| 5 |
aurelien |
1 |
<?php
|
|
|
2 |
/**
|
|
|
3 |
* Tokenizes PHP code.
|
|
|
4 |
*
|
|
|
5 |
* PHP version 5
|
|
|
6 |
*
|
|
|
7 |
* @category PHP
|
|
|
8 |
* @package PHP_CodeSniffer
|
|
|
9 |
* @author Greg Sherwood <gsherwood@squiz.net>
|
|
|
10 |
* @copyright 2006 Squiz Pty Ltd (ABN 77 084 670 600)
|
|
|
11 |
* @license http://matrix.squiz.net/developer/tools/php_cs/licence BSD Licence
|
| 34 |
aurelien |
12 |
* @version CVS: $Id: PHP.php 34 2009-04-09 07:34:39Z aurelien $
|
| 5 |
aurelien |
13 |
* @link http://pear.php.net/package/PHP_CodeSniffer
|
|
|
14 |
*/
|
|
|
15 |
|
|
|
16 |
/**
|
|
|
17 |
* Tokenizes PHP code.
|
|
|
18 |
*
|
|
|
19 |
* @category PHP
|
|
|
20 |
* @package PHP_CodeSniffer
|
|
|
21 |
* @author Greg Sherwood <gsherwood@squiz.net>
|
|
|
22 |
* @copyright 2006 Squiz Pty Ltd (ABN 77 084 670 600)
|
|
|
23 |
* @license http://matrix.squiz.net/developer/tools/php_cs/licence BSD Licence
|
|
|
24 |
* @version Release: 1.2.0RC1
|
|
|
25 |
* @link http://pear.php.net/package/PHP_CodeSniffer
|
|
|
26 |
*/
|
|
|
27 |
class PHP_CodeSniffer_Tokenizers_PHP
|
|
|
28 |
{
|
|
|
29 |
|
|
|
30 |
/**
|
|
|
31 |
* A list of tokens that are allowed to open a scope.
|
|
|
32 |
*
|
|
|
33 |
* This array also contains information about what kind of token the scope
|
|
|
34 |
* opener uses to open and close the scope, if the token strictly requires
|
|
|
35 |
* an opener, if the token can share a scope closer, and who it can be shared
|
|
|
36 |
* with. An example of a token that shares a scope closer is a CASE scope.
|
|
|
37 |
*
|
|
|
38 |
* @var array
|
|
|
39 |
*/
|
|
|
40 |
public $scopeOpeners = array(
|
|
|
41 |
T_IF => array(
|
|
|
42 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
43 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
44 |
'strict' => false,
|
|
|
45 |
'shared' => false,
|
|
|
46 |
'with' => array(),
|
|
|
47 |
),
|
|
|
48 |
T_TRY => array(
|
|
|
49 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
50 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
51 |
'strict' => true,
|
|
|
52 |
'shared' => false,
|
|
|
53 |
'with' => array(),
|
|
|
54 |
),
|
|
|
55 |
T_CATCH => array(
|
|
|
56 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
57 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
58 |
'strict' => true,
|
|
|
59 |
'shared' => false,
|
|
|
60 |
'with' => array(),
|
|
|
61 |
),
|
|
|
62 |
T_ELSE => array(
|
|
|
63 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
64 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
65 |
'strict' => false,
|
|
|
66 |
'shared' => false,
|
|
|
67 |
'with' => array(),
|
|
|
68 |
),
|
|
|
69 |
T_ELSEIF => array(
|
|
|
70 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
71 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
72 |
'strict' => false,
|
|
|
73 |
'shared' => false,
|
|
|
74 |
'with' => array(),
|
|
|
75 |
),
|
|
|
76 |
T_FOR => array(
|
|
|
77 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
78 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
79 |
'strict' => false,
|
|
|
80 |
'shared' => false,
|
|
|
81 |
'with' => array(),
|
|
|
82 |
),
|
|
|
83 |
T_FOREACH => array(
|
|
|
84 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
85 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
86 |
'strict' => false,
|
|
|
87 |
'shared' => false,
|
|
|
88 |
'with' => array(),
|
|
|
89 |
),
|
|
|
90 |
T_INTERFACE => array(
|
|
|
91 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
92 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
93 |
'strict' => true,
|
|
|
94 |
'shared' => false,
|
|
|
95 |
'with' => array(),
|
|
|
96 |
),
|
|
|
97 |
T_FUNCTION => array(
|
|
|
98 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
99 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
100 |
'strict' => false,
|
|
|
101 |
'shared' => false,
|
|
|
102 |
'with' => array(),
|
|
|
103 |
),
|
|
|
104 |
T_CLASS => array(
|
|
|
105 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
106 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
107 |
'strict' => true,
|
|
|
108 |
'shared' => false,
|
|
|
109 |
'with' => array(),
|
|
|
110 |
),
|
|
|
111 |
T_WHILE => array(
|
|
|
112 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
113 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
114 |
'strict' => false,
|
|
|
115 |
'shared' => false,
|
|
|
116 |
'with' => array(),
|
|
|
117 |
),
|
|
|
118 |
T_DO => array(
|
|
|
119 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
120 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
121 |
'strict' => true,
|
|
|
122 |
'shared' => false,
|
|
|
123 |
'with' => array(),
|
|
|
124 |
),
|
|
|
125 |
T_SWITCH => array(
|
|
|
126 |
'start' => T_OPEN_CURLY_BRACKET,
|
|
|
127 |
'end' => T_CLOSE_CURLY_BRACKET,
|
|
|
128 |
'strict' => true,
|
|
|
129 |
'shared' => false,
|
|
|
130 |
'with' => array(),
|
|
|
131 |
),
|
|
|
132 |
T_CASE => array(
|
|
|
133 |
'start' => T_COLON,
|
|
|
134 |
'end' => T_BREAK,
|
|
|
135 |
'strict' => true,
|
|
|
136 |
'shared' => true,
|
|
|
137 |
'with' => array(
|
|
|
138 |
T_DEFAULT,
|
|
|
139 |
T_CASE,
|
|
|
140 |
T_SWITCH,
|
|
|
141 |
),
|
|
|
142 |
),
|
|
|
143 |
T_DEFAULT => array(
|
|
|
144 |
'start' => T_COLON,
|
|
|
145 |
'end' => T_BREAK,
|
|
|
146 |
'strict' => true,
|
|
|
147 |
'shared' => true,
|
|
|
148 |
'with' => array(
|
|
|
149 |
T_CASE,
|
|
|
150 |
T_SWITCH,
|
|
|
151 |
),
|
|
|
152 |
),
|
|
|
153 |
T_START_HEREDOC => array(
|
|
|
154 |
'start' => T_START_HEREDOC,
|
|
|
155 |
'end' => T_END_HEREDOC,
|
|
|
156 |
'strict' => true,
|
|
|
157 |
'shared' => false,
|
|
|
158 |
'with' => array(),
|
|
|
159 |
),
|
|
|
160 |
);
|
|
|
161 |
|
|
|
162 |
/**
|
|
|
163 |
* A list of tokens that end the scope.
|
|
|
164 |
*
|
|
|
165 |
* This array is just a unique collection of the end tokens
|
|
|
166 |
* from the _scopeOpeners array. The data is duplicated here to
|
|
|
167 |
* save time during parsing of the file.
|
|
|
168 |
*
|
|
|
169 |
* @var array
|
|
|
170 |
*/
|
|
|
171 |
public $endScopeTokens = array(
|
|
|
172 |
T_CLOSE_CURLY_BRACKET,
|
|
|
173 |
T_BREAK,
|
|
|
174 |
T_END_HEREDOC,
|
|
|
175 |
);
|
|
|
176 |
|
|
|
177 |
|
|
|
178 |
/**
|
|
|
179 |
* Creates an array of tokens when given some PHP code.
|
|
|
180 |
*
|
|
|
181 |
* Starts by using token_get_all() but does a lot of extra processing
|
|
|
182 |
* to insert information about the context of the token.
|
|
|
183 |
*
|
|
|
184 |
* @param string $string The string to tokenize.
|
|
|
185 |
* @param string $eolChar The EOL character to use for splitting strings.
|
|
|
186 |
*
|
|
|
187 |
* @return array
|
|
|
188 |
*/
|
|
|
189 |
public function tokenizeString($string, $eolChar='\n')
|
|
|
190 |
{
|
|
|
191 |
$tokens = @token_get_all($string);
|
|
|
192 |
$finalTokens = array();
|
|
|
193 |
|
|
|
194 |
$newStackPtr = 0;
|
|
|
195 |
$numTokens = count($tokens);
|
|
|
196 |
for ($stackPtr = 0; $stackPtr < $numTokens; $stackPtr++) {
|
|
|
197 |
$token = $tokens[$stackPtr];
|
|
|
198 |
$tokenIsArray = is_array($token);
|
|
|
199 |
|
|
|
200 |
/*
|
|
|
201 |
If we are using \r\n newline characters, the \r and \n are sometimes
|
|
|
202 |
split over two tokens. This normally occurs after comments. We need
|
|
|
203 |
to merge these two characters together so that our line endings are
|
|
|
204 |
consistent for all lines.
|
|
|
205 |
*/
|
|
|
206 |
|
|
|
207 |
if ($tokenIsArray === true && substr($token[1], -1) === "\r") {
|
|
|
208 |
if (isset($tokens[($stackPtr + 1)]) === true
|
|
|
209 |
&& is_array($tokens[($stackPtr + 1)]) === true
|
|
|
210 |
&& $tokens[($stackPtr + 1)][1][0] === "\n"
|
|
|
211 |
) {
|
|
|
212 |
$token[1] .= "\n";
|
|
|
213 |
|
|
|
214 |
if ($tokens[($stackPtr + 1)][1] === "\n") {
|
|
|
215 |
// The next token's content has been merged into this token,
|
|
|
216 |
// so we can skip it.
|
|
|
217 |
$stackPtr++;
|
|
|
218 |
} else {
|
|
|
219 |
$tokens[($stackPtr + 1)][1]
|
|
|
220 |
= substr($tokens[($stackPtr + 1)][1], 1);
|
|
|
221 |
}
|
|
|
222 |
}
|
|
|
223 |
}//end if
|
|
|
224 |
|
|
|
225 |
/*
|
|
|
226 |
If this is a double quoted string, PHP will tokenise the whole
|
|
|
227 |
thing which causes problems with the scope map when braces are
|
|
|
228 |
within the string. So we need to merge the tokens together to
|
|
|
229 |
provide a single string.
|
|
|
230 |
*/
|
|
|
231 |
|
|
|
232 |
if ($tokenIsArray === false && $token === '"') {
|
|
|
233 |
$tokenContent = '"';
|
|
|
234 |
$nestedVars = array();
|
|
|
235 |
for ($i = ($stackPtr + 1); $i < $numTokens; $i++) {
|
|
|
236 |
$subTokenIsArray = is_array($tokens[$i]);
|
|
|
237 |
|
|
|
238 |
if ($subTokenIsArray === true) {
|
|
|
239 |
$tokenContent .= $tokens[$i][1];
|
|
|
240 |
if ($tokens[$i][1] === '{'
|
|
|
241 |
&& $tokens[$i][0] !== T_ENCAPSED_AND_WHITESPACE
|
|
|
242 |
) {
|
|
|
243 |
$nestedVars[] = $i;
|
|
|
244 |
}
|
|
|
245 |
} else {
|
|
|
246 |
$tokenContent .= $tokens[$i];
|
|
|
247 |
if ($tokens[$i] === '}') {
|
|
|
248 |
array_pop($nestedVars);
|
|
|
249 |
}
|
|
|
250 |
}
|
|
|
251 |
|
|
|
252 |
if ($subTokenIsArray === false
|
|
|
253 |
&& $tokens[$i] === '"'
|
|
|
254 |
&& empty($nestedVars) === true
|
|
|
255 |
) {
|
|
|
256 |
// We found the other end of the double quoted string.
|
|
|
257 |
break;
|
|
|
258 |
}
|
|
|
259 |
}
|
|
|
260 |
|
|
|
261 |
$stackPtr = $i;
|
|
|
262 |
|
|
|
263 |
// Convert each line within the double quoted string to a
|
|
|
264 |
// new token, so it conforms with other multiple line tokens.
|
|
|
265 |
$tokenLines = explode($eolChar, $tokenContent);
|
|
|
266 |
$numLines = count($tokenLines);
|
|
|
267 |
$newToken = array();
|
|
|
268 |
|
|
|
269 |
for ($j = 0; $j < $numLines; $j++) {
|
|
|
270 |
$newToken['content'] = $tokenLines[$j];
|
|
|
271 |
if ($j === ($numLines - 1)) {
|
|
|
272 |
if ($tokenLines[$j] === '') {
|
|
|
273 |
break;
|
|
|
274 |
}
|
|
|
275 |
} else {
|
|
|
276 |
$newToken['content'] .= $eolChar;
|
|
|
277 |
}
|
|
|
278 |
|
|
|
279 |
$newToken['code'] = T_DOUBLE_QUOTED_STRING;
|
|
|
280 |
$newToken['type'] = 'T_DOUBLE_QUOTED_STRING';
|
|
|
281 |
$finalTokens[$newStackPtr] = $newToken;
|
|
|
282 |
$newStackPtr++;
|
|
|
283 |
}
|
|
|
284 |
|
|
|
285 |
// Continue, as we're done with this token.
|
|
|
286 |
continue;
|
|
|
287 |
}//end if
|
|
|
288 |
|
|
|
289 |
/*
|
|
|
290 |
If this is a heredoc, PHP will tokenise the whole
|
|
|
291 |
thing which causes problems when heredocs don't
|
|
|
292 |
contain real PHP code, which is almost never.
|
|
|
293 |
We want to leave the start and end heredoc tokens
|
|
|
294 |
alone though.
|
|
|
295 |
*/
|
|
|
296 |
|
|
|
297 |
if ($tokenIsArray === true && $token[0] === T_START_HEREDOC) {
|
|
|
298 |
// Add the start heredoc token to the final array.
|
|
|
299 |
$finalTokens[$newStackPtr]
|
|
|
300 |
= PHP_CodeSniffer::standardiseToken($token);
|
|
|
301 |
$newStackPtr++;
|
|
|
302 |
|
|
|
303 |
$tokenContent = '';
|
|
|
304 |
for ($i = ($stackPtr + 1); $i < $numTokens; $i++) {
|
|
|
305 |
$subTokenIsArray = is_array($tokens[$i]);
|
|
|
306 |
if ($subTokenIsArray === true
|
|
|
307 |
&& $tokens[$i][0] === T_END_HEREDOC
|
|
|
308 |
) {
|
|
|
309 |
// We found the other end of the heredoc.
|
|
|
310 |
break;
|
|
|
311 |
}
|
|
|
312 |
|
|
|
313 |
if ($subTokenIsArray === true) {
|
|
|
314 |
$tokenContent .= $tokens[$i][1];
|
|
|
315 |
} else {
|
|
|
316 |
$tokenContent .= $tokens[$i];
|
|
|
317 |
}
|
|
|
318 |
}
|
|
|
319 |
|
|
|
320 |
$stackPtr = $i;
|
|
|
321 |
|
|
|
322 |
// Convert each line within the heredoc to a
|
|
|
323 |
// new token, so it conforms with other multiple line tokens.
|
|
|
324 |
$tokenLines = explode($eolChar, $tokenContent);
|
|
|
325 |
$numLines = count($tokenLines);
|
|
|
326 |
$newToken = array();
|
|
|
327 |
|
|
|
328 |
for ($j = 0; $j < $numLines; $j++) {
|
|
|
329 |
$newToken['content'] = $tokenLines[$j];
|
|
|
330 |
if ($j === ($numLines - 1)) {
|
|
|
331 |
if ($tokenLines[$j] === '') {
|
|
|
332 |
break;
|
|
|
333 |
}
|
|
|
334 |
} else {
|
|
|
335 |
$newToken['content'] .= $eolChar;
|
|
|
336 |
}
|
|
|
337 |
|
|
|
338 |
$newToken['code'] = T_HEREDOC;
|
|
|
339 |
$newToken['type'] = 'T_HEREDOC';
|
|
|
340 |
$finalTokens[$newStackPtr] = $newToken;
|
|
|
341 |
$newStackPtr++;
|
|
|
342 |
}
|
|
|
343 |
|
|
|
344 |
// Add the end heredoc token to the final array.
|
|
|
345 |
$finalTokens[$newStackPtr]
|
|
|
346 |
= PHP_CodeSniffer::standardiseToken($tokens[$stackPtr]);
|
|
|
347 |
$newStackPtr++;
|
|
|
348 |
|
|
|
349 |
// Continue, as we're done with this token.
|
|
|
350 |
continue;
|
|
|
351 |
}//end if
|
|
|
352 |
|
|
|
353 |
/*
|
|
|
354 |
If this token has newlines in its content, split each line up
|
|
|
355 |
and create a new token for each line. We do this so it's easier
|
|
|
356 |
to asertain where errors occur on a line.
|
|
|
357 |
Note that $token[1] is the token's content.
|
|
|
358 |
*/
|
|
|
359 |
|
|
|
360 |
if ($tokenIsArray === true && strpos($token[1], $eolChar) !== false) {
|
|
|
361 |
$tokenLines = explode($eolChar, $token[1]);
|
|
|
362 |
$numLines = count($tokenLines);
|
|
|
363 |
$tokenName = token_name($token[0]);
|
|
|
364 |
|
|
|
365 |
for ($i = 0; $i < $numLines; $i++) {
|
|
|
366 |
$newToken['content'] = $tokenLines[$i];
|
|
|
367 |
if ($i === ($numLines - 1)) {
|
|
|
368 |
if ($tokenLines[$i] === '') {
|
|
|
369 |
break;
|
|
|
370 |
}
|
|
|
371 |
} else {
|
|
|
372 |
$newToken['content'] .= $eolChar;
|
|
|
373 |
}
|
|
|
374 |
|
|
|
375 |
$newToken['type'] = $tokenName;
|
|
|
376 |
$newToken['code'] = $token[0];
|
|
|
377 |
$finalTokens[$newStackPtr] = $newToken;
|
|
|
378 |
$newStackPtr++;
|
|
|
379 |
}
|
|
|
380 |
} else {
|
|
|
381 |
$newToken = PHP_CodeSniffer::standardiseToken($token);
|
|
|
382 |
|
|
|
383 |
// This is a special condition for T_ARRAY tokens use to
|
|
|
384 |
// type hint function arguments as being arrays. We want to keep
|
|
|
385 |
// the parenthsis map clean, so let's tag these tokens as
|
|
|
386 |
// T_ARRAY_HINT.
|
|
|
387 |
if ($newToken['code'] === T_ARRAY) {
|
|
|
388 |
// Recalculate number of tokens.
|
|
|
389 |
$numTokens = count($tokens);
|
|
|
390 |
for ($i = $stackPtr; $i < $numTokens; $i++) {
|
|
|
391 |
if (is_array($tokens[$i]) === false) {
|
|
|
392 |
if ($tokens[$i] === '(') {
|
|
|
393 |
break;
|
|
|
394 |
}
|
|
|
395 |
} else if ($tokens[$i][0] === T_VARIABLE) {
|
|
|
396 |
$newToken['code'] = T_ARRAY_HINT;
|
|
|
397 |
$newToken['type'] = 'T_ARRAY_HINT';
|
|
|
398 |
break;
|
|
|
399 |
}
|
|
|
400 |
}
|
|
|
401 |
}
|
|
|
402 |
|
|
|
403 |
$finalTokens[$newStackPtr] = $newToken;
|
|
|
404 |
$newStackPtr++;
|
|
|
405 |
}//end if
|
|
|
406 |
}//end for
|
|
|
407 |
|
|
|
408 |
return $finalTokens;
|
|
|
409 |
|
|
|
410 |
}//end tokenizeString()
|
|
|
411 |
|
|
|
412 |
|
|
|
413 |
/**
|
|
|
414 |
* Performs additional processing after main tokenizing.
|
|
|
415 |
*
|
|
|
416 |
* @param array &$tokens The array of tokens to process.
|
|
|
417 |
* @param string $eolChar The EOL character to use for splitting strings.
|
|
|
418 |
*
|
|
|
419 |
* @return array
|
|
|
420 |
*/
|
|
|
421 |
public function processAdditional(&$tokens, $eolChar)
|
|
|
422 |
{
|
|
|
423 |
|
|
|
424 |
}//end processAdditional()
|
|
|
425 |
|
|
|
426 |
|
|
|
427 |
}//end class
|
|
|
428 |
|
|
|
429 |
?>
|