeBody'); |
| 287 |
if ($this->_encoding > 62) |
| 288 |
$decode = preg_replace('/\\\\w/', '[\\xa1-\\xff]', $decode); |
| 289 |
// perform the encoding inline for lower ascii values |
| 290 |
elseif ($ascii < 36) |
| 291 |
$decode = preg_replace($ENCODE, $inline, $decode); |
| 292 |
// special case: when $count==0 there are no keywords. I want to keep |
| 293 |
// the basic shape of the unpacking funcion so i'll frig the code... |
| 294 |
if ($count == 0) |
| 295 |
$decode = preg_replace($this->_safeRegExp('($count)\\s*=\\s*1'), '$1=0', $decode, 1); |
| 296 |
} |
| 297 |
// boot function |
| 298 |
$unpack = $this->_getJSFunction('_unpack'); |
| 299 |
if ($this->_fastDecode) { |
| 300 |
// insert the decoder |
| 301 |
$this->buffer = $decode; |
| 302 |
$unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastDecode'), $unpack, 1); |
| 303 |
} |
| 304 |
$unpack = preg_replace('/"/', "'", $unpack); |
| 305 |
if ($this->_encoding > 62) { // high-ascii |
| 306 |
// get rid of the word-boundaries for regexp matches |
| 307 |
$unpack = preg_replace('/\'\\\\\\\\b\'\s*\\+|\\+\s*\'\\\\\\\\b\'/', '', $unpack); |
| 308 |
} |
| 309 |
if ($ascii > 36 || $this->_encoding > 62 || $this->_fastDecode) { |
| 310 |
// insert the encode function |
| 311 |
$this->buffer = $encode; |
| 312 |
$unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastEncode'), $unpack, 1); |
| 313 |
} else { |
| 314 |
// perform the encoding inline |
| 315 |
$unpack = preg_replace($ENCODE, $inline, $unpack); |
| 316 |
} |
| 317 |
// pack the boot function too |
| 318 |
$unpackPacker = new JavaScriptPacker($unpack, 0, false, true); |
| 319 |
$unpack = $unpackPacker->pack(); |
| 320 |
// arguments |
| 321 |
$params = array($packed, $ascii, $count, $keywords); |
| 322 |
if ($this->_fastDecode) { |
| 323 |
$params[] = 0; |
| 324 |
$params[] = '{}'; |
| 325 |
} |
| 326 |
$params = implode(',', $params); |
| 327 |
// the whole thing |
| 328 |
return 'eval(' . $unpack . '(' . $params . "))\n"; |
| 329 |
} |
| 330 |
private $buffer; |
| 331 |
private function _insertFastDecode($match) { |
| 332 |
return '{' . $this->buffer . ';'; |
| 333 |
} |
| 334 |
private function _insertFastEncode($match) { |
| 335 |
return '{$encode=' . $this->buffer . ';'; |
| 336 |
} |
| 337 |
// mmm.. ..which one do i need ?? |
| 338 |
private function _getEncoder($ascii) { |
| 339 |
return $ascii > 10 ? $ascii > 36 ? $ascii > 62 ? |
| 340 |
'_encode95' : '_encode62' : '_encode36' : '_encode10'; |
| 341 |
} |
| 342 |
// zero encoding |
| 343 |
// characters: 0123456789 |
| 344 |
private function _encode10($charCode) { |
| 345 |
return $charCode; |
| 346 |
} |
| 347 |
// inherent base36 support |
| 348 |
// characters: 0123456789abcdefghijklmnopqrstuvwxyz |
| 349 |
private function _encode36($charCode) { |
| 350 |
return base_convert($charCode, 10, 36); |
| 351 |
} |
| 352 |
// hitch a ride on base36 and add the upper case alpha characters |
| 353 |
// characters: 0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ |
| 354 |
private function _encode62($charCode) { |
| 355 |
$res = ''; |
| 356 |
if ($charCode >= $this->_encoding) { |
| 357 |
$res = $this->_encode62((int)($charCode / $this->_encoding)); |
| 358 |
} |
| 359 |
$charCode = $charCode % $this->_encoding; |
| 360 |
if ($charCode > 35) |
| 361 |
return $res . chr($charCode + 29); |
| 362 |
else |
| 363 |
return $res . base_convert($charCode, 10, 36); |
| 364 |
} |
| 365 |
// use high-ascii values |
| 366 |
// characters: ¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄ� |
| 367 |
ÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþ |
| 368 |
private function _encode95($charCode) { |
| 369 |
$res = ''; |
| 370 |
if ($charCode >= $this->_encoding) |
| 371 |
$res = $this->_encode95($charCode / $this->_encoding); |
| 372 |
return $res . chr(($charCode % $this->_encoding) + 161); |
| 373 |
} |
| 374 |
private function _safeRegExp($string) { |
| 375 |
return '/'.preg_replace('/\$/', '\\\$', $string).'/'; |
| 376 |
} |
| 377 |
private function _encodePrivate($charCode) { |
| 378 |
return "_" . $charCode; |
| 379 |
} |
| 380 |
// protect characters used by the parser |
| 381 |
private function _escape($script) { |
| 382 |
return preg_replace('/([\\\\\'])/', '\\\$1', $script); |
| 383 |
} |
| 384 |
// protect high-ascii characters already in the script |
| 385 |
private function _escape95($script) { |
| 386 |
return preg_replace_callback( |
| 387 |
'/[\\xa1-\\xff]/', |
| 388 |
array(&$this, '_escape95Bis'), |
| 389 |
$script |
| 390 |
); |
| 391 |
} |
| 392 |
private function _escape95Bis($match) { |
| 393 |
return '\x'.((string)dechex(ord($match))); |
| 394 |
} |
| 395 |
private function _getJSFunction($aName) { |
| 396 |
if (defined('self::JSFUNCTION'.$aName)) |
| 397 |
return constant('self::JSFUNCTION'.$aName); |
| 398 |
else |
| 399 |
return ''; |
| 400 |
} |
| 401 |
// JavaScript Functions used. |
| 402 |
// Note : In Dean's version, these functions are converted |
| 403 |
// with 'String(aFunctionName);'. |
| 404 |
// This internal conversion complete the original code, ex : |
| 405 |
// 'while (aBool) anAction();' is converted to |
| 406 |
// 'while (aBool) { anAction(); }'. |
| 407 |
// The JavaScript functions below are corrected. |
| 408 |
// unpacking function - this is the boot strap function |
| 409 |
// data extracted from this packing routine is passed to |
| 410 |
// this function when decoded in the target |
| 411 |
// NOTE ! : without the ';' final. |
| 412 |
const JSFUNCTION_unpack = |
| 413 |
'function($packed, $ascii, $count, $keywords, $encode, $decode) { |
| 414 |
while ($count--) { |
| 415 |
if ($keywords[$count]) { |
| 416 |
$packed = $packed.replace(new RegExp(\'\\\\b\' + $encode($count) + \'\\\\b\', \'g\'), $keywords[$count]); |
| 417 |
} |
| 418 |
} |
| 419 |
return $packed; |
| 420 |
}'; |
| 421 |
/* |
| 422 |
'function($packed, $ascii, $count, $keywords, $encode, $decode) { |
| 423 |
while ($count--) |
| 424 |
if ($keywords[$count]) |
| 425 |
$packed = $packed.replace(new RegExp(\'\\\\b\' + $encode($count) + \'\\\\b\', \'g\'), $keywords[$count]); |
| 426 |
return $packed; |
| 427 |
}'; |
| 428 |
*/ |
| 429 |
// code-snippet inserted into the unpacker to speed up decoding |
| 430 |
const JSFUNCTION_decodeBody = |
| 431 |
//_decode = function() { |
| 432 |
// does the browser support String.replace where the |
| 433 |
// replacement value is a function? |
| 434 |
' if (!\'\'.replace(/^/, String)) { |
| 435 |
// decode all the values we need |
| 436 |
while ($count--) { |
| 437 |
$decode[$encode($count)] = $keywords[$count] || $encode($count); |
| 438 |
} |
| 439 |
// global replacement function |
| 440 |
$keywords = [function ($encoded) {return $decode[$encoded]}]; |
| 441 |
// generic match |
| 442 |
$encode = function () {return \'\\\\w+\'}; |
| 443 |
// reset the loop counter - we are now doing a global replace |
| 444 |
$count = 1; |
| 445 |
} |
| 446 |
'; |
| 447 |
//}; |
| 448 |
/* |
| 449 |
' if (!\'\'.replace(/^/, String)) { |
| 450 |
// decode all the values we need |
| 451 |
while ($count--) $decode[$encode($count)] = $keywords[$count] || $encode($count); |
| 452 |
// global replacement function |
| 453 |
$keywords = [function ($encoded) {return $decode[$encoded]}]; |
| 454 |
// generic match |
| 455 |
$encode = function () {return\'\\\\w+\'}; |
| 456 |
// reset the loop counter - we are now doing a global replace |
| 457 |
$count = 1; |
| 458 |
}'; |
| 459 |
*/ |
| 460 |
// zero encoding |
| 461 |
// characters: 0123456789 |
| 462 |
const JSFUNCTION_encode10 = |
| 463 |
'function($charCode) { |
| 464 |
return $charCode; |
| 465 |
}';//;'; |
| 466 |
// inherent base36 support |
| 467 |
// characters: 0123456789abcdefghijklmnopqrstuvwxyz |
| 468 |
const JSFUNCTION_encode36 = |
| 469 |
'function($charCode) { |
| 470 |
return $charCode.toString(36); |
| 471 |
}';//;'; |
| 472 |
// hitch a ride on base36 and add the upper case alpha characters |
| 473 |
// characters: 0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ |
| 474 |
const JSFUNCTION_encode62 = |
| 475 |
'function($charCode) { |
| 476 |
return ($charCode < _encoding ? \'\' : arguments.callee(parseInt($charCode / _encoding))) + |
| 477 |
(($charCode = $charCode % _encoding) > 35 ? String.fromCharCode($charCode + 29) : $charCode.toString(36)); |
| 478 |
}'; |
| 479 |
// use high-ascii values |
| 480 |
// characters: ¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄ� |
| 481 |
ÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþ |
| 482 |
const JSFUNCTION_encode95 = |
| 483 |
'function($charCode) { |
| 484 |
return ($charCode < _encoding ? \'\' : arguments.callee($charCode / _encoding)) + |
| 485 |
String.fromCharCode($charCode % _encoding + 161); |
| 486 |
}'; |
| 487 |
} |
| 488 |
class ParseMaster { |
| 489 |
public $ignoreCase = false; |
| 490 |
public $escapeChar = ''; |
| 491 |
// constants |
| 492 |
const EXPRESSION = 0; |
| 493 |
const REPLACEMENT = 1; |
| 494 |
const LENGTH = 2; |
| 495 |
// used to determine nesting levels |
| 496 |
private $GROUPS = '/\\(/';//g |
| 497 |
private $SUB_REPLACE = '/\\$\\d/'; |
| 498 |
private $INDEXED = '/^\\$\\d+$/'; |
| 499 |
private $TRIM = '/([\'"])\\1\\.(.*)\\.\\1\\1$/'; |
| 500 |
private $ESCAPE = '/\\\./';//g |
| 501 |
private $QUOTE = '/\'/'; |
| 502 |
private $DELETED = '/\\x01[^\\x01]*\\x01/';//g |
| 503 |
public function add($expression, $replacement = '') { |
| 504 |
// count the number of sub-expressions |
| 505 |
// - add one because each pattern is itself a sub-expression |
| 506 |
$length = 1 + preg_match_all($this->GROUPS, $this->_internalEscape((string)$expression), $out); |
| 507 |
// treat only strings $replacement |
| 508 |
if (is_string($replacement)) { |
| 509 |
// does the pattern deal with sub-expressions? |
| 510 |
if (preg_match($this->SUB_REPLACE, $replacement)) { |
| 511 |
// a simple lookup? (e.g. "$2") |
| 512 |
if (preg_match($this->INDEXED, $replacement)) { |
| 513 |
// store the index (used for fast retrieval of matched strings) |
| 514 |
$replacement = (int)(substr($replacement, 1)) - 1; |
| 515 |
} else { // a complicated lookup (e.g. "Hello $2 $1") |
| 516 |
// build a function to do the lookup |
| 517 |
$quote = preg_match($this->QUOTE, $this->_internalEscape($replacement)) |
| 518 |
? '"' : "'"; |
| 519 |
$replacement = array( |
| 520 |
'fn' => '_backReferences', |
| 521 |
'data' => array( |
| 522 |
'replacement' => $replacement, |
| 523 |
'length' => $length, |
| 524 |
'quote' => $quote |
| 525 |
) |
| 526 |
); |
| 527 |
} |
| 528 |
} |
| 529 |
} |
| 530 |
// pass the modified arguments |
| 531 |
if (!empty($expression)) $this->_add($expression, $replacement, $length); |
| 532 |
else $this->_add('/^$/', $replacement, $length); |
| 533 |
} |
| 534 |
public function exec($string) { |
| 535 |
// execute the global replacement |
| 536 |
$this->_escaped = array(); |
| 537 |
// simulate the _patterns.toSTring of Dean |
| 538 |
$regexp = '/'; |
| 539 |
foreach ($this->_patterns as $reg) { |
| 540 |
$regexp .= '(' . substr($reg[self::EXPRESSION], 1, -1) . ')|'; |
| 541 |
} |
| 542 |
$regexp = substr($regexp, 0, -1) . '/'; |
| 543 |
$regexp .= ($this->ignoreCase) ? 'i' : ''; |
| 544 |
$string = $this->_escape($string, $this->escapeChar); |
| 545 |
$string = preg_replace_callback( |
| 546 |
$regexp, |
| 547 |
array( |
| 548 |
&$this, |
| 549 |
'_replacement' |
| 550 |
), |
| 551 |
$string |
| 552 |
); |
| 553 |
$string = $this->_unescape($string, $this->escapeChar); |
| 554 |
return preg_replace($this->DELETED, '', $string); |
| 555 |
} |
| 556 |
public function reset() { |
| 557 |
// clear the patterns collection so that this object may be re-used |
| 558 |
$this->_patterns = array(); |
| 559 |
} |
| 560 |
// private |
| 561 |
private $_escaped = array(); // escaped characters |
| 562 |
private $_patterns = array(); // patterns stored by index |
| 563 |
// create and add a new pattern to the patterns collection |
| 564 |
private function _add() { |
| 565 |
$arguments = func_get_args(); |
| 566 |
$this->_patterns[] = $arguments; |
| 567 |
} |
| 568 |
// this is the global replace function (it's quite complicated) |
| 569 |
private function _replacement($arguments) { |
| 570 |
if (empty($arguments)) return ''; |
| 571 |
$i = 1; $j = 0; |
| 572 |
// loop through the patterns |
| 573 |
while (isset($this->_patterns[$j])) { |
| 574 |
$pattern = $this->_patterns[$j++]; |
| 575 |
// do we have a result? |
| 576 |
if (isset($arguments[$i]) && ($arguments[$i] != '')) { |
| 577 |
$replacement = $pattern[self::REPLACEMENT]; |
| 578 |
if (is_array($replacement) && isset($replacement['fn'])) { |
| 579 |
if (isset($replacement['data'])) $this->buffer = $replacement['data']; |
| 580 |
return call_user_func(array(&$this, $replacement['fn']), $arguments, $i); |
| 581 |
} elseif (is_int($replacement)) { |
| 582 |
return $arguments[$replacement + $i]; |
| 583 |
} |
| 584 |
$delete = ($this->escapeChar == '' || |
| 585 |
strpos($arguments[$i], $this->escapeChar) === false) |
| 586 |
? '' : "\x01" . $arguments[$i] . "\x01"; |
| 587 |
return $delete . $replacement; |
| 588 |
// skip over references to sub-expressions |
| 589 |
} else { |
| 590 |
$i += $pattern[self::LENGTH]; |
| 591 |
} |
| 592 |
} |
| 593 |
} |
| 594 |
private function _backReferences($match, $offset) { |
| 595 |
$replacement = $this->buffer['replacement']; |
| 596 |
$quote = $this->buffer['quote']; |
| 597 |
$i = $this->buffer['length']; |
| 598 |
while ($i) { |
| 599 |
$replacement = str_replace('$'.$i--, $match[$offset + $i], $replacement); |
| 600 |
} |
| 601 |
return $replacement; |
| 602 |
} |
| 603 |
private function _replace_name($match, $offset){ |
| 604 |
$length = strlen($match[$offset + 2]); |
| 605 |
$start = $length - max($length - strlen($match[$offset + 3]), 0); |
| 606 |
return substr($match[$offset + 1], $start, $length) . $match[$offset + 4]; |
| 607 |
} |
| 608 |
private function _replace_encoded($match, $offset) { |
| 609 |
return $this->buffer[$match[$offset]]; |
| 610 |
} |
| 611 |
// php : we cannot pass additional data to preg_replace_callback, |
| 612 |
// and we cannot use &$this in create_function, so let's go to lower level |
| 613 |
private $buffer; |
| 614 |
// encode escaped characters |
| 615 |
private function _escape($string, $escapeChar) { |
| 616 |
if ($escapeChar) { |
| 617 |
$this->buffer = $escapeChar; |
| 618 |
return preg_replace_callback( |
| 619 |
'/\\' . $escapeChar . '(.)' .'/', |
| 620 |
array(&$this, '_escapeBis'), |
| 621 |
$string |
| 622 |
); |
| 623 |
} else { |
| 624 |
return $string; |
| 625 |
} |
| 626 |
} |
| 627 |
private function _escapeBis($match) { |
| 628 |
$this->_escaped[] = $match[1]; |
| 629 |
return $this->buffer; |
| 630 |
} |
| 631 |
// decode escaped characters |
| 632 |
private function _unescape($string, $escapeChar) { |
| 633 |
if ($escapeChar) { |
| 634 |
$regexp = '/'.'\\'.$escapeChar.'/'; |
| 635 |
$this->buffer = array('escapeChar'=> $escapeChar, 'i' => 0); |
| 636 |
return preg_replace_callback |
| 637 |
( |
| 638 |
$regexp, |
| 639 |
array(&$this, '_unescapeBis'), |
| 640 |
$string |
| 641 |
); |
| 642 |
} else { |
| 643 |
return $string; |
| 644 |
} |
| 645 |
} |
| 646 |
private function _unescapeBis() { |
| 647 |
if (isset($this->_escaped[$this->buffer['i']]) |
| 648 |
&& $this->_escaped[$this->buffer['i']] != '') |
| 649 |
{ |
| 650 |
$temp = $this->_escaped[$this->buffer['i']]; |
| 651 |
} else { |
| 652 |
$temp = ''; |
| 653 |
} |
| 654 |
$this->buffer['i']++; |
| 655 |
return $this->buffer['escapeChar'] . $temp; |
| 656 |
} |
| 657 |
private function _internalEscape($string) { |
| 658 |
return preg_replace($this->ESCAPE, '', $string); |
| 659 |
} |
| 660 |
} |
| 661 |
?> |
