3 * General API for generating and formatting diffs - the differences between
4 * two sequences of strings.
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.
9 * Copyright 2004 Geoffrey T. Dairiki <dairiki@dairiki.org>
10 * Copyright 2004-2010 The Horde Project (http://www.horde.org/)
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.
16 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
28 * Computes diffs between sequences of strings.
30 * @param string $engine Name of the diffing engine to use. 'auto'
31 * will automatically select the best.
32 * @param array $params Parameters to pass to the diffing engine.
33 * Normally an array of two arrays, each
34 * containing the lines from a file.
36 function Text_Diff($engine, $params)
38 // Backward compatibility workaround.
39 if (!is_string($engine)) {
40 $params = array($engine, $params);
44 if ($engine == 'auto') {
45 $engine = extension_loaded('xdiff') ? 'xdiff' : 'native';
47 $engine = basename($engine);
51 require_once dirname(__FILE__).'/Diff/Engine/' . $engine . '.php';
52 $class = 'Text_Diff_Engine_' . $engine;
53 $diff_engine = new $class();
55 $this->_edits = call_user_func_array(array($diff_engine, 'diff'), $params);
59 * Returns the array of differences.
67 * returns the number of new (added) lines in a given diff.
69 * @since Text_Diff 1.1.0
71 * @return integer The number of new lines
73 function countAddedLines()
76 foreach ($this->_edits as $edit) {
77 if (is_a($edit, 'Text_Diff_Op_add') ||
78 is_a($edit, 'Text_Diff_Op_change')) {
79 $count += $edit->nfinal();
86 * Returns the number of deleted (removed) lines in a given diff.
88 * @since Text_Diff 1.1.0
90 * @return integer The number of deleted lines
92 function countDeletedLines()
95 foreach ($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();
105 * Computes a reversed diff.
109 * $diff = new Text_Diff($lines1, $lines2);
110 * $rev = $diff->reverse();
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()
120 if (version_compare(zend_version(), '2', '>')) {
125 $rev->_edits = array();
126 foreach ($this->_edits as $edit) {
127 $rev->_edits[] = $edit->reverse();
133 * Checks for an empty diff.
135 * @return boolean True if two sequences were identical.
139 foreach ($this->_edits as $edit) {
140 if (!is_a($edit, 'Text_Diff_Op_copy')) {
148 * Computes the length of the Longest Common Subsequence (LCS).
150 * This is mostly for diagnostic purposes.
152 * @return integer The length of the LCS.
157 foreach ($this->_edits as $edit) {
158 if (is_a($edit, 'Text_Diff_Op_copy')) {
159 $lcs += count($edit->orig);
166 * Gets the original set of lines.
168 * This reconstructs the $from_lines parameter passed to the constructor.
170 * @return array The original sequence of strings.
172 function getOriginal()
175 foreach ($this->_edits as $edit) {
177 array_splice($lines, count($lines), 0, $edit->orig);
184 * Gets the final set of lines.
186 * This reconstructs the $to_lines parameter passed to the constructor.
188 * @return array The sequence of strings.
193 foreach ($this->_edits as $edit) {
195 array_splice($lines, count($lines), 0, $edit->final);
202 * Removes trailing newlines from a line of text. This is meant to be used
205 * @param string $line The line to trim.
206 * @param integer $key The index of the line in the array. Not used.
208 function trimNewlines(&$line, $key)
210 $line = str_replace(array("\n", "\r"), '', $line);
214 * Determines the location of the system temporary directory.
220 * @return string A directory name which can be used for temp files.
221 * Returns false if one could not be found.
223 function _getTempDir()
225 $tmp_locations = array('/tmp', '/var/tmp', 'c:\WUTemp', 'c:\temp',
226 'c:\windows\temp', 'c:\winnt\temp');
228 /* Try PHP's upload_tmp_dir directive. */
229 $tmp = ini_get('upload_tmp_dir');
231 /* Otherwise, try to determine the TMPDIR environment variable. */
233 $tmp = getenv('TMPDIR');
236 /* If we still cannot determine a value, then cycle through a list of
237 * preset possibilities. */
238 while (!strlen($tmp) && count($tmp_locations)) {
239 $tmp_check = array_shift($tmp_locations);
240 if (@is_dir($tmp_check)) {
245 /* If it is still empty, we have failed, so return false; otherwise
246 * return the directory determined. */
247 return strlen($tmp) ? $tmp : false;
251 * Checks a diff for validity.
253 * This is here only for debugging purposes.
255 function _check($from_lines, $to_lines)
257 if (serialize($from_lines) != serialize($this->getOriginal())) {
258 trigger_error("Reconstructed original doesn't match", E_USER_ERROR);
260 if (serialize($to_lines) != serialize($this->getFinal())) {
261 trigger_error("Reconstructed final doesn't match", E_USER_ERROR);
264 $rev = $this->reverse();
265 if (serialize($to_lines) != serialize($rev->getOriginal())) {
266 trigger_error("Reversed original doesn't match", E_USER_ERROR);
268 if (serialize($from_lines) != serialize($rev->getFinal())) {
269 trigger_error("Reversed final doesn't match", E_USER_ERROR);
273 foreach ($this->_edits as $edit) {
274 if ($prevtype == get_class($edit)) {
275 trigger_error("Edit sequence is non-optimal", E_USER_ERROR);
277 $prevtype = get_class($edit);
287 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
289 class Text_MappedDiff extends Text_Diff {
292 * Computes a diff between sequences of strings.
294 * This can be used to compute things like case-insensitve diffs, or diffs
295 * which ignore changes in white-space.
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 $to_lines.
307 function Text_MappedDiff($from_lines, $to_lines,
308 $mapped_from_lines, $mapped_to_lines)
310 assert(count($from_lines) == count($mapped_from_lines));
311 assert(count($to_lines) == count($mapped_to_lines));
313 parent::Text_Diff($mapped_from_lines, $mapped_to_lines);
316 for ($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));
323 $final = &$this->_edits[$i]->final;
324 if (is_array($final)) {
325 $final = array_slice($to_lines, $yi, count($final));
326 $yi += count($final);
335 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
346 trigger_error('Abstract method', E_USER_ERROR);
351 return $this->orig ? count($this->orig) : 0;
356 return $this->final ? count($this->final) : 0;
363 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
367 class Text_Diff_Op_copy extends Text_Diff_Op {
369 function Text_Diff_Op_copy($orig, $final = false)
371 if (!is_array($final)) {
375 $this->final = $final;
380 $reverse = &new Text_Diff_Op_copy($this->final, $this->orig);
388 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
392 class Text_Diff_Op_delete extends Text_Diff_Op {
394 function Text_Diff_Op_delete($lines)
396 $this->orig = $lines;
397 $this->final = false;
402 $reverse = &new Text_Diff_Op_add($this->orig);
410 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
414 class Text_Diff_Op_add extends Text_Diff_Op {
416 function Text_Diff_Op_add($lines)
418 $this->final = $lines;
424 $reverse = &new Text_Diff_Op_delete($this->final);
432 * @author Geoffrey T. Dairiki <dairiki@dairiki.org>
436 class Text_Diff_Op_change extends Text_Diff_Op {
438 function Text_Diff_Op_change($orig, $final)
441 $this->final = $final;
446 $reverse = &new Text_Diff_Op_change($this->final, $this->orig);