1 <?php
2 
3 /**
4  * Utilities for handling pagenames
5  *
6  * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
7  * @author     Andreas Gohr <andi@splitbrain.org>
8  * @todo       Combine similar functions like {wiki,media,meta}FN()
9  */
10 
11 use dokuwiki\Utf8\PhpString;
12 use dokuwiki\Utf8\Clean;
13 use dokuwiki\File\Resolver;
14 use dokuwiki\Extension\Event;
15 use dokuwiki\ChangeLog\MediaChangeLog;
16 use dokuwiki\ChangeLog\PageChangeLog;
17 use dokuwiki\File\MediaResolver;
18 use dokuwiki\File\PageResolver;
19 
20 /**
21  * Fetch the an ID from request
22  *
23  * Uses either standard $_REQUEST variable or extracts it from
24  * the full request URI when userewrite is set to 2
25  *
26  * For $param='id' $conf['start'] is returned if no id was found.
27  * If the second parameter is true (default) the ID is cleaned.
28  *
29  * @author Andreas Gohr <andi@splitbrain.org>
30  *
31  * @param string $param  the $_REQUEST variable name, default 'id'
32  * @param bool   $clean  if true, ID is cleaned
33  * @return string
34  */
35 function getID($param = 'id', $clean = true)
36 {
37     /** @var Input $INPUT */
38     global $INPUT;
39     global $conf;
40     global $ACT;
41 
42     $id = $INPUT->str($param);
43 
44     //construct page id from request URI
45     if (empty($id) && $conf['userewrite'] == 2) {
46         $request = $INPUT->server->str('REQUEST_URI');
47         $script = '';
48 
49         //get the script URL
50         if ($conf['basedir']) {
51             $relpath = '';
52             if ($param != 'id') {
53                 $relpath = 'lib/exe/';
54             }
55             $script = $conf['basedir'] . $relpath .
56                 PhpString::basename($INPUT->server->str('SCRIPT_FILENAME'));
57         } elseif ($INPUT->server->str('PATH_INFO')) {
58             $request = $INPUT->server->str('PATH_INFO');
59         } elseif ($INPUT->server->str('SCRIPT_NAME')) {
60             $script = $INPUT->server->str('SCRIPT_NAME');
61         } elseif ($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')) {
62             $script = preg_replace(
63                 '/^' . preg_quote($INPUT->server->str('DOCUMENT_ROOT'), '/') . '/',
64                 '',
65                 $INPUT->server->str('SCRIPT_FILENAME')
66             );
67             $script = '/' . $script;
68         }
69 
70         //clean script and request (fixes a windows problem)
71         $script  = preg_replace('/\/\/+/', '/', $script);
72         $request = preg_replace('/\/\/+/', '/', $request);
73 
74         //remove script URL and Querystring to gain the id
75         if (preg_match('/^' . preg_quote($script, '/') . '(.*)/', $request, $match)) {
76             $id = preg_replace('/\?.*/', '', $match[1]);
77         }
78         $id = urldecode($id);
79         //strip leading slashes
80         $id = preg_replace('!^/+!', '', $id);
81     }
82 
83     // Namespace autolinking from URL
84     if (str_ends_with($id, ':') || ($conf['useslash'] && str_ends_with($id, '/'))) {
85         if (page_exists($id . $conf['start'])) {
86             // start page inside namespace
87             $id .= $conf['start'];
88         } elseif (page_exists($id . noNS(cleanID($id)))) {
89             // page named like the NS inside the NS
90             $id .= noNS(cleanID($id));
91         } elseif (page_exists($id)) {
92             // page like namespace exists
93             $id = substr($id, 0, -1);
94         } else {
95             // fall back to default
96             $id .= $conf['start'];
97         }
98         if (isset($ACT) && $ACT === 'show') {
99             $urlParameters = $_GET;
100             if (isset($urlParameters['id'])) {
101                 unset($urlParameters['id']);
102             }
103             send_redirect(wl($id, $urlParameters, true, '&'));
104         }
105     }
106     if ($clean) $id = cleanID($id);
107     if ($id === '' && $param == 'id') $id = $conf['start'];
108 
109     return $id;
110 }
111 
112 /**
113  * Remove unwanted chars from ID
114  *
115  * Cleans a given ID to only use allowed characters. Accented characters are
116  * converted to unaccented ones
117  *
118  * @author Andreas Gohr <andi@splitbrain.org>
119  *
120  * @param  string  $raw_id    The pageid to clean
121  * @param  boolean $ascii     Force ASCII
122  * @return string cleaned id
123  */
124 function cleanID($raw_id, $ascii = false)
125 {
126     global $conf;
127     static $sepcharpat = null;
128 
129     global $cache_cleanid;
130     $cache = & $cache_cleanid;
131 
132     // check if it's already in the memory cache
133     if (!$ascii && isset($cache[(string)$raw_id])) {
134         return $cache[(string)$raw_id];
135     }
136 
137     $sepchar = $conf['sepchar'];
138     if ($sepcharpat == null) // build string only once to save clock cycles
139         $sepcharpat = '#\\' . $sepchar . '+#';
140 
141     $id = trim((string)$raw_id);
142     $id = PhpString::strtolower($id);
143 
144     //alternative namespace seperator
145     if ($conf['useslash']) {
146         $id = strtr($id, ';/', '::');
147     } else {
148         $id = strtr($id, ';/', ':' . $sepchar);
149     }
150 
151     if ($conf['deaccent'] == 2 || $ascii) $id = Clean::romanize($id);
152     if ($conf['deaccent'] || $ascii) $id = Clean::deaccent($id, -1);
153 
154     //remove specials
155     $id = Clean::stripspecials($id, $sepchar, '\*');
156 
157     if ($ascii) $id = Clean::strip($id);
158 
159     //clean up
160     $id = preg_replace($sepcharpat, $sepchar, $id);
161     $id = preg_replace('#:+#', ':', $id);
162     $id = trim($id, ':._-');
163     $id = preg_replace('#:[:\._\-]+#', ':', $id);
164     $id = preg_replace('#[:\._\-]+:#', ':', $id);
165 
166     if (!$ascii) $cache[(string)$raw_id] = $id;
167     return($id);
168 }
169 
170 /**
171  * Return namespacepart of a wiki ID
172  *
173  * @author Andreas Gohr <andi@splitbrain.org>
174  *
175  * @param string $id
176  * @return string|false the namespace part or false if the given ID has no namespace (root)
177  */
178 function getNS($id)
179 {
180     $pos = strrpos((string)$id, ':');
181     if ($pos !== false) {
182         return substr((string)$id, 0, $pos);
183     }
184     return false;
185 }
186 
187 /**
188  * Returns the ID without the namespace
189  *
190  * @author Andreas Gohr <andi@splitbrain.org>
191  *
192  * @param string $id
193  * @return string
194  */
195 function noNS($id)
196 {
197     $pos = strrpos($id, ':');
198     if ($pos !== false) {
199         return substr($id, $pos + 1);
200     } else {
201         return $id;
202     }
203 }
204 
205 /**
206  * Returns the current namespace
207  *
208  * @author Nathan Fritz <fritzn@crown.edu>
209  *
210  * @param string $id
211  * @return string
212  */
213 function curNS($id)
214 {
215     return noNS(getNS($id));
216 }
217 
218 /**
219  * Returns the ID without the namespace or current namespace for 'start' pages
220  *
221  * @author Nathan Fritz <fritzn@crown.edu>
222  *
223  * @param string $id
224  * @return string
225  */
226 function noNSorNS($id)
227 {
228     global $conf;
229 
230     $p = noNS($id);
231     if ($p === $conf['start'] || $p === false || $p === '') {
232         $p = curNS($id);
233         if ($p === false || $p === '') {
234             return $conf['start'];
235         }
236     }
237     return $p;
238 }
239 
240 /**
241  * Creates a XHTML valid linkid from a given headline title
242  *
243  * @param string  $title   The headline title
244  * @param array|bool   $check   Existing IDs
245  * @return string the title
246  *
247  * @author Andreas Gohr <andi@splitbrain.org>
248  */
249 function sectionID($title, &$check)
250 {
251     $title = str_replace([':', '.'], '', cleanID($title));
252     $new = ltrim($title, '0123456789_-');
253     if (empty($new)) {
254         $title = 'section' . preg_replace('/[^0-9]+/', '', $title); //keep numbers from headline
255     } else {
256         $title = $new;
257     }
258 
259     if (is_array($check)) {
260         $suffix = 0;
261         $candidateTitle = $title;
262         while (in_array($candidateTitle, $check)) {
263             $candidateTitle = $title . ++$suffix;
264         }
265         $check [] = $candidateTitle;
266         return $candidateTitle;
267     } else {
268         return $title;
269     }
270 }
271 
272 /**
273  * Wiki page existence check
274  *
275  * parameters as for wikiFN
276  *
277  * @author Chris Smith <chris@jalakai.co.uk>
278  *
279  * @param string $id page id
280  * @param string|int $rev empty or revision timestamp
281  * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
282  * @param bool $date_at
283  * @return bool exists?
284  */
285 function page_exists($id, $rev = '', $clean = true, $date_at = false)
286 {
287     $id = (explode('#', $id, 2))[0]; // #3608
288 
289     if ($rev !== '' && $date_at) {
290         $pagelog = new PageChangeLog($id);
291         $pagelog_rev = $pagelog->getLastRevisionAt($rev);
292         if ($pagelog_rev !== false)
293             $rev = $pagelog_rev;
294     }
295     return file_exists(wikiFN($id, $rev, $clean));
296 }
297 
298 /**
299  * Media existence check
300  *
301  * @param string $id page id
302  * @param string|int $rev empty or revision timestamp
303  * @param bool $clean flag indicating that $id should be cleaned (see mediaFN as well)
304  * @param bool $date_at
305  * @return bool exists?
306  */
307 function media_exists($id, $rev = '', $clean = true, $date_at = false)
308 {
309     if ($rev !== '' && $date_at) {
310         $changeLog = new MediaChangeLog($id);
311         $changelog_rev = $changeLog->getLastRevisionAt($rev);
312         if ($changelog_rev !== false) {
313             $rev = $changelog_rev;
314         }
315     }
316     return file_exists(mediaFN($id, $rev, $clean));
317 }
318 
319 /**
320  * returns the full path to the datafile specified by ID and optional revision
321  *
322  * The filename is URL encoded to protect Unicode chars
323  *
324  * @param  $raw_id  string   id of wikipage
325  * @param  $rev     int|string   page revision, empty string for current
326  * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
327  *                           when $id is guaranteed to have been cleaned already.
328  * @return string full path
329  *
330  * @author Andreas Gohr <andi@splitbrain.org>
331  */
332 function wikiFN($raw_id, $rev = '', $clean = true)
333 {
334     global $conf;
335 
336     global $cache_wikifn;
337     $cache = & $cache_wikifn;
338 
339     $id = $raw_id;
340 
341     if ($clean) $id = cleanID($id);
342     $id = str_replace(':', '/', $id);
343 
344     if (isset($cache[$id]) && isset($cache[$id][$rev])) {
345         return $cache[$id][$rev];
346     }
347 
348     if (empty($rev)) {
349         $fn = $conf['datadir'] . '/' . utf8_encodeFN($id) . '.txt';
350     } else {
351         $fn = $conf['olddir'] . '/' . utf8_encodeFN($id) . '.' . $rev . '.txt';
352         if ($conf['compression']) {
353             //test for extensions here, we want to read both compressions
354             if (file_exists($fn . '.gz')) {
355                 $fn .= '.gz';
356             } elseif (file_exists($fn . '.bz2')) {
357                 $fn .= '.bz2';
358             } else {
359                 //file doesnt exist yet, so we take the configured extension
360                 $fn .= '.' . $conf['compression'];
361             }
362         }
363     }
364 
365     if (!isset($cache[$id])) {
366         $cache[$id] = [];
367     }
368     $cache[$id][$rev] = $fn;
369     return $fn;
370 }
371 
372 /**
373  * Returns the full path to the file for locking the page while editing.
374  *
375  * @author Ben Coburn <btcoburn@silicodon.net>
376  *
377  * @param string $id page id
378  * @return string full path
379  */
380 function wikiLockFN($id)
381 {
382     global $conf;
383     return $conf['lockdir'] . '/' . md5(cleanID($id)) . '.lock';
384 }
385 
386 
387 /**
388  * returns the full path to the meta file specified by ID and extension
389  *
390  * @author Steven Danz <steven-danz@kc.rr.com>
391  *
392  * @param string $id   page id
393  * @param string $ext  file extension
394  * @return string full path
395  */
396 function metaFN($id, $ext)
397 {
398     global $conf;
399     $id = cleanID($id);
400     $id = str_replace(':', '/', $id);
401 
402     $fn = $conf['metadir'] . '/' . utf8_encodeFN($id) . $ext;
403     return $fn;
404 }
405 
406 /**
407  * returns the full path to the media's meta file specified by ID and extension
408  *
409  * @author Kate Arzamastseva <pshns@ukr.net>
410  *
411  * @param string $id   media id
412  * @param string $ext  extension of media
413  * @return string
414  */
415 function mediaMetaFN($id, $ext)
416 {
417     global $conf;
418     $id = cleanID($id);
419     $id = str_replace(':', '/', $id);
420 
421     $fn = $conf['mediametadir'] . '/' . utf8_encodeFN($id) . $ext;
422     return $fn;
423 }
424 
425 /**
426  * returns an array of full paths to all metafiles of a given ID
427  *
428  * @author Esther Brunner <esther@kaffeehaus.ch>
429  * @author Michael Hamann <michael@content-space.de>
430  *
431  * @param string $id page id
432  * @return array
433  */
434 function metaFiles($id)
435 {
436     $basename = metaFN($id, '');
437     $files    = glob($basename . '.*', GLOB_MARK);
438     // filter files like foo.bar.meta when $id == 'foo'
439     return    $files ? preg_grep('/^' . preg_quote($basename, '/') . '\.[^.\/]*$/u', $files) : [];
440 }
441 
442 /**
443  * returns the full path to the mediafile specified by ID
444  *
445  * The filename is URL encoded to protect Unicode chars
446  *
447  * @author Andreas Gohr <andi@splitbrain.org>
448  * @author Kate Arzamastseva <pshns@ukr.net>
449  *
450  * @param string     $id  media id
451  * @param string|int $rev empty string or revision timestamp
452  * @param bool $clean
453  *
454  * @return string full path
455  */
456 function mediaFN($id, $rev = '', $clean = true)
457 {
458     global $conf;
459     if ($clean) $id = cleanID($id);
460     $id = str_replace(':', '/', $id);
461     if (empty($rev)) {
462         $fn = $conf['mediadir'] . '/' . utf8_encodeFN($id);
463     } else {
464         $ext = mimetype($id);
465         $name = substr($id, 0, -1 * strlen($ext[0]) - 1);
466         $fn = $conf['mediaolddir'] . '/' . utf8_encodeFN($name . '.' . ( (int) $rev ) . '.' . $ext[0]);
467     }
468     return $fn;
469 }
470 
471 /**
472  * Returns the full filepath to a localized file if local
473  * version isn't found the english one is returned
474  *
475  * @param  string $id  The id of the local file
476  * @param  string $ext The file extension (usually txt)
477  * @return string full filepath to localized file
478  *
479  * @author Andreas Gohr <andi@splitbrain.org>
480  */
481 function localeFN($id, $ext = 'txt')
482 {
483     global $conf;
484     $file = DOKU_CONF . 'lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
485     if (!file_exists($file)) {
486         $file = DOKU_INC . 'inc/lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
487         if (!file_exists($file)) {
488             //fall back to english
489             $file = DOKU_INC . 'inc/lang/en/' . $id . '.' . $ext;
490         }
491     }
492     return $file;
493 }
494 
495 /**
496  * Resolve relative paths in IDs
497  *
498  * Do not call directly use resolve_mediaid or resolve_pageid
499  * instead
500  *
501  * Partyly based on a cleanPath function found at
502  * http://php.net/manual/en/function.realpath.php#57016
503  *
504  * @deprecated 2020-09-30
505  * @param string $ns     namespace which is context of id
506  * @param string $id     relative id
507  * @param bool   $clean  flag indicating that id should be cleaned
508  * @return string
509  */
510 function resolve_id($ns, $id, $clean = true)
511 {
512     global $conf;
513     dbg_deprecated(Resolver::class . ' and its children');
514 
515     // some pre cleaning for useslash:
516     if ($conf['useslash']) $id = str_replace('/', ':', $id);
517 
518     // if the id starts with a dot we need to handle the
519     // relative stuff
520     if ($id && $id[0] == '.') {
521         // normalize initial dots without a colon
522         $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/', '\1\3:', $id);
523         // prepend the current namespace
524         $id = $ns . ':' . $id;
525 
526         // cleanup relatives
527         $result = [];
528         $pathA  = explode(':', $id);
529         if (!$pathA[0]) $result[] = '';
530         foreach ($pathA as $dir) {
531             if ($dir == '..') {
532                 if (end($result) == '..') {
533                     $result[] = '..';
534                 } elseif (!array_pop($result)) {
535                     $result[] = '..';
536                 }
537             } elseif ($dir && $dir != '.') {
538                 $result[] = $dir;
539             }
540         }
541         if (!end($pathA)) $result[] = '';
542         $id = implode(':', $result);
543     } elseif ($ns !== false && strpos($id, ':') === false) {
544         //if link contains no namespace. add current namespace (if any)
545         $id = $ns . ':' . $id;
546     }
547 
548     if ($clean) $id = cleanID($id);
549     return $id;
550 }
551 
552 /**
553  * Returns a full media id
554  *
555  * @param string $ns namespace which is context of id
556  * @param string &$media (reference) relative media id, updated to resolved id
557  * @param bool &$exists (reference) updated with existance of media
558  * @param int|string $rev
559  * @param bool $date_at
560  * @deprecated 2020-09-30
561  */
562 function resolve_mediaid($ns, &$media, &$exists, $rev = '', $date_at = false)
563 {
564     dbg_deprecated(MediaResolver::class);
565     $resolver = new MediaResolver("$ns:deprecated");
566     $media = $resolver->resolveId($media, $rev, $date_at);
567     $exists = media_exists($media, $rev, false, $date_at);
568 }
569 
570 /**
571  * Returns a full page id
572  *
573  * @deprecated 2020-09-30
574  * @param string $ns namespace which is context of id
575  * @param string &$page (reference) relative page id, updated to resolved id
576  * @param bool &$exists (reference) updated with existance of media
577  * @param string $rev
578  * @param bool $date_at
579  */
580 function resolve_pageid($ns, &$page, &$exists, $rev = '', $date_at = false)
581 {
582     dbg_deprecated(PageResolver::class);
583 
584     global $ID;
585     if (getNS($ID) == $ns) {
586         $context = $ID; // this is usually the case
587     } else {
588         $context = "$ns:deprecated"; // only used when a different context namespace was given
589     }
590 
591     $resolver = new PageResolver($context);
592     $page = $resolver->resolveId($page, $rev, $date_at);
593     $exists = page_exists($page, $rev, false, $date_at);
594 }
595 
596 /**
597  * Returns the name of a cachefile from given data
598  *
599  * The needed directory is created by this function!
600  *
601  * @author Andreas Gohr <andi@splitbrain.org>
602  *
603  * @param string $data  This data is used to create a unique md5 name
604  * @param string $ext   This is appended to the filename if given
605  * @return string       The filename of the cachefile
606  */
607 function getCacheName($data, $ext = '')
608 {
609     global $conf;
610     $md5  = md5($data);
611     $file = $conf['cachedir'] . '/' . $md5[0] . '/' . $md5 . $ext;
612     io_makeFileDir($file);
613     return $file;
614 }
615 
616 /**
617  * Checks a pageid against $conf['hidepages']
618  *
619  * @author Andreas Gohr <gohr@cosmocode.de>
620  *
621  * @param string $id page id
622  * @return bool
623  */
624 function isHiddenPage($id)
625 {
626     $data = ['id' => $id, 'hidden' => false];
627     Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
628     return $data['hidden'];
629 }
630 
631 /**
632  * callback checks if page is hidden
633  *
634  * @param array $data event data    - see isHiddenPage()
635  */
636 function _isHiddenPage(&$data)
637 {
638     global $conf;
639     global $ACT;
640 
641     if ($data['hidden']) return;
642     if (empty($conf['hidepages'])) return;
643     if ($ACT == 'admin') return;
644 
645     if (preg_match('/' . $conf['hidepages'] . '/ui', ':' . $data['id'])) {
646         $data['hidden'] = true;
647     }
648 }
649 
650 /**
651  * Reverse of isHiddenPage
652  *
653  * @author Andreas Gohr <gohr@cosmocode.de>
654  *
655  * @param string $id page id
656  * @return bool
657  */
658 function isVisiblePage($id)
659 {
660     return !isHiddenPage($id);
661 }
662 
663 /**
664  * Format an id for output to a user
665  *
666  * Namespaces are denoted by a trailing “:*”. The root namespace is
667  * “*”. Output is escaped.
668  *
669  * @author Adrian Lang <lang@cosmocode.de>
670  *
671  * @param string $id page id
672  * @return string
673  */
674 function prettyprint_id($id)
675 {
676     if (!$id || $id === ':') {
677         return '*';
678     }
679     if (str_ends_with($id, ':')) {
680         $id .= '*';
681     }
682     return hsc($id);
683 }
684 
685 /**
686  * Encode a UTF-8 filename to use on any filesystem
687  *
688  * Uses the 'fnencode' option to determine encoding
689  *
690  * When the second parameter is true the string will
691  * be encoded only if non ASCII characters are detected -
692  * This makes it safe to run it multiple times on the
693  * same string (default is true)
694  *
695  * @author Andreas Gohr <andi@splitbrain.org>
696  * @see    urlencode
697  *
698  * @param string $file file name
699  * @param bool   $safe if true, only encoded when non ASCII characters detected
700  * @return string
701  */
702 function utf8_encodeFN($file, $safe = true)
703 {
704     global $conf;
705     if ($conf['fnencode'] == 'utf-8') return $file;
706 
707     if ($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#', $file)) {
708         return $file;
709     }
710 
711     if ($conf['fnencode'] == 'safe') {
712         return SafeFN::encode($file);
713     }
714 
715     $file = urlencode($file);
716     $file = str_replace('%2F', '/', $file);
717     return $file;
718 }
719 
720 /**
721  * Decode a filename back to UTF-8
722  *
723  * Uses the 'fnencode' option to determine encoding
724  *
725  * @author Andreas Gohr <andi@splitbrain.org>
726  * @see    urldecode
727  *
728  * @param string $file file name
729  * @return string
730  */
731 function utf8_decodeFN($file)
732 {
733     global $conf;
734     if ($conf['fnencode'] == 'utf-8') return $file;
735 
736     if ($conf['fnencode'] == 'safe') {
737         return SafeFN::decode($file);
738     }
739 
740     return urldecode($file);
741 }
742 
743 /**
744  * Find a page in the current namespace (determined from $ID) or any
745  * higher namespace that can be accessed by the current user,
746  * this condition can be overriden by an optional parameter.
747  *
748  * Used for sidebars, but can be used other stuff as well
749  *
750  * @todo   add event hook
751  *
752  * @param  string $page the pagename you're looking for
753  * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
754  * @return false|string the full page id of the found page, false if any
755  */
756 function page_findnearest($page, $useacl = true)
757 {
758     if ((string) $page === '') return false;
759     global $ID;
760 
761     $ns = $ID;
762     do {
763         $ns = getNS($ns);
764         $pageid = cleanID("$ns:$page");
765         if (page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)) {
766             return $pageid;
767         }
768     } while ($ns !== false);
769 
770     return false;
771 }
772