2 // PukiWiki - Yet another WikiWikiWeb clone.
3 // $Id: file.php,v 1.49 2006/03/05 14:57:35 henoheno Exp $
5 // 2002-2006 PukiWiki Developers Team
6 // 2001-2002 Originally written by yu-ji
7 // License: GPL v2 or (at your option) any later version
9 // File related functions
11 // Get source(wiki text) data of the page
12 function get_source($page = NULL, $lock = TRUE)
17 $path = get_filename($page);
20 $fp = @fopen($path, 'r');
21 if ($fp == FALSE) return $array;
25 // Removing line-feeds: Because file() doesn't remove them.
26 $array = str_replace("\r", '', file($path));
37 // Get last-modified filetime of the page
38 function get_filetime($page)
40 return is_page($page) ? filemtime(get_filename($page)) - LOCALZONE : 0;
43 // Get physical file name of the page
44 function get_filename($page)
46 return DATA_DIR . encode($page) . '.txt';
49 // Put a data(wiki text) into a physical file(diff, backup, text)
50 function page_write($page, $postdata, $notimestamp = FALSE)
54 if (PKWK_READONLY) return; // Do nothing
56 $postdata = make_str_rules($postdata);
58 // Create and write diff
59 $oldpostdata = is_page($page) ? join('', get_source($page)) : '';
60 $diffdata = do_diff($oldpostdata, $postdata);
61 file_write(DIFF_DIR, $page, $diffdata);
64 make_backup($page, $postdata == ''); // Is $postdata null?
67 file_write(DATA_DIR, $page, $postdata, $notimestamp);
71 $_diff = explode("\n", $diffdata);
72 $plus = join("\n", preg_replace('/^\+/', '', preg_grep('/^\+/', $_diff)));
73 $minus = join("\n", preg_replace('/^-/', '', preg_grep('/^-/', $_diff)));
74 tb_send($page, $plus, $minus);
80 // Modify original text with user-defined / system-defined rules
81 function make_str_rules($source)
83 global $str_rules, $fixed_heading_anchor;
85 $lines = explode("\n", $source);
86 $count = count($lines);
91 for ($i = 0; $i < $count; $i++) {
92 $line = & $lines[$i]; // Modify directly
94 // Ignore null string and preformatted texts
95 if ($line == '' || $line{0} == ' ' || $line{0} == "\t") continue;
99 if (! PKWKEXP_DISABLE_MULTILINE_PLUGIN_HACK &&
101 preg_match('/#[^{]*(\{\{+)\s*$/', $line, $matches)) {
102 // Multiline convert plugin start
104 $multiline = strlen($matches[1]); // Set specific number
107 if (! PKWKEXP_DISABLE_MULTILINE_PLUGIN_HACK &&
109 preg_match('/^\}{' . $multiline . '}\s*$/', $line)) {
110 // Multiline convert plugin end
115 if ($modify === FALSE) continue;
117 // Replace with $str_rules
118 foreach ($str_rules as $pattern => $replacement)
119 $line = preg_replace('/' . $pattern . '/', $replacement, $line);
121 // Adding fixed anchor into headings
122 if ($fixed_heading_anchor &&
123 preg_match('/^(\*{1,3}.*?)(?:\[#([A-Za-z][\w-]*)\]\s*)?$/', $line, $matches) &&
124 (! isset($matches[2]) || $matches[2] == '')) {
125 // Generate unique id
126 $anchor = generate_fixed_heading_anchor_id($matches[1]);
127 $line = rtrim($matches[1]) . ' [#' . $anchor . ']';
131 // Multiline part has no stopper
132 if (! PKWKEXP_DISABLE_MULTILINE_PLUGIN_HACK &&
133 $modify === FALSE && $multiline != 0)
134 $lines[] = str_repeat('}', $multiline);
136 return implode("\n", $lines);
140 function generate_fixed_heading_anchor_id($seed)
142 // A random alphabetic letter + 7 letters of random strings from md()
143 return chr(mt_rand(ord('a'), ord('z'))) .
144 substr(md5(uniqid(substr($seed, 0, 100), TRUE)),
148 // Read top N lines as an array
149 // (Use PHP file() function if you want to get ALL lines)
150 function file_head($file, $count = 1)
156 $fp = @fopen($file, 'r');
157 if ($fp == FALSE) return FALSE;
158 set_file_buffer($fp, 0);
161 while (! feof($fp)) {
162 $line = fgets($fp, $buffer);
163 if ($line != FALSE) $array[] = $line;
164 if (++$index >= $count) break;
168 if(! fclose($fp)) return FALSE;
174 function file_write($dir, $page, $str, $notimestamp = FALSE)
176 global $update_exec, $_msg_invalidiwn, $notify, $notify_diff_only, $notify_subject;
177 global $whatsdeleted, $maxshow_deleted;
179 if (PKWK_READONLY) return; // Do nothing
181 if (! is_pagename($page))
182 die_message(str_replace('$1', htmlspecialchars($page),
183 str_replace('$2', 'WikiName', $_msg_invalidiwn)));
185 $page = strip_bracket($page);
186 $file = $dir . encode($page) . '.txt';
190 if ($dir == DATA_DIR && file_exists($file)) {
193 add_recent($page, $whatsdeleted, '', $maxshow_deleted); // RecentDeleted
196 // File replacement (Edit)
197 $str = rtrim(preg_replace('/' . "\r" . '/', '', $str)) . "\n";
199 if ($notimestamp && file_exists($file))
200 $timestamp = filemtime($file) - LOCALZONE;
202 $fp = fopen($file, 'a') or die('fopen() failed: ' .
203 htmlspecialchars(basename($dir) . '/' . encode($page) . '.txt') .
205 'Maybe permission is not writable or filename is too long');
206 set_file_buffer($fp, 0);
219 if ($timestamp) pkwk_touch_file($file, $timestamp + LOCALZONE);
222 // Clear is_page() cache
223 is_page($page, TRUE);
225 // Update RecentChanges
226 if (! $timestamp && $dir == DATA_DIR)
229 // Execute $update_exec here
230 if ($update_exec && $dir == DATA_DIR)
231 system($update_exec . ' > /dev/null &');
233 if ($notify && $dir == DIFF_DIR) {
234 if ($notify_diff_only) $str = preg_replace('/^[^-+].*\n/m', '', $str);
236 $footer['ACTION'] = 'Page update';
237 $footer['PAGE'] = & $page;
238 $footer['URI'] = get_script_uri() . '?' . rawurlencode($page);
239 $footer['USER_AGENT'] = TRUE;
240 $footer['REMOTE_ADDR'] = TRUE;
242 pkwk_mail_notify($notify_subject, $str, $footer) or
243 die('pkwk_mail_notify(): Failed');
247 // Update RecentDeleted
248 function add_recent($page, $recentpage, $subject = '', $limit = 0)
250 if (PKWK_READONLY || $limit == 0 || $page == '' || $recentpage == '' ||
251 check_non_list($page)) return;
254 $lines = $matches = array();
255 foreach (get_source($recentpage) as $line)
256 if (preg_match('/^-(.+) - (\[\[.+\]\])$/', $line, $matches))
257 $lines[$matches[2]] = $line;
259 $_page = '[[' . $page . ']]';
261 // Remove a report about the same page
262 if (isset($lines[$_page])) unset($lines[$_page]);
265 array_unshift($lines, '-' . format_date(UTIME) . ' - ' . $_page .
266 htmlspecialchars($subject) . "\n");
268 // Get latest $limit reports
269 $lines = array_splice($lines, 0, $limit);
272 $fp = fopen(get_filename($recentpage), 'w') or
273 die_message('Cannot write page file ' .
274 htmlspecialchars($recentpage) .
275 '<br />Maybe permission is not writable or filename is too long');
276 set_file_buffer($fp, 0);
279 fputs($fp, '#freeze' . "\n");
280 fputs($fp, '#norelated' . "\n"); // :)
281 fputs($fp, join('', $lines));
286 // Update RecentChanges
287 function put_lastmodified()
289 global $maxshow, $whatsnew, $autolink;
291 if (PKWK_READONLY) return; // Do nothing
293 // Get whole page list
294 $pages = get_existpages();
296 // Check ALL filetime
297 $recent_pages = array();
298 foreach($pages as $page)
299 if ($page != $whatsnew && ! check_non_list($page))
300 $recent_pages[$page] = get_filetime($page);
302 // Sort decending order of last-modification date
303 arsort($recent_pages, SORT_NUMERIC);
305 // Create recent.dat (for recent.inc.php)
306 $file = 'recent.dat';
307 $fp = fopen(CACHE_DIR . $file, 'w') or
308 die_message('Cannot write file ' .
309 'CACHE_DIR/' . $file . '<br />' . "\n" .
310 'Maybe permission is not writable');
312 set_file_buffer($fp, 0);
315 foreach ($recent_pages as $page=>$time)
316 fputs($fp, $time . "\t" . $page . "\n");
320 // Create RecentChanges
321 $fp = fopen(get_filename($whatsnew), 'w') or
322 die_message('Cannot write page file ' .
323 htmlspecialchars($whatsnew) .
324 '<br />Maybe permission is not writable or filename is too long');
326 set_file_buffer($fp, 0);
330 // BugTrack2/106: Only variables can be passed by reference from PHP 5.0.5
331 $tmp_array = array_keys($recent_pages); // with array_splice()
333 foreach (array_splice($tmp_array, 0, $maxshow) as $page) {
334 $time = $recent_pages[$page];
335 $s_lastmod = htmlspecialchars(format_date($time));
336 $s_page = htmlspecialchars($page);
337 fputs($fp, '-' . $s_lastmod . ' - [[' . $s_page . ']]' . "\n");
339 fputs($fp, '#norelated' . "\n"); // :)
345 list($pattern, $pattern_a, $forceignorelist) =
346 get_autolink_pattern($pages);
348 $fp = fopen(CACHE_DIR . 'autolink.dat', 'w') or
349 die_message('Cannot write autolink file ' .
350 CACHE_DIR . '/autolink.dat' .
351 '<br />Maybe permission is not writable');
352 set_file_buffer($fp, 0);
355 fputs($fp, $pattern . "\n");
356 fputs($fp, $pattern_a . "\n");
357 fputs($fp, join("\t", $forceignorelist) . "\n");
363 // Get elapsed date of the page
364 function get_pg_passage($page, $sw = TRUE)
366 global $show_passage;
367 if (! $show_passage) return '';
369 $time = get_filetime($page);
370 $pg_passage = ($time != 0) ? get_passage($time) : '';
372 return $sw ? '<small>' . $pg_passage . '</small>' : ' ' . $pg_passage;
375 // Last-Modified header
376 function header_lastmod($page = NULL)
380 if ($lastmod && is_page($page)) {
382 header('Last-Modified: ' .
383 date('D, d M Y H:i:s', get_filetime($page)) . ' GMT');
387 // Get a page list of this wiki
388 function get_existpages($dir = DATA_DIR, $ext = '.txt')
392 $pattern = '((?:[0-9A-F]{2})+)';
393 if ($ext != '') $ext = preg_quote($ext, '/');
394 $pattern = '/^' . $pattern . $ext . '$/';
396 $dp = @opendir($dir) or
397 die_message($dir . ' is not found or not readable.');
399 while ($file = readdir($dp))
400 if (preg_match($pattern, $file, $matches))
401 $aryret[$file] = decode($matches[1]);
407 // Get PageReading(pronounce-annotated) data in an array()
408 function get_readings()
410 global $pagereading_enable, $pagereading_kanji2kana_converter;
411 global $pagereading_kanji2kana_encoding, $pagereading_chasen_path;
412 global $pagereading_kakasi_path, $pagereading_config_page;
413 global $pagereading_config_dict;
415 $pages = get_existpages();
418 foreach ($pages as $page)
419 $readings[$page] = '';
421 $deletedPage = FALSE;
423 foreach (get_source($pagereading_config_page) as $line) {
425 if(preg_match('/^-\[\[([^]]+)\]\]\s+(.+)$/', $line, $matches)) {
426 if(isset($readings[$matches[1]])) {
427 // This page is not clear how to be pronounced
428 $readings[$matches[1]] = $matches[2];
430 // This page seems deleted
436 // If enabled ChaSen/KAKASI execution
437 if($pagereading_enable) {
439 // Check there's non-clear-pronouncing page
440 $unknownPage = FALSE;
441 foreach ($readings as $page => $reading) {
448 // Execute ChaSen/KAKASI, and get annotation
450 switch(strtolower($pagereading_kanji2kana_converter)) {
452 if(! file_exists($pagereading_chasen_path))
453 die_message('ChaSen not found: ' . $pagereading_chasen_path);
455 $tmpfname = tempnam(realpath(CACHE_DIR), 'PageReading');
456 $fp = fopen($tmpfname, 'w') or
457 die_message('Cannot write temporary file "' . $tmpfname . '".' . "\n");
458 foreach ($readings as $page => $reading) {
459 if($reading != '') continue;
460 fputs($fp, mb_convert_encoding($page . "\n",
461 $pagereading_kanji2kana_encoding, SOURCE_ENCODING));
465 $chasen = "$pagereading_chasen_path -F %y $tmpfname";
466 $fp = popen($chasen, 'r');
469 die_message('ChaSen execution failed: ' . $chasen);
471 foreach ($readings as $page => $reading) {
472 if($reading != '') continue;
475 $line = mb_convert_encoding($line, SOURCE_ENCODING,
476 $pagereading_kanji2kana_encoding);
478 $readings[$page] = $line;
483 die_message('Temporary file can not be removed: ' . $tmpfname);
486 case 'kakasi': /*FALLTHROUGH*/
488 if(! file_exists($pagereading_kakasi_path))
489 die_message('KAKASI not found: ' . $pagereading_kakasi_path);
491 $tmpfname = tempnam(realpath(CACHE_DIR), 'PageReading');
492 $fp = fopen($tmpfname, 'w') or
493 die_message('Cannot write temporary file "' . $tmpfname . '".' . "\n");
494 foreach ($readings as $page => $reading) {
495 if($reading != '') continue;
496 fputs($fp, mb_convert_encoding($page . "\n",
497 $pagereading_kanji2kana_encoding, SOURCE_ENCODING));
501 $kakasi = "$pagereading_kakasi_path -kK -HK -JK < $tmpfname";
502 $fp = popen($kakasi, 'r');
505 die_message('KAKASI execution failed: ' . $kakasi);
508 foreach ($readings as $page => $reading) {
509 if($reading != '') continue;
512 $line = mb_convert_encoding($line, SOURCE_ENCODING,
513 $pagereading_kanji2kana_encoding);
515 $readings[$page] = $line;
520 die_message('Temporary file can not be removed: ' . $tmpfname);
524 $patterns = $replacements = $matches = array();
525 foreach (get_source($pagereading_config_dict) as $line) {
527 if(preg_match('|^ /([^/]+)/,\s*(.+)$|', $line, $matches)) {
528 $patterns[] = $matches[1];
529 $replacements[] = $matches[2];
532 foreach ($readings as $page => $reading) {
533 if($reading != '') continue;
535 $readings[$page] = $page;
536 foreach ($patterns as $no => $pattern)
537 $readings[$page] = mb_convert_kana(mb_ereg_replace($pattern,
538 $replacements[$no], $readings[$page]), 'aKCV');
543 die_message('Unknown kanji-kana converter: ' . $pagereading_kanji2kana_converter . '.');
548 if($unknownPage || $deletedPage) {
550 asort($readings); // Sort by pronouncing(alphabetical/reading) order
552 foreach ($readings as $page => $reading)
553 $body .= '-[[' . $page . ']] ' . $reading . "\n";
555 page_write($pagereading_config_page, $body);
559 // Pages that are not prounouncing-clear, return pagenames of themselves
560 foreach ($pages as $page) {
561 if($readings[$page] == '')
562 $readings[$page] = $page;
568 // Get a list of encoded files (must specify a directory and a suffix)
569 function get_existfiles($dir, $ext)
571 $pattern = '/^(?:[0-9A-F]{2})+' . preg_quote($ext, '/') . '$/';
573 $dp = @opendir($dir) or die_message($dir . ' is not found or not readable.');
574 while ($file = readdir($dp))
575 if (preg_match($pattern, $file))
576 $aryret[] = $dir . $file;
581 // Get a list of related pages of the page
582 function links_get_related($page)
584 global $vars, $related;
585 static $links = array();
587 if (isset($links[$page])) return $links[$page];
589 // If possible, merge related pages generated by make_link()
590 $links[$page] = ($page == $vars['page']) ? $related : array();
592 // Get repated pages from DB
593 $links[$page] += links_get_related_db($vars['page']);
595 return $links[$page];
598 // _If needed_, re-create the file to change/correct ownership into PHP's
599 // NOTE: Not works for Windows
600 function pkwk_chown($filename, $preserve_time = TRUE)
602 static $php_uid; // PHP's UID
604 if (! isset($php_uid)) {
605 if (extension_loaded('posix')) {
606 $php_uid = posix_getuid(); // Unix
608 $php_uid = 0; // Windows
612 // Lock for pkwk_chown()
613 $lockfile = CACHE_DIR . 'pkwk_chown.lock';
614 $flock = fopen($lockfile, 'a') or
615 die('pkwk_chown(): fopen() failed for: CACHEDIR/' .
616 basename(htmlspecialchars($lockfile)));
617 flock($flock, LOCK_EX) or die('pkwk_chown(): flock() failed for lock');
620 $stat = stat($filename) or
621 die('pkwk_chown(): stat() failed for: ' . basename(htmlspecialchars($filename)));
622 if ($stat[4] === $php_uid) {
623 // NOTE: Windows always here
624 $result = TRUE; // Seems the same UID. Nothing to do
626 $tmp = $filename . '.' . getmypid() . '.tmp';
628 // Lock source $filename to avoid file corruption
629 // NOTE: Not 'r+'. Don't check write permission here
630 $ffile = fopen($filename, 'r') or
631 die('pkwk_chown(): fopen() failed for: ' .
632 basename(htmlspecialchars($filename)));
634 // Try to chown by re-creating files
636 // * touch() before copy() is for 'rw-r--r--' instead of 'rwxr-xr-x' (with umask 022).
637 // * (PHP 4 < PHP 4.2.0) touch() with the third argument is not implemented and retuns NULL and Warn.
638 // * @unlink() before rename() is for Windows but here's for Unix only
639 flock($ffile, LOCK_EX) or die('pkwk_chown(): flock() failed');
640 $result = touch($tmp) && copy($filename, $tmp) &&
641 ($preserve_time ? (touch($tmp, $stat[9], $stat[8]) || touch($tmp, $stat[9])) : TRUE) &&
642 rename($tmp, $filename);
643 flock($ffile, LOCK_UN) or die('pkwk_chown(): flock() failed');
645 fclose($ffile) or die('pkwk_chown(): fclose() failed');
647 if ($result === FALSE) @unlink($tmp);
650 // Unlock for pkwk_chown()
651 flock($flock, LOCK_UN) or die('pkwk_chown(): flock() failed for lock');
652 fclose($flock) or die('pkwk_chown(): fclose() failed for lock');
657 // touch() with trying pkwk_chown()
658 function pkwk_touch_file($filename, $time = FALSE, $atime = FALSE)
660 // Is the owner incorrected and unable to correct?
661 if (! file_exists($filename) || pkwk_chown($filename)) {
662 if ($time === FALSE) {
663 $result = touch($filename);
664 } else if ($atime === FALSE) {
665 $result = touch($filename, $time);
667 $result = touch($filename, $time, $atime);
671 die('pkwk_touch_file(): Invalid UID and (not writable for the directory or not a flie): ' .
672 htmlspecialchars(basename($filename)));