0
|
1 |
<?php |
|
2 |
/** |
|
3 |
* General API for generating and formatting diffs - the differences between |
|
4 |
* two sequences of strings. |
|
5 |
* |
|
6 |
* The original PHP version of this code was written by Geoffrey T. Dairiki |
|
7 |
* <dairiki@dairiki.org>, and is used/adapted with his permission. |
|
8 |
* |
|
9 |
* Copyright 2004 Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
10 |
* Copyright 2004-2010 The Horde Project (http://www.horde.org/) |
|
11 |
* |
|
12 |
* See the enclosed file COPYING for license information (LGPL). If you did |
|
13 |
* not receive this file, see http://opensource.org/licenses/lgpl-license.php. |
|
14 |
* |
|
15 |
* @package Text_Diff |
|
16 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
17 |
*/ |
|
18 |
class Text_Diff { |
|
19 |
|
|
20 |
/** |
|
21 |
* Array of changes. |
|
22 |
* |
|
23 |
* @var array |
|
24 |
*/ |
|
25 |
var $_edits; |
|
26 |
|
|
27 |
/** |
|
28 |
* Computes diffs between sequences of strings. |
|
29 |
* |
|
30 |
* @param string $engine Name of the diffing engine to use. 'auto' |
|
31 |
* will automatically select the best. |
|
32 |
* @param array $params Parameters to pass to the diffing engine. |
|
33 |
* Normally an array of two arrays, each |
|
34 |
* containing the lines from a file. |
|
35 |
*/ |
|
36 |
function Text_Diff($engine, $params) |
|
37 |
{ |
|
38 |
// Backward compatibility workaround. |
|
39 |
if (!is_string($engine)) { |
|
40 |
$params = array($engine, $params); |
|
41 |
$engine = 'auto'; |
|
42 |
} |
|
43 |
|
|
44 |
if ($engine == 'auto') { |
|
45 |
$engine = extension_loaded('xdiff') ? 'xdiff' : 'native'; |
|
46 |
} else { |
|
47 |
$engine = basename($engine); |
|
48 |
} |
|
49 |
|
|
50 |
// WP #7391 |
|
51 |
require_once dirname(__FILE__).'/Diff/Engine/' . $engine . '.php'; |
|
52 |
$class = 'Text_Diff_Engine_' . $engine; |
|
53 |
$diff_engine = new $class(); |
|
54 |
|
|
55 |
$this->_edits = call_user_func_array(array($diff_engine, 'diff'), $params); |
|
56 |
} |
|
57 |
|
|
58 |
/** |
|
59 |
* Returns the array of differences. |
|
60 |
*/ |
|
61 |
function getDiff() |
|
62 |
{ |
|
63 |
return $this->_edits; |
|
64 |
} |
|
65 |
|
|
66 |
/** |
|
67 |
* returns the number of new (added) lines in a given diff. |
|
68 |
* |
|
69 |
* @since Text_Diff 1.1.0 |
|
70 |
* |
|
71 |
* @return integer The number of new lines |
|
72 |
*/ |
|
73 |
function countAddedLines() |
|
74 |
{ |
|
75 |
$count = 0; |
|
76 |
foreach ($this->_edits as $edit) { |
|
77 |
if (is_a($edit, 'Text_Diff_Op_add') || |
|
78 |
is_a($edit, 'Text_Diff_Op_change')) { |
|
79 |
$count += $edit->nfinal(); |
|
80 |
} |
|
81 |
} |
|
82 |
return $count; |
|
83 |
} |
|
84 |
|
|
85 |
/** |
|
86 |
* Returns the number of deleted (removed) lines in a given diff. |
|
87 |
* |
|
88 |
* @since Text_Diff 1.1.0 |
|
89 |
* |
|
90 |
* @return integer The number of deleted lines |
|
91 |
*/ |
|
92 |
function countDeletedLines() |
|
93 |
{ |
|
94 |
$count = 0; |
|
95 |
foreach ($this->_edits as $edit) { |
|
96 |
if (is_a($edit, 'Text_Diff_Op_delete') || |
|
97 |
is_a($edit, 'Text_Diff_Op_change')) { |
|
98 |
$count += $edit->norig(); |
|
99 |
} |
|
100 |
} |
|
101 |
return $count; |
|
102 |
} |
|
103 |
|
|
104 |
/** |
|
105 |
* Computes a reversed diff. |
|
106 |
* |
|
107 |
* Example: |
|
108 |
* <code> |
|
109 |
* $diff = new Text_Diff($lines1, $lines2); |
|
110 |
* $rev = $diff->reverse(); |
|
111 |
* </code> |
|
112 |
* |
|
113 |
* @return Text_Diff A Diff object representing the inverse of the |
|
114 |
* original diff. Note that we purposely don't return a |
|
115 |
* reference here, since this essentially is a clone() |
|
116 |
* method. |
|
117 |
*/ |
|
118 |
function reverse() |
|
119 |
{ |
|
120 |
if (version_compare(zend_version(), '2', '>')) { |
|
121 |
$rev = clone($this); |
|
122 |
} else { |
|
123 |
$rev = $this; |
|
124 |
} |
|
125 |
$rev->_edits = array(); |
|
126 |
foreach ($this->_edits as $edit) { |
|
127 |
$rev->_edits[] = $edit->reverse(); |
|
128 |
} |
|
129 |
return $rev; |
|
130 |
} |
|
131 |
|
|
132 |
/** |
|
133 |
* Checks for an empty diff. |
|
134 |
* |
|
135 |
* @return boolean True if two sequences were identical. |
|
136 |
*/ |
|
137 |
function isEmpty() |
|
138 |
{ |
|
139 |
foreach ($this->_edits as $edit) { |
|
140 |
if (!is_a($edit, 'Text_Diff_Op_copy')) { |
|
141 |
return false; |
|
142 |
} |
|
143 |
} |
|
144 |
return true; |
|
145 |
} |
|
146 |
|
|
147 |
/** |
|
148 |
* Computes the length of the Longest Common Subsequence (LCS). |
|
149 |
* |
|
150 |
* This is mostly for diagnostic purposes. |
|
151 |
* |
|
152 |
* @return integer The length of the LCS. |
|
153 |
*/ |
|
154 |
function lcs() |
|
155 |
{ |
|
156 |
$lcs = 0; |
|
157 |
foreach ($this->_edits as $edit) { |
|
158 |
if (is_a($edit, 'Text_Diff_Op_copy')) { |
|
159 |
$lcs += count($edit->orig); |
|
160 |
} |
|
161 |
} |
|
162 |
return $lcs; |
|
163 |
} |
|
164 |
|
|
165 |
/** |
|
166 |
* Gets the original set of lines. |
|
167 |
* |
|
168 |
* This reconstructs the $from_lines parameter passed to the constructor. |
|
169 |
* |
|
170 |
* @return array The original sequence of strings. |
|
171 |
*/ |
|
172 |
function getOriginal() |
|
173 |
{ |
|
174 |
$lines = array(); |
|
175 |
foreach ($this->_edits as $edit) { |
|
176 |
if ($edit->orig) { |
|
177 |
array_splice($lines, count($lines), 0, $edit->orig); |
|
178 |
} |
|
179 |
} |
|
180 |
return $lines; |
|
181 |
} |
|
182 |
|
|
183 |
/** |
|
184 |
* Gets the final set of lines. |
|
185 |
* |
|
186 |
* This reconstructs the $to_lines parameter passed to the constructor. |
|
187 |
* |
|
188 |
* @return array The sequence of strings. |
|
189 |
*/ |
|
190 |
function getFinal() |
|
191 |
{ |
|
192 |
$lines = array(); |
|
193 |
foreach ($this->_edits as $edit) { |
|
194 |
if ($edit->final) { |
|
195 |
array_splice($lines, count($lines), 0, $edit->final); |
|
196 |
} |
|
197 |
} |
|
198 |
return $lines; |
|
199 |
} |
|
200 |
|
|
201 |
/** |
|
202 |
* Removes trailing newlines from a line of text. This is meant to be used |
|
203 |
* with array_walk(). |
|
204 |
* |
|
205 |
* @param string $line The line to trim. |
|
206 |
* @param integer $key The index of the line in the array. Not used. |
|
207 |
*/ |
|
208 |
static function trimNewlines(&$line, $key) |
|
209 |
{ |
|
210 |
$line = str_replace(array("\n", "\r"), '', $line); |
|
211 |
} |
|
212 |
|
|
213 |
/** |
|
214 |
* Determines the location of the system temporary directory. |
|
215 |
* |
|
216 |
* @static |
|
217 |
* |
|
218 |
* @access protected |
|
219 |
* |
|
220 |
* @return string A directory name which can be used for temp files. |
|
221 |
* Returns false if one could not be found. |
|
222 |
*/ |
|
223 |
function _getTempDir() |
|
224 |
{ |
|
225 |
$tmp_locations = array('/tmp', '/var/tmp', 'c:\WUTemp', 'c:\temp', |
|
226 |
'c:\windows\temp', 'c:\winnt\temp'); |
|
227 |
|
|
228 |
/* Try PHP's upload_tmp_dir directive. */ |
|
229 |
$tmp = ini_get('upload_tmp_dir'); |
|
230 |
|
|
231 |
/* Otherwise, try to determine the TMPDIR environment variable. */ |
|
232 |
if (!strlen($tmp)) { |
|
233 |
$tmp = getenv('TMPDIR'); |
|
234 |
} |
|
235 |
|
|
236 |
/* If we still cannot determine a value, then cycle through a list of |
|
237 |
* preset possibilities. */ |
|
238 |
while (!strlen($tmp) && count($tmp_locations)) { |
|
239 |
$tmp_check = array_shift($tmp_locations); |
|
240 |
if (@is_dir($tmp_check)) { |
|
241 |
$tmp = $tmp_check; |
|
242 |
} |
|
243 |
} |
|
244 |
|
|
245 |
/* If it is still empty, we have failed, so return false; otherwise |
|
246 |
* return the directory determined. */ |
|
247 |
return strlen($tmp) ? $tmp : false; |
|
248 |
} |
|
249 |
|
|
250 |
/** |
|
251 |
* Checks a diff for validity. |
|
252 |
* |
|
253 |
* This is here only for debugging purposes. |
|
254 |
*/ |
|
255 |
function _check($from_lines, $to_lines) |
|
256 |
{ |
|
257 |
if (serialize($from_lines) != serialize($this->getOriginal())) { |
|
258 |
trigger_error("Reconstructed original doesn't match", E_USER_ERROR); |
|
259 |
} |
|
260 |
if (serialize($to_lines) != serialize($this->getFinal())) { |
|
261 |
trigger_error("Reconstructed final doesn't match", E_USER_ERROR); |
|
262 |
} |
|
263 |
|
|
264 |
$rev = $this->reverse(); |
|
265 |
if (serialize($to_lines) != serialize($rev->getOriginal())) { |
|
266 |
trigger_error("Reversed original doesn't match", E_USER_ERROR); |
|
267 |
} |
|
268 |
if (serialize($from_lines) != serialize($rev->getFinal())) { |
|
269 |
trigger_error("Reversed final doesn't match", E_USER_ERROR); |
|
270 |
} |
|
271 |
|
|
272 |
$prevtype = null; |
|
273 |
foreach ($this->_edits as $edit) { |
|
274 |
if ($prevtype == get_class($edit)) { |
|
275 |
trigger_error("Edit sequence is non-optimal", E_USER_ERROR); |
|
276 |
} |
|
277 |
$prevtype = get_class($edit); |
|
278 |
} |
|
279 |
|
|
280 |
return true; |
|
281 |
} |
|
282 |
|
|
283 |
} |
|
284 |
|
|
285 |
/** |
|
286 |
* @package Text_Diff |
|
287 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
288 |
*/ |
|
289 |
class Text_MappedDiff extends Text_Diff { |
|
290 |
|
|
291 |
/** |
|
292 |
* Computes a diff between sequences of strings. |
|
293 |
* |
|
294 |
* This can be used to compute things like case-insensitve diffs, or diffs |
|
295 |
* which ignore changes in white-space. |
|
296 |
* |
|
297 |
* @param array $from_lines An array of strings. |
|
298 |
* @param array $to_lines An array of strings. |
|
299 |
* @param array $mapped_from_lines This array should have the same size |
|
300 |
* number of elements as $from_lines. The |
|
301 |
* elements in $mapped_from_lines and |
|
302 |
* $mapped_to_lines are what is actually |
|
303 |
* compared when computing the diff. |
|
304 |
* @param array $mapped_to_lines This array should have the same number |
|
305 |
* of elements as $to_lines. |
|
306 |
*/ |
|
307 |
function Text_MappedDiff($from_lines, $to_lines, |
|
308 |
$mapped_from_lines, $mapped_to_lines) |
|
309 |
{ |
|
310 |
assert(count($from_lines) == count($mapped_from_lines)); |
|
311 |
assert(count($to_lines) == count($mapped_to_lines)); |
|
312 |
|
|
313 |
parent::Text_Diff($mapped_from_lines, $mapped_to_lines); |
|
314 |
|
|
315 |
$xi = $yi = 0; |
|
316 |
for ($i = 0; $i < count($this->_edits); $i++) { |
|
317 |
$orig = &$this->_edits[$i]->orig; |
|
318 |
if (is_array($orig)) { |
|
319 |
$orig = array_slice($from_lines, $xi, count($orig)); |
|
320 |
$xi += count($orig); |
|
321 |
} |
|
322 |
|
|
323 |
$final = &$this->_edits[$i]->final; |
|
324 |
if (is_array($final)) { |
|
325 |
$final = array_slice($to_lines, $yi, count($final)); |
|
326 |
$yi += count($final); |
|
327 |
} |
|
328 |
} |
|
329 |
} |
|
330 |
|
|
331 |
} |
|
332 |
|
|
333 |
/** |
|
334 |
* @package Text_Diff |
|
335 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
336 |
* |
|
337 |
* @access private |
|
338 |
*/ |
|
339 |
class Text_Diff_Op { |
|
340 |
|
|
341 |
var $orig; |
|
342 |
var $final; |
|
343 |
|
|
344 |
function &reverse() |
|
345 |
{ |
|
346 |
trigger_error('Abstract method', E_USER_ERROR); |
|
347 |
} |
|
348 |
|
|
349 |
function norig() |
|
350 |
{ |
|
351 |
return $this->orig ? count($this->orig) : 0; |
|
352 |
} |
|
353 |
|
|
354 |
function nfinal() |
|
355 |
{ |
|
356 |
return $this->final ? count($this->final) : 0; |
|
357 |
} |
|
358 |
|
|
359 |
} |
|
360 |
|
|
361 |
/** |
|
362 |
* @package Text_Diff |
|
363 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
364 |
* |
|
365 |
* @access private |
|
366 |
*/ |
|
367 |
class Text_Diff_Op_copy extends Text_Diff_Op { |
|
368 |
|
|
369 |
function Text_Diff_Op_copy($orig, $final = false) |
|
370 |
{ |
|
371 |
if (!is_array($final)) { |
|
372 |
$final = $orig; |
|
373 |
} |
|
374 |
$this->orig = $orig; |
|
375 |
$this->final = $final; |
|
376 |
} |
|
377 |
|
|
378 |
function &reverse() |
|
379 |
{ |
|
380 |
$reverse = new Text_Diff_Op_copy($this->final, $this->orig); |
|
381 |
return $reverse; |
|
382 |
} |
|
383 |
|
|
384 |
} |
|
385 |
|
|
386 |
/** |
|
387 |
* @package Text_Diff |
|
388 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
389 |
* |
|
390 |
* @access private |
|
391 |
*/ |
|
392 |
class Text_Diff_Op_delete extends Text_Diff_Op { |
|
393 |
|
|
394 |
function Text_Diff_Op_delete($lines) |
|
395 |
{ |
|
396 |
$this->orig = $lines; |
|
397 |
$this->final = false; |
|
398 |
} |
|
399 |
|
|
400 |
function &reverse() |
|
401 |
{ |
|
402 |
$reverse = new Text_Diff_Op_add($this->orig); |
|
403 |
return $reverse; |
|
404 |
} |
|
405 |
|
|
406 |
} |
|
407 |
|
|
408 |
/** |
|
409 |
* @package Text_Diff |
|
410 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
411 |
* |
|
412 |
* @access private |
|
413 |
*/ |
|
414 |
class Text_Diff_Op_add extends Text_Diff_Op { |
|
415 |
|
|
416 |
function Text_Diff_Op_add($lines) |
|
417 |
{ |
|
418 |
$this->final = $lines; |
|
419 |
$this->orig = false; |
|
420 |
} |
|
421 |
|
|
422 |
function &reverse() |
|
423 |
{ |
|
424 |
$reverse = new Text_Diff_Op_delete($this->final); |
|
425 |
return $reverse; |
|
426 |
} |
|
427 |
|
|
428 |
} |
|
429 |
|
|
430 |
/** |
|
431 |
* @package Text_Diff |
|
432 |
* @author Geoffrey T. Dairiki <dairiki@dairiki.org> |
|
433 |
* |
|
434 |
* @access private |
|
435 |
*/ |
|
436 |
class Text_Diff_Op_change extends Text_Diff_Op { |
|
437 |
|
|
438 |
function Text_Diff_Op_change($orig, $final) |
|
439 |
{ |
|
440 |
$this->orig = $orig; |
|
441 |
$this->final = $final; |
|
442 |
} |
|
443 |
|
|
444 |
function &reverse() |
|
445 |
{ |
|
446 |
$reverse = new Text_Diff_Op_change($this->final, $this->orig); |
|
447 |
return $reverse; |
|
448 |
} |
|
449 |
|
|
450 |
} |