Total Complexity | 316 |
Total Lines | 1001 |
Duplicated Lines | 0 % |
Changes | 0 |
Complex classes like JSqueeze often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use JSqueeze, and based on these observations, apply Extract Interface, too.
1 | <?php |
||
58 | class JSqueeze |
||
59 | { |
||
60 | const |
||
61 | |||
62 | SPECIAL_VAR_PACKER = '(\$+[a-zA-Z_]|_[a-zA-Z0-9$])[a-zA-Z0-9_$]*'; |
||
63 | |||
64 | public |
||
65 | |||
66 | $charFreq; |
||
67 | |||
68 | protected |
||
69 | |||
70 | $strings, |
||
71 | $closures, |
||
72 | $str0, |
||
73 | $str1, |
||
74 | $argFreq, |
||
75 | $specialVarRx, |
||
76 | $keepImportantComments, |
||
77 | |||
78 | $varRx = '(?:[a-zA-Z_$])[a-zA-Z0-9_$]*', |
||
79 | $reserved = array( |
||
80 | // Literals |
||
81 | 'true','false','null', |
||
82 | // ES6 |
||
83 | 'break','case','class','catch','const','continue','debugger','default','delete','do','else','export','extends','finally','for','function','if','import','in','instanceof','new','return','super','switch','this','throw','try','typeof','var','void','while','with','yield', |
||
84 | // Future |
||
85 | 'enum', |
||
86 | // Strict mode |
||
87 | 'implements','package','protected','static','let','interface','private','public', |
||
88 | // Module |
||
89 | 'await', |
||
90 | // Older standards |
||
91 | 'abstract','boolean','byte','char','double','final','float','goto','int','long','native','short','synchronized','throws','transient','volatile', |
||
92 | ); |
||
93 | |||
94 | public function __construct() |
||
98 | } |
||
99 | |||
100 | /** |
||
101 | * Squeezes a JavaScript source code. |
||
102 | * |
||
103 | * Set $singleLine to false if you want optional |
||
104 | * semi-colons to be replaced by line feeds. |
||
105 | * |
||
106 | * Set $keepImportantComments to false if you want /*! comments to be removed. |
||
107 | * |
||
108 | * $specialVarRx defines the regular expression of special variables names |
||
109 | * for global vars, methods, properties and in string substitution. |
||
110 | * Set it to false if you don't want any. |
||
111 | * |
||
112 | * If the analysed javascript source contains a single line comment like |
||
113 | * this one, then the directive will overwrite $specialVarRx: |
||
114 | * |
||
115 | * // jsqueeze.specialVarRx = your_special_var_regexp_here |
||
116 | * |
||
117 | * Only the first directive is parsed, others are ignored. It is not possible |
||
118 | * to redefine $specialVarRx in the middle of the javascript source. |
||
119 | * |
||
120 | * Example: |
||
121 | * $parser = new JSqueeze; |
||
122 | * $squeezed_js = $parser->squeeze($fat_js); |
||
123 | */ |
||
124 | public function squeeze($code, $singleLine = true, $keepImportantComments = true, $specialVarRx = false) |
||
125 | { |
||
126 | $code = trim($code); |
||
127 | if ('' === $code) { |
||
128 | return ''; |
||
129 | } |
||
130 | |||
131 | $this->argFreq = array(-1 => 0); |
||
132 | $this->specialVarRx = $specialVarRx; |
||
133 | $this->keepImportantComments = !!$keepImportantComments; |
||
134 | |||
135 | if (preg_match("#//[ \t]*jsqueeze\.specialVarRx[ \t]*=[ \t]*([\"']?)(.*)\\1#i", $code, $key)) { |
||
136 | if (!$key[1]) { |
||
137 | $key[2] = trim($key[2]); |
||
138 | $key[1] = strtolower($key[2]); |
||
139 | $key[1] = $key[1] && $key[1] != 'false' && $key[1] != 'none' && $key[1] != 'off'; |
||
140 | } |
||
141 | |||
142 | $this->specialVarRx = $key[1] ? $key[2] : false; |
||
143 | } |
||
144 | |||
145 | // Remove capturing parentheses |
||
146 | $this->specialVarRx && $this->specialVarRx = preg_replace('/(?<!\\\\)((?:\\\\\\\\)*)\((?!\?)/', '(?:', $this->specialVarRx); |
||
147 | |||
148 | false !== strpos($code, "\r") && $code = strtr(str_replace("\r\n", "\n", $code), "\r", "\n"); |
||
149 | false !== strpos($code, "\xC2\x85") && $code = str_replace("\xC2\x85", "\n", $code); // Next Line |
||
150 | false !== strpos($code, "\xE2\x80\xA8") && $code = str_replace("\xE2\x80\xA8", "\n", $code); // Line Separator |
||
151 | false !== strpos($code, "\xE2\x80\xA9") && $code = str_replace("\xE2\x80\xA9", "\n", $code); // Paragraph Separator |
||
152 | |||
153 | list($code, $this->strings) = $this->extractStrings($code); |
||
154 | list($code, $this->closures) = $this->extractClosures($code); |
||
155 | |||
156 | $key = "//''\"\"#0'"; // This crap has a wonderful property: it can not happen in any valid javascript, even in strings |
||
157 | $this->closures[$key] = &$code; |
||
158 | |||
159 | $tree = array($key => array('parent' => false)); |
||
160 | $this->makeVars($code, $tree[$key], $key); |
||
161 | $this->renameVars($tree[$key], true); |
||
162 | |||
163 | $code = substr($tree[$key]['code'], 1); |
||
164 | $code = preg_replace("'\breturn !'", 'return!', $code); |
||
165 | $code = preg_replace("'\}(?=(else|while)[^\$.a-zA-Z0-9_])'", "}\r", $code); |
||
166 | $code = str_replace(array_keys($this->strings), array_values($this->strings), $code); |
||
167 | |||
168 | if ($singleLine) { |
||
169 | $code = strtr($code, "\n", ';'); |
||
170 | } else { |
||
171 | $code = str_replace("\n", ";\n", $code); |
||
172 | } |
||
173 | false !== strpos($code, "\r") && $code = strtr(trim($code), "\r", "\n"); |
||
174 | |||
175 | // Cleanup memory |
||
176 | $this->charFreq = array_fill(0, 256, 0); |
||
177 | $this->strings = $this->closures = $this->argFreq = array(); |
||
178 | $this->str0 = $this->str1 = ''; |
||
179 | |||
180 | return $code; |
||
181 | } |
||
182 | |||
183 | protected function extractStrings($f) |
||
184 | { |
||
185 | if ($cc_on = false !== strpos($f, '@cc_on')) { |
||
186 | // Protect conditional comments from being removed |
||
187 | $f = str_replace('#', '##', $f); |
||
188 | $f = str_replace('/*@', '1#@', $f); |
||
189 | $f = preg_replace("'//@([^\n]+)'", '2#@$1@#3', $f); |
||
190 | $f = str_replace('@*/', '@#1', $f); |
||
191 | } |
||
192 | |||
193 | $len = strlen($f); |
||
194 | $code = str_repeat(' ', $len); |
||
195 | $j = 0; |
||
196 | |||
197 | $strings = array(); |
||
198 | $K = 0; |
||
199 | |||
200 | $instr = false; |
||
201 | |||
202 | $q = array( |
||
203 | "'", '"', |
||
204 | "'" => 0, |
||
205 | '"' => 0, |
||
206 | ); |
||
207 | |||
208 | // Extract strings, removes comments |
||
209 | for ($i = 0; $i < $len; ++$i) { |
||
210 | if ($instr) { |
||
211 | if ('//' == $instr) { |
||
212 | if ("\n" == $f[$i]) { |
||
213 | $f[$i--] = ' '; |
||
214 | $instr = false; |
||
215 | } |
||
216 | } elseif ($f[$i] == $instr || ('/' == $f[$i] && "/'" == $instr)) { |
||
217 | if ('!' == $instr) { |
||
218 | } elseif ('*' == $instr) { |
||
219 | if ('/' == $f[$i + 1]) { |
||
220 | ++$i; |
||
221 | $instr = false; |
||
222 | } |
||
223 | } else { |
||
224 | if ("/'" == $instr) { |
||
225 | while (isset($f[$i + 1]) && false !== strpos('gmi', $f[$i + 1])) { |
||
226 | $s[] = $f[$i++]; |
||
227 | } |
||
228 | $s[] = $f[$i]; |
||
229 | } |
||
230 | |||
231 | $instr = false; |
||
232 | } |
||
233 | } elseif ('*' == $instr) { |
||
234 | } elseif ('!' == $instr) { |
||
235 | if ('*' == $f[$i] && '/' == $f[$i + 1]) { |
||
236 | $s[] = "*/\r"; |
||
237 | ++$i; |
||
238 | $instr = false; |
||
239 | } elseif ("\n" == $f[$i]) { |
||
240 | $s[] = "\r"; |
||
241 | } else { |
||
242 | $s[] = $f[$i]; |
||
243 | } |
||
244 | } elseif ('\\' == $f[$i]) { |
||
245 | ++$i; |
||
246 | |||
247 | if ("\n" != $f[$i]) { |
||
248 | isset($q[$f[$i]]) && ++$q[$f[$i]]; |
||
249 | $s[] = '\\'.$f[$i]; |
||
250 | } |
||
251 | } elseif ('[' == $f[$i] && "/'" == $instr) { |
||
252 | $instr = '/['; |
||
253 | $s[] = '['; |
||
254 | } elseif (']' == $f[$i] && '/[' == $instr) { |
||
255 | $instr = "/'"; |
||
256 | $s[] = ']'; |
||
257 | } elseif ("'" == $f[$i] || '"' == $f[$i]) { |
||
258 | ++$q[$f[$i]]; |
||
259 | $s[] = '\\'.$f[$i]; |
||
260 | } else { |
||
261 | $s[] = $f[$i]; |
||
262 | } |
||
263 | } else { |
||
264 | switch ($f[$i]) { |
||
265 | case ';': |
||
266 | // Remove triple semi-colon |
||
267 | if ($i > 0 && ';' == $f[$i - 1] && $i + 1 < $len && ';' == $f[$i + 1]) { |
||
268 | $f[$i] = $f[$i + 1] = '/'; |
||
269 | } else { |
||
270 | $code[++$j] = ';'; |
||
271 | break; |
||
272 | } |
||
273 | |||
274 | case '/': |
||
275 | if ('*' == $f[$i + 1]) { |
||
276 | ++$i; |
||
277 | $instr = '*'; |
||
278 | |||
279 | if ($this->keepImportantComments && '!' == $f[$i + 1]) { |
||
280 | ++$i; |
||
281 | // no break here |
||
282 | } else { |
||
283 | break; |
||
284 | } |
||
285 | } elseif ('/' == $f[$i + 1]) { |
||
286 | ++$i; |
||
287 | $instr = '//'; |
||
288 | break; |
||
289 | } else { |
||
290 | $a = $j && (' ' == $code[$j] || "\x7F" == $code[$j]) ? $code[$j - 1] : $code[$j]; |
||
291 | if (false !== strpos('-!%&;<=>~:^+|,()*?[{} ', $a) |
||
292 | || (false !== strpos('oenfd', $a) |
||
293 | && preg_match( |
||
294 | "'(?<![\$.a-zA-Z0-9_])(do|else|return|typeof|yield[ \x7F]?\*?)[ \x7F]?$'", |
||
295 | substr($code, $j - 7, 8) |
||
296 | )) |
||
297 | ) { |
||
298 | if (')' === $a && $j > 1) { |
||
299 | $a = 1; |
||
300 | $k = $j - (' ' == $code[$j] || "\x7F" == $code[$j]) - 1; |
||
301 | while ($k >= 0 && $a) { |
||
302 | if ('(' === $code[$k]) { |
||
303 | --$a; |
||
304 | } elseif (')' === $code[$k]) { |
||
305 | ++$a; |
||
306 | } |
||
307 | --$k; |
||
308 | } |
||
309 | if (!preg_match("'(?<![\$.a-zA-Z0-9_])(if|for|while)[ \x7F]?$'", substr($code, 0, $k + 1))) { |
||
310 | $code[++$j] = '/'; |
||
311 | break; |
||
312 | } |
||
313 | } |
||
314 | |||
315 | $key = "//''\"\"".$K++.$instr = "/'"; |
||
316 | $a = $j; |
||
317 | $code .= $key; |
||
318 | while (isset($key[++$j - $a - 1])) { |
||
319 | $code[$j] = $key[$j - $a - 1]; |
||
320 | } |
||
321 | --$j; |
||
322 | isset($s) && ($s = implode('', $s)) && $cc_on && $this->restoreCc($s); |
||
323 | $strings[$key] = array('/'); |
||
324 | $s = &$strings[$key]; |
||
325 | } else { |
||
326 | $code[++$j] = '/'; |
||
327 | } |
||
328 | |||
329 | break; |
||
330 | } |
||
331 | |||
332 | case "'": |
||
333 | case '"': |
||
334 | $instr = $f[$i]; |
||
335 | $key = "//''\"\"".$K++.('!' == $instr ? ']' : "'"); |
||
336 | $a = $j; |
||
337 | $code .= $key; |
||
338 | while (isset($key[++$j - $a - 1])) { |
||
339 | $code[$j] = $key[$j - $a - 1]; |
||
340 | } |
||
341 | --$j; |
||
342 | isset($s) && ($s = implode('', $s)) && $cc_on && $this->restoreCc($s); |
||
343 | $strings[$key] = array(); |
||
344 | $s = &$strings[$key]; |
||
345 | '!' == $instr && $s[] = "\r/*!"; |
||
346 | |||
347 | break; |
||
348 | |||
349 | case "\n": |
||
350 | if ($j > 3) { |
||
351 | if (' ' == $code[$j] || "\x7F" == $code[$j]) { |
||
352 | --$j; |
||
353 | } |
||
354 | |||
355 | $code[++$j] = |
||
356 | false !== strpos('kend+-', $code[$j - 1]) |
||
357 | && preg_match( |
||
358 | "'(?:\+\+|--|(?<![\$.a-zA-Z0-9_])(break|continue|return|yield[ \x7F]?\*?))[ \x7F]?$'", |
||
359 | substr($code, $j - 8, 9) |
||
360 | ) |
||
361 | ? ';' : "\x7F"; |
||
362 | |||
363 | break; |
||
364 | } |
||
365 | |||
366 | case "\t": $f[$i] = ' '; |
||
367 | case ' ': |
||
368 | if (!$j || ' ' == $code[$j] || "\x7F" == $code[$j]) { |
||
369 | break; |
||
370 | } |
||
371 | |||
372 | default: |
||
373 | $code[++$j] = $f[$i]; |
||
374 | } |
||
375 | } |
||
376 | } |
||
377 | |||
378 | isset($s) && ($s = implode('', $s)) && $cc_on && $this->restoreCc($s); |
||
379 | unset($s); |
||
380 | |||
381 | $code = substr($code, 0, $j + 1); |
||
382 | $cc_on && $this->restoreCc($code, false); |
||
383 | |||
384 | // Protect wanted spaces and remove unwanted ones |
||
385 | $code = strtr($code, "\x7F", ' '); |
||
386 | $code = str_replace('- -', "-\x7F-", $code); |
||
387 | $code = str_replace('+ +', "+\x7F+", $code); |
||
388 | $code = preg_replace("'(\d)\s+\.\s*([a-zA-Z\$_[(])'", "$1\x7F.$2", $code); |
||
389 | $code = preg_replace("# ([-!%&;<=>~:.^+|,()*?[\]{}/']+)#", '$1', $code); |
||
390 | $code = preg_replace("#([-!%&;<=>~:.^+|,()*?[\]{}/]+) #", '$1', $code); |
||
391 | $cc_on && $code = preg_replace_callback("'//[^\'].*?@#3'", function ($m) { return strtr($m[0], ' ', "\x7F"); }, $code); |
||
392 | |||
393 | // Replace new Array/Object by []/{} |
||
394 | false !== strpos($code, 'new Array') && $code = preg_replace("'new Array(?:\(\)|([;\])},:]))'", '[]$1', $code); |
||
395 | false !== strpos($code, 'new Object') && $code = preg_replace("'new Object(?:\(\)|([;\])},:]))'", '{}$1', $code); |
||
396 | |||
397 | // Add missing semi-colons after curly braces |
||
398 | // This adds more semi-colons than strictly needed, |
||
399 | // but it seems that later gzipping is favorable to the repetition of "};" |
||
400 | $code = preg_replace("'\}(?![:,;.()\[\]}\|&]|(else|catch|finally|while)[^\$.a-zA-Z0-9_])'", '};', $code); |
||
401 | |||
402 | // Tag possible empty instruction for easy detection |
||
403 | $code = preg_replace("'(?<![\$.a-zA-Z0-9_])if\('", '1#(', $code); |
||
404 | $code = preg_replace("'(?<![\$.a-zA-Z0-9_])for\('", '2#(', $code); |
||
405 | $code = preg_replace("'(?<![\$.a-zA-Z0-9_])while\('", '3#(', $code); |
||
406 | |||
407 | $forPool = array(); |
||
408 | $instrPool = array(); |
||
409 | $s = 0; |
||
410 | |||
411 | $f = array(); |
||
412 | $j = -1; |
||
413 | |||
414 | // Remove as much semi-colon as possible |
||
415 | $len = strlen($code); |
||
416 | for ($i = 0; $i < $len; ++$i) { |
||
417 | switch ($code[$i]) { |
||
418 | case '(': |
||
419 | if ($j >= 0 && "\n" == $f[$j]) { |
||
420 | $f[$j] = ';'; |
||
421 | } |
||
422 | |||
423 | ++$s; |
||
424 | |||
425 | if ($i && '#' == $code[$i - 1]) { |
||
426 | $instrPool[$s - 1] = 1; |
||
427 | if ('2' == $code[$i - 2]) { |
||
428 | $forPool[$s] = 1; |
||
429 | } |
||
430 | } |
||
431 | |||
432 | $f[++$j] = '('; |
||
433 | break; |
||
434 | |||
435 | case ']': |
||
436 | case ')': |
||
437 | if ($i + 1 < $len && !isset($forPool[$s]) && !isset($instrPool[$s - 1]) && preg_match("'[a-zA-Z0-9_\$]'", $code[$i + 1])) { |
||
438 | $f[$j] .= $code[$i]; |
||
439 | $f[++$j] = "\n"; |
||
440 | } else { |
||
441 | $f[++$j] = $code[$i]; |
||
442 | } |
||
443 | |||
444 | if (')' == $code[$i]) { |
||
445 | unset($forPool[$s]); |
||
446 | --$s; |
||
447 | } |
||
448 | |||
449 | continue 2; |
||
450 | |||
451 | case '}': |
||
452 | if ("\n" == $f[$j]) { |
||
453 | $f[$j] = '}'; |
||
454 | } else { |
||
455 | $f[++$j] = '}'; |
||
456 | } |
||
457 | break; |
||
458 | |||
459 | case ';': |
||
460 | if (isset($forPool[$s]) || isset($instrPool[$s])) { |
||
461 | $f[++$j] = ';'; |
||
462 | } elseif ($j >= 0 && "\n" != $f[$j] && ';' != $f[$j]) { |
||
463 | $f[++$j] = "\n"; |
||
464 | } |
||
465 | |||
466 | break; |
||
467 | |||
468 | case '#': |
||
469 | switch ($f[$j]) { |
||
470 | case '1': $f[$j] = 'if'; break 2; |
||
471 | case '2': $f[$j] = 'for'; break 2; |
||
472 | case '3': $f[$j] = 'while'; break 2; |
||
473 | } |
||
474 | |||
475 | case '['; |
||
476 | if ($j >= 0 && "\n" == $f[$j]) { |
||
477 | $f[$j] = ';'; |
||
478 | } |
||
479 | |||
480 | default: $f[++$j] = $code[$i]; |
||
481 | } |
||
482 | |||
483 | unset($instrPool[$s]); |
||
484 | } |
||
485 | |||
486 | $f = implode('', $f); |
||
487 | $cc_on && $f = str_replace('@#3', "\r", $f); |
||
488 | |||
489 | // Fix "else ;" empty instructions |
||
490 | $f = preg_replace("'(?<![\$.a-zA-Z0-9_])else\n'", "\n", $f); |
||
491 | |||
492 | $r1 = array( // keywords with a direct object |
||
493 | 'case','delete','do','else','function','in','instanceof','of','break', |
||
494 | 'new','return','throw','typeof','var','void','yield','let','if', |
||
495 | 'const','get','set', |
||
496 | ); |
||
497 | |||
498 | $r2 = array( // keywords with a subject |
||
499 | 'in','instanceof','of', |
||
500 | ); |
||
501 | |||
502 | // Fix missing semi-colons |
||
503 | $f = preg_replace("'(?<!(?<![a-zA-Z0-9_\$])".implode(')(?<!(?<![a-zA-Z0-9_\$])', $r1).') (?!('.implode('|', $r2).")(?![a-zA-Z0-9_\$]))'", "\n", $f); |
||
504 | $f = preg_replace("'(?<!(?<![a-zA-Z0-9_\$])do)(?<!(?<![a-zA-Z0-9_\$])else) if\('", "\nif(", $f); |
||
505 | $f = preg_replace("'(?<=--|\+\+)(?<![a-zA-Z0-9_\$])(".implode('|', $r1).")(?![a-zA-Z0-9_\$])'", "\n$1", $f); |
||
506 | $f = preg_replace("'(?<![a-zA-Z0-9_\$])for\neach\('", 'for each(', $f); |
||
507 | $f = preg_replace("'(?<![a-zA-Z0-9_\$])\n(".implode('|', $r2).")(?![a-zA-Z0-9_\$])'", '$1', $f); |
||
508 | |||
509 | // Merge strings |
||
510 | if ($q["'"] > $q['"']) { |
||
511 | $q = array($q[1], $q[0]); |
||
512 | } |
||
513 | $f = preg_replace("#//''\"\"[0-9]+'#", $q[0].'$0'.$q[0], $f); |
||
514 | strpos($f, $q[0].'+'.$q[0]) && $f = str_replace($q[0].'+'.$q[0], '', $f); |
||
515 | $len = count($strings); |
||
516 | foreach ($strings as $r1 => &$r2) { |
||
517 | $r2 = "/'" == substr($r1, -2) |
||
518 | ? str_replace(array("\\'", '\\"'), array("'", '"'), $r2) |
||
519 | : str_replace('\\'.$q[1], $q[1], $r2); |
||
520 | } |
||
521 | |||
522 | // Restore wanted spaces |
||
523 | $f = strtr($f, "\x7F", ' '); |
||
524 | |||
525 | return array($f, $strings); |
||
526 | } |
||
527 | |||
528 | protected function extractClosures($code) |
||
529 | { |
||
530 | $code = ';'.$code; |
||
531 | |||
532 | $this->argFreq[-1] += substr_count($code, '}catch('); |
||
533 | |||
534 | if ($this->argFreq[-1]) { |
||
535 | // Special catch scope handling |
||
536 | |||
537 | // FIXME: this implementation doesn't work with nested catch scopes who need |
||
538 | // access to their parent's caught variable (but who needs that?). |
||
539 | |||
540 | $f = preg_split("@}catch\(({$this->varRx})@", $code, -1, PREG_SPLIT_DELIM_CAPTURE); |
||
541 | |||
542 | $code = 'catch$scope$var'.mt_rand(); |
||
543 | $this->specialVarRx = $this->specialVarRx ? '(?:'.$this->specialVarRx.'|'.preg_quote($code).')' : preg_quote($code); |
||
544 | $i = count($f) - 1; |
||
545 | |||
546 | while ($i) { |
||
547 | $c = 1; |
||
548 | $j = 0; |
||
549 | $l = strlen($f[$i]); |
||
550 | |||
551 | while ($c && $j < $l) { |
||
552 | $s = $f[$i][$j++]; |
||
553 | $c += '(' == $s ? 1 : (')' == $s ? -1 : 0); |
||
554 | } |
||
555 | |||
556 | if (!$c) { |
||
557 | do { |
||
558 | $s = $f[$i][$j++]; |
||
559 | $c += '{' == $s ? 1 : ('}' == $s ? -1 : 0); |
||
560 | } while ($c && $j < $l); |
||
561 | } |
||
562 | |||
563 | $c = preg_quote($f[$i - 1], '#'); |
||
564 | $f[$i - 2] .= '}catch('.preg_replace("#([.,{]?)(?<![a-zA-Z0-9_\$@]){$c}\\b#", '$1'.$code, $f[$i - 1].substr($f[$i], 0, $j)).substr($f[$i], $j); |
||
565 | |||
566 | unset($f[$i--], $f[$i--]); |
||
567 | } |
||
568 | |||
569 | $code = $f[0]; |
||
570 | } |
||
571 | |||
572 | $f = preg_split("'(?<![a-zA-Z0-9_\$])((?:function[ (]|get |set ).*?\{)'", $code, -1, PREG_SPLIT_DELIM_CAPTURE); |
||
573 | $i = count($f) - 1; |
||
574 | $closures = array(); |
||
575 | |||
576 | while ($i) { |
||
577 | $c = 1; |
||
578 | $j = 0; |
||
579 | $l = strlen($f[$i]); |
||
580 | |||
581 | while ($c && $j < $l) { |
||
582 | $s = $f[$i][$j++]; |
||
583 | $c += '{' == $s ? 1 : ('}' == $s ? -1 : 0); |
||
584 | } |
||
585 | |||
586 | switch (substr($f[$i - 2], -1)) { |
||
587 | default: |
||
588 | if (false !== $c = strpos($f[$i - 1], ' ', 8)) { |
||
589 | break; |
||
590 | } |
||
591 | case false: case "\n": case ';': case '{': case '}': case ')': case ']': |
||
592 | $c = strpos($f[$i - 1], '(', 4); |
||
593 | } |
||
594 | |||
595 | $l = "//''\"\"#$i'"; |
||
596 | $code = substr($f[$i - 1], $c); |
||
597 | $closures[$l] = $code.substr($f[$i], 0, $j); |
||
598 | $f[$i - 2] .= substr($f[$i - 1], 0, $c).$l.substr($f[$i], $j); |
||
599 | |||
600 | if ('(){' !== $code) { |
||
601 | $j = substr_count($code, ','); |
||
602 | do { |
||
603 | isset($this->argFreq[$j]) ? ++$this->argFreq[$j] : $this->argFreq[$j] = 1; |
||
604 | } while ($j--); |
||
605 | } |
||
606 | |||
607 | $i -= 2; |
||
608 | } |
||
609 | |||
610 | return array($f[0], $closures); |
||
611 | } |
||
612 | |||
613 | protected function makeVars($closure, &$tree, $key) |
||
614 | { |
||
615 | $tree['code'] = &$closure; |
||
616 | $tree['nfe'] = false; |
||
617 | $tree['used'] = array(); |
||
618 | $tree['local'] = array(); |
||
619 | |||
620 | // Replace multiple "var" declarations by a single one |
||
621 | $closure = preg_replace_callback("'(?<=[\n\{\}])var [^\n\{\};]+(?:\nvar [^\n\{\};]+)+'", array($this, 'mergeVarDeclarations'), $closure); |
||
622 | |||
623 | // Get all local vars (functions, arguments and "var" prefixed) |
||
624 | |||
625 | $vars = &$tree['local']; |
||
626 | |||
627 | if (preg_match("'^( [^(]*)?\((.*?)\)\{'", $closure, $v)) { |
||
628 | if ($v[1]) { |
||
629 | $vars[$tree['nfe'] = substr($v[1], 1)] = -1; |
||
630 | $tree['parent']['local'][';'.$key] = &$vars[$tree['nfe']]; |
||
631 | } |
||
632 | |||
633 | if ($v[2]) { |
||
634 | $i = 0; |
||
635 | $v = explode(',', $v[2]); |
||
636 | foreach ($v as $w) { |
||
637 | $vars[$w] = $this->argFreq[$i++] - 1; // Give a bonus to argument variables |
||
638 | } |
||
639 | } |
||
640 | } |
||
641 | |||
642 | $v = preg_split("'(?<![\$.a-zA-Z0-9_])var '", $closure); |
||
643 | if ($i = count($v) - 1) { |
||
644 | $w = array(); |
||
645 | |||
646 | while ($i) { |
||
647 | $j = $c = 0; |
||
648 | $l = strlen($v[$i]); |
||
649 | |||
650 | while ($j < $l) { |
||
651 | switch ($v[$i][$j]) { |
||
652 | case '(': case '[': case '{': |
||
653 | ++$c; |
||
654 | break; |
||
655 | |||
656 | case ')': case ']': case '}': |
||
657 | if ($c-- <= 0) { |
||
658 | break 2; |
||
659 | } |
||
660 | break; |
||
661 | |||
662 | case ';': case "\n": |
||
663 | if (!$c) { |
||
664 | break 2; |
||
665 | } |
||
666 | |||
667 | default: |
||
668 | $c || $w[] = $v[$i][$j]; |
||
669 | } |
||
670 | |||
671 | ++$j; |
||
672 | } |
||
673 | |||
674 | $w[] = ','; |
||
675 | --$i; |
||
676 | } |
||
677 | |||
678 | $v = explode(',', implode('', $w)); |
||
679 | foreach ($v as $w) { |
||
680 | if (preg_match("'^{$this->varRx}'", $w, $v)) { |
||
681 | isset($vars[$v[0]]) || $vars[$v[0]] = 0; |
||
682 | } |
||
683 | } |
||
684 | } |
||
685 | |||
686 | if (preg_match_all("@function ({$this->varRx})//''\"\"#@", $closure, $v)) { |
||
687 | foreach ($v[1] as $w) { |
||
688 | isset($vars[$w]) || $vars[$w] = 0; |
||
689 | } |
||
690 | } |
||
691 | |||
692 | if ($this->argFreq[-1] && preg_match_all("@}catch\(({$this->varRx})@", $closure, $v)) { |
||
693 | $v[0] = array(); |
||
694 | foreach ($v[1] as $w) { |
||
695 | isset($v[0][$w]) ? ++$v[0][$w] : $v[0][$w] = 1; |
||
696 | } |
||
697 | foreach ($v[0] as $w => $v) { |
||
698 | $vars[$w] = $this->argFreq[-1] - $v; |
||
699 | } |
||
700 | } |
||
701 | |||
702 | // Get all used vars, local and non-local |
||
703 | |||
704 | $vars = &$tree['used']; |
||
705 | |||
706 | if (preg_match_all("#([.,{]?(?:[gs]et )?)(?<![a-zA-Z0-9_\$])({$this->varRx})(:?)#", $closure, $w, PREG_SET_ORDER)) { |
||
707 | foreach ($w as $k) { |
||
708 | if (isset($k[1][0]) && (',' === $k[1][0] || '{' === $k[1][0])) { |
||
709 | if (':' === $k[3]) { |
||
710 | $k = '.'.$k[2]; |
||
711 | } elseif ('get ' === substr($k[1], 1, 4) || 'set ' === substr($k[1], 1, 4)) { |
||
712 | ++$this->charFreq[ord($k[1][1])]; // "g" or "s" |
||
713 | ++$this->charFreq[101]; // "e" |
||
714 | ++$this->charFreq[116]; // "t" |
||
715 | $k = '.'.$k[2]; |
||
716 | } else { |
||
717 | $k = $k[2]; |
||
718 | } |
||
719 | } else { |
||
720 | $k = $k[1].$k[2]; |
||
721 | } |
||
722 | |||
723 | isset($vars[$k]) ? ++$vars[$k] : $vars[$k] = 1; |
||
724 | } |
||
725 | } |
||
726 | |||
727 | if (preg_match_all("#//''\"\"[0-9]+(?:['!]|/')#", $closure, $w)) { |
||
728 | foreach ($w[0] as $a) { |
||
729 | $v = "'" === substr($a, -1) && "/'" !== substr($a, -2) && $this->specialVarRx |
||
730 | ? preg_split("#([.,{]?(?:[gs]et )?(?<![a-zA-Z0-9_\$@]){$this->specialVarRx}:?)#", $this->strings[$a], -1, PREG_SPLIT_DELIM_CAPTURE) |
||
731 | : array($this->strings[$a]); |
||
732 | $a = count($v); |
||
733 | |||
734 | for ($i = 0; $i < $a; ++$i) { |
||
735 | $k = $v[$i]; |
||
736 | |||
737 | if (1 === $i % 2) { |
||
738 | if (',' === $k[0] || '{' === $k[0]) { |
||
739 | if (':' === substr($k, -1)) { |
||
740 | $k = '.'.substr($k, 1, -1); |
||
741 | } elseif ('get ' === substr($k, 1, 4) || 'set ' === substr($k, 1, 4)) { |
||
742 | ++$this->charFreq[ord($k[1])]; // "g" or "s" |
||
743 | ++$this->charFreq[101]; // "e" |
||
744 | ++$this->charFreq[116]; // "t" |
||
745 | $k = '.'.substr($k, 5); |
||
746 | } else { |
||
747 | $k = substr($k, 1); |
||
748 | } |
||
749 | } elseif (':' === substr($k, -1)) { |
||
750 | $k = substr($k, 0, -1); |
||
751 | } |
||
752 | |||
753 | $w = &$tree; |
||
754 | |||
755 | while (isset($w['parent']) && !(isset($w['used'][$k]) || isset($w['local'][$k]))) { |
||
756 | $w = &$w['parent']; |
||
757 | } |
||
758 | |||
759 | (isset($w['used'][$k]) || isset($w['local'][$k])) && (isset($vars[$k]) ? ++$vars[$k] : $vars[$k] = 1); |
||
760 | |||
761 | unset($w); |
||
762 | } |
||
763 | |||
764 | if (0 === $i % 2 || !isset($vars[$k])) { |
||
765 | foreach (count_chars($v[$i], 1) as $k => $w) { |
||
766 | $this->charFreq[$k] += $w; |
||
767 | } |
||
768 | } |
||
769 | } |
||
770 | } |
||
771 | } |
||
772 | |||
773 | // Propagate the usage number to parents |
||
774 | |||
775 | foreach ($vars as $w => $a) { |
||
776 | $k = &$tree; |
||
777 | $chain = array(); |
||
778 | do { |
||
779 | $vars = &$k['local']; |
||
780 | $chain[] = &$k; |
||
781 | if (isset($vars[$w])) { |
||
782 | unset($k['used'][$w]); |
||
783 | if (isset($vars[$w])) { |
||
784 | $vars[$w] += $a; |
||
785 | } else { |
||
786 | $vars[$w] = $a; |
||
787 | } |
||
788 | $a = false; |
||
789 | break; |
||
790 | } |
||
791 | } while ($k['parent'] && $k = &$k['parent']); |
||
792 | |||
793 | if ($a && !$k['parent']) { |
||
794 | if (isset($vars[$w])) { |
||
795 | $vars[$w] += $a; |
||
796 | } else { |
||
797 | $vars[$w] = $a; |
||
798 | } |
||
799 | } |
||
800 | |||
801 | if (isset($tree['used'][$w]) && isset($vars[$w])) { |
||
802 | foreach ($chain as &$b) { |
||
803 | isset($b['local'][$w]) || $b['used'][$w] = &$vars[$w]; |
||
804 | } |
||
805 | } |
||
806 | } |
||
807 | |||
808 | // Analyse childs |
||
809 | |||
810 | $tree['childs'] = array(); |
||
811 | $vars = &$tree['childs']; |
||
812 | |||
813 | if (preg_match_all("@//''\"\"#[0-9]+'@", $closure, $w)) { |
||
814 | foreach ($w[0] as $a) { |
||
815 | $vars[$a] = array('parent' => &$tree); |
||
816 | $this->makeVars($this->closures[$a], $vars[$a], $a); |
||
817 | } |
||
818 | } |
||
819 | } |
||
820 | |||
821 | protected function mergeVarDeclarations($m) |
||
824 | } |
||
825 | |||
826 | protected function renameVars(&$tree, $root) |
||
827 | { |
||
828 | if ($root) { |
||
829 | $tree['local'] += $tree['used']; |
||
830 | $tree['used'] = array(); |
||
831 | |||
832 | foreach ($tree['local'] as $k => $v) { |
||
833 | if ('.' == $k[0]) { |
||
834 | $k = substr($k, 1); |
||
835 | } |
||
836 | |||
837 | if ('true' === $k) { |
||
838 | $this->charFreq[48] += $v; |
||
839 | } elseif ('false' === $k) { |
||
840 | $this->charFreq[49] += $v; |
||
841 | } elseif (!$this->specialVarRx || !preg_match("#^{$this->specialVarRx}$#", $k)) { |
||
842 | foreach (count_chars($k, 1) as $k => $w) { |
||
843 | $this->charFreq[$k] += $w * $v; |
||
844 | } |
||
845 | } elseif (2 == strlen($k)) { |
||
846 | $tree['used'][] = $k[1]; |
||
847 | } |
||
848 | } |
||
849 | |||
850 | $this->charFreq = $this->rsort($this->charFreq); |
||
851 | |||
852 | $this->str0 = ''; |
||
853 | $this->str1 = ''; |
||
854 | |||
855 | foreach ($this->charFreq as $k => $v) { |
||
856 | if (!$v) { |
||
857 | break; |
||
858 | } |
||
859 | |||
860 | $v = chr($k); |
||
861 | |||
862 | if ((64 < $k && $k < 91) || (96 < $k && $k < 123)) { // A-Z a-z |
||
863 | $this->str0 .= $v; |
||
864 | $this->str1 .= $v; |
||
865 | } elseif (47 < $k && $k < 58) { // 0-9 |
||
866 | $this->str1 .= $v; |
||
867 | } |
||
868 | } |
||
869 | |||
870 | if ('' === $this->str0) { |
||
871 | $this->str0 = 'claspemitdbfrugnjvhowkxqyzCLASPEMITDBFRUGNJVHOWKXQYZ'; |
||
872 | $this->str1 = $this->str0.'0123456789'; |
||
873 | } |
||
874 | |||
875 | foreach ($tree['local'] as $var => $root) { |
||
876 | if ('.' != substr($var, 0, 1) && isset($tree['local'][".{$var}"])) { |
||
877 | $tree['local'][$var] += $tree['local'][".{$var}"]; |
||
878 | } |
||
879 | } |
||
880 | |||
881 | foreach ($tree['local'] as $var => $root) { |
||
882 | if ('.' == substr($var, 0, 1) && isset($tree['local'][substr($var, 1)])) { |
||
883 | $tree['local'][$var] = $tree['local'][substr($var, 1)]; |
||
884 | } |
||
885 | } |
||
886 | |||
887 | $tree['local'] = $this->rsort($tree['local']); |
||
888 | |||
889 | foreach ($tree['local'] as $var => $root) { |
||
890 | switch (substr($var, 0, 1)) { |
||
891 | case '.': |
||
892 | if (!isset($tree['local'][substr($var, 1)])) { |
||
893 | $tree['local'][$var] = '#'.($this->specialVarRx && 3 < strlen($var) && preg_match("'^\.{$this->specialVarRx}$'", $var) ? $this->getNextName($tree).'$' : substr($var, 1)); |
||
894 | } |
||
895 | break; |
||
896 | |||
897 | case ';': $tree['local'][$var] = 0 === $root ? '' : $this->getNextName($tree); |
||
898 | case '#': break; |
||
899 | |||
900 | default: |
||
901 | $root = $this->specialVarRx && 2 < strlen($var) && preg_match("'^{$this->specialVarRx}$'", $var) ? $this->getNextName($tree).'$' : $var; |
||
902 | $tree['local'][$var] = $root; |
||
903 | if (isset($tree['local'][".{$var}"])) { |
||
904 | $tree['local'][".{$var}"] = '#'.$root; |
||
905 | } |
||
906 | } |
||
907 | } |
||
908 | |||
909 | foreach ($tree['local'] as $var => $root) { |
||
910 | $tree['local'][$var] = preg_replace("'^#'", '.', $tree['local'][$var]); |
||
911 | } |
||
912 | } else { |
||
913 | $tree['local'] = $this->rsort($tree['local']); |
||
914 | if (false !== $tree['nfe']) { |
||
915 | $tree['used'][] = $tree['local'][$tree['nfe']]; |
||
916 | } |
||
917 | |||
918 | foreach ($tree['local'] as $var => $root) { |
||
919 | if ($tree['nfe'] !== $var) { |
||
920 | $tree['local'][$var] = 0 === $root ? '' : $this->getNextName($tree); |
||
921 | } |
||
922 | } |
||
923 | } |
||
924 | |||
925 | $this->local_tree = &$tree['local']; |
||
926 | $this->used_tree = &$tree['used']; |
||
927 | |||
928 | $tree['code'] = preg_replace_callback("#[.,{ ]?(?:[gs]et )?(?<![a-zA-Z0-9_\$@]){$this->varRx}:?#", array($this, 'getNewName'), $tree['code']); |
||
929 | |||
930 | if ($this->specialVarRx && preg_match_all("#//''\"\"[0-9]+'#", $tree['code'], $b)) { |
||
931 | foreach ($b[0] as $a) { |
||
932 | $this->strings[$a] = preg_replace_callback( |
||
933 | "#[.,{]?(?:[gs]et )?(?<![a-zA-Z0-9_\$@]){$this->specialVarRx}:?#", |
||
934 | array($this, 'getNewName'), |
||
935 | $this->strings[$a] |
||
936 | ); |
||
937 | } |
||
938 | } |
||
939 | |||
940 | foreach ($tree['childs'] as $a => &$b) { |
||
941 | $this->renameVars($b, false); |
||
942 | $tree['code'] = str_replace($a, $b['code'], $tree['code']); |
||
943 | unset($tree['childs'][$a]); |
||
944 | } |
||
945 | } |
||
946 | |||
947 | protected function getNewName($m) |
||
948 | { |
||
949 | $m = $m[0]; |
||
950 | |||
951 | $pre = '.' === $m[0] ? '.' : ''; |
||
952 | $post = ''; |
||
953 | |||
954 | if (',' === $m[0] || '{' === $m[0] || ' ' === $m[0]) { |
||
955 | $pre = $m[0]; |
||
956 | |||
957 | if (':' === substr($m, -1)) { |
||
958 | $post = ':'; |
||
959 | $m = (' ' !== $m[0] ? '.' : '').substr($m, 1, -1); |
||
960 | } elseif ('get ' === substr($m, 1, 4) || 'set ' === substr($m, 1, 4)) { |
||
961 | $pre .= substr($m, 1, 4); |
||
962 | $m = '.'.substr($m, 5); |
||
963 | } else { |
||
964 | $m = substr($m, 1); |
||
965 | } |
||
966 | } elseif (':' === substr($m, -1)) { |
||
967 | $post = ':'; |
||
968 | $m = substr($m, 0, -1); |
||
969 | } |
||
970 | |||
971 | $post = (isset($this->reserved[$m]) |
||
972 | ? ('true' === $m ? '!0' : ('false' === $m ? '!1' : $m)) |
||
973 | : ( |
||
974 | isset($this->local_tree[$m]) |
||
975 | ? $this->local_tree[$m] |
||
976 | : ( |
||
977 | isset($this->used_tree[$m]) |
||
978 | ? $this->used_tree[$m] |
||
979 | : $m |
||
980 | ) |
||
981 | ) |
||
982 | ).$post; |
||
983 | |||
984 | return '' === $post ? '' : ($pre.('.' === $post[0] ? substr($post, 1) : $post)); |
||
985 | } |
||
986 | |||
987 | protected function getNextName(&$tree = array(), &$counter = false) |
||
988 | { |
||
989 | if (false === $counter) { |
||
990 | $counter = &$tree['counter']; |
||
991 | isset($counter) || $counter = -1; |
||
992 | $exclude = array_flip($tree['used']); |
||
993 | } else { |
||
994 | $exclude = $tree; |
||
995 | } |
||
996 | |||
997 | ++$counter; |
||
998 | |||
999 | $len0 = strlen($this->str0); |
||
1000 | $len1 = strlen($this->str0); |
||
1001 | |||
1002 | $name = $this->str0[$counter % $len0]; |
||
1003 | |||
1004 | $i = intval($counter / $len0) - 1; |
||
1005 | while ($i >= 0) { |
||
1006 | $name .= $this->str1[ $i % $len1 ]; |
||
1007 | $i = intval($i / $len1) - 1; |
||
1008 | } |
||
1009 | |||
1010 | return !(isset($this->reserved[$name]) || isset($exclude[$name])) ? $name : $this->getNextName($exclude, $counter); |
||
1011 | } |
||
1012 | |||
1013 | protected function restoreCc(&$s, $lf = true) |
||
1014 | { |
||
1015 | $lf && $s = str_replace('@#3', '', $s); |
||
1016 | |||
1017 | $s = str_replace('@#1', '@*/', $s); |
||
1018 | $s = str_replace('2#@', '//@', $s); |
||
1019 | $s = str_replace('1#@', '/*@', $s); |
||
1020 | $s = str_replace('##', '#', $s); |
||
1021 | } |
||
1022 | |||
1023 | private function rsort($array) |
||
1059 | } |
||
1060 | } |
||
1061 | |||
1062 | ?> |
||
1063 |