xref: /dokuwiki/inc/pageutils.php (revision 093fe67e98c0cdb4b73fd46938e49b64971483c2) 
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
11use dokuwiki\Utf8\PhpString;
12use dokuwiki\Utf8\Clean;
13use dokuwiki\File\Resolver;
14use dokuwiki\Extension\Event;
15use dokuwiki\ChangeLog\MediaChangeLog;
16use dokuwiki\ChangeLog\PageChangeLog;
17use dokuwiki\File\MediaResolver;
18use 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   $clean  if true, ID is cleaned
33 * @return string
34 */
35function getID($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 $ascii     Force ASCII
122 * @return string cleaned id
123 */
124function cleanID($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 $id
176 * @return string|false the namespace part or false if the given ID has no namespace (root)
177 */
178function getNS($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 $id
193 * @return string
194 */
195function noNS($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 $id
211 * @return string
212 */
213function curNS($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 $id
224 * @return string
225 */
226function noNSorNS($id)
227{
228    global $conf;
229
230    $p = noNS($id);
231    if (in_array($p, [$conf['start'], false, ''], true)) {
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   $check   Existing IDs
245 * @return string the title
246 *
247 * @author Andreas Gohr <andi@splitbrain.org>
248 */
249function sectionID($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 $date_at
283 * @return bool exists?
284 */
285function page_exists($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 $date_at
305 * @return bool exists?
306 */
307function media_exists($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 $id is guaranteed to have been cleaned already.
328 * @return string full path
329 *
330 * @author Andreas Gohr <andi@splitbrain.org>
331 */
332function wikiFN($raw_id, $rev = '', $clean = true)
333{
334    global $conf;
335
336    global $cache_wikifn;
337    $cache = & $cache_wikifn;
338
339    $id = $raw_id;
340    $rev = (int) $rev; // any falsy rev will be rev 0 in the cache
341
342    if ($clean) $id = cleanID($id);
343    $id = str_replace(':', '/', $id);
344
345    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
346        return $cache[$id][$rev];
347    }
348
349    if (empty($rev)) {
350        $fn = $conf['datadir'] . '/' . utf8_encodeFN($id) . '.txt';
351    } else {
352        $fn = $conf['olddir'] . '/' . utf8_encodeFN($id) . '.' . $rev . '.txt';
353        if ($conf['compression']) {
354            //test for extensions here, we want to read both compressions
355            if (file_exists($fn . '.gz')) {
356                $fn .= '.gz';
357            } elseif (file_exists($fn . '.bz2')) {
358                $fn .= '.bz2';
359            } else {
360                //file doesnt exist yet, so we take the configured extension
361                $fn .= '.' . $conf['compression'];
362            }
363        }
364    }
365
366    if (!isset($cache[$id])) {
367        $cache[$id] = [];
368    }
369    $cache[$id][$rev] = $fn;
370    return $fn;
371}
372
373/**
374 * Returns the full path to the file for locking the page while editing.
375 *
376 * @author Ben Coburn <btcoburn@silicodon.net>
377 *
378 * @param string $id page id
379 * @return string full path
380 */
381function wikiLockFN($id)
382{
383    global $conf;
384    return $conf['lockdir'] . '/' . md5(cleanID($id)) . '.lock';
385}
386
387
388/**
389 * returns the full path to the meta file specified by ID and extension
390 *
391 * @author Steven Danz <steven-danz@kc.rr.com>
392 *
393 * @param string $id   page id
394 * @param string $ext  file extension
395 * @return string full path
396 */
397function metaFN($id, $ext)
398{
399    global $conf;
400    $id = cleanID($id);
401    $id = str_replace(':', '/', $id);
402
403    $fn = $conf['metadir'] . '/' . utf8_encodeFN($id) . $ext;
404    return $fn;
405}
406
407/**
408 * returns the full path to the media's meta file specified by ID and extension
409 *
410 * @author Kate Arzamastseva <pshns@ukr.net>
411 *
412 * @param string $id   media id
413 * @param string $ext  extension of media
414 * @return string
415 */
416function mediaMetaFN($id, $ext)
417{
418    global $conf;
419    $id = cleanID($id);
420    $id = str_replace(':', '/', $id);
421
422    $fn = $conf['mediametadir'] . '/' . utf8_encodeFN($id) . $ext;
423    return $fn;
424}
425
426/**
427 * returns an array of full paths to all metafiles of a given ID
428 *
429 * @author Esther Brunner <esther@kaffeehaus.ch>
430 * @author Michael Hamann <michael@content-space.de>
431 *
432 * @param string $id page id
433 * @return array
434 */
435function metaFiles($id)
436{
437    $basename = metaFN($id, '');
438    $files    = glob($basename . '.*', GLOB_MARK);
439    // filter files like foo.bar.meta when $id == 'foo'
440    return    $files ? preg_grep('/^' . preg_quote($basename, '/') . '\.[^.\/]*$/u', $files) : [];
441}
442
443/**
444 * returns the full path to the mediafile specified by ID
445 *
446 * The filename is URL encoded to protect Unicode chars
447 *
448 * @author Andreas Gohr <andi@splitbrain.org>
449 * @author Kate Arzamastseva <pshns@ukr.net>
450 *
451 * @param string     $id  media id
452 * @param string|int $rev empty string or revision timestamp
453 * @param bool $clean
454 *
455 * @return string full path
456 */
457function mediaFN($id, $rev = '', $clean = true)
458{
459    global $conf;
460    if ($clean) $id = cleanID($id);
461    $id = str_replace(':', '/', $id);
462    if (empty($rev)) {
463        $fn = $conf['mediadir'] . '/' . utf8_encodeFN($id);
464    } else {
465        $ext = mimetype($id);
466        $name = substr($id, 0, -1 * strlen($ext[0]) - 1);
467        $fn = $conf['mediaolddir'] . '/' . utf8_encodeFN($name . '.' . ( (int) $rev ) . '.' . $ext[0]);
468    }
469    return $fn;
470}
471
472/**
473 * Returns the full filepath to a localized file if local
474 * version isn't found the english one is returned
475 *
476 * @param  string $id  The id of the local file
477 * @param  string $ext The file extension (usually txt)
478 * @return string full filepath to localized file
479 *
480 * @author Andreas Gohr <andi@splitbrain.org>
481 */
482function localeFN($id, $ext = 'txt')
483{
484    global $conf;
485    $file = DOKU_CONF . 'lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
486    if (!file_exists($file)) {
487        $file = DOKU_INC . 'inc/lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
488        if (!file_exists($file)) {
489            //fall back to english
490            $file = DOKU_INC . 'inc/lang/en/' . $id . '.' . $ext;
491        }
492    }
493    return $file;
494}
495
496/**
497 * Resolve relative paths in IDs
498 *
499 * Do not call directly use resolve_mediaid or resolve_pageid
500 * instead
501 *
502 * Partyly based on a cleanPath function found at
503 * http://php.net/manual/en/function.realpath.php#57016
504 *
505 * @deprecated 2020-09-30
506 * @param string $ns     namespace which is context of id
507 * @param string $id     relative id
508 * @param bool   $clean  flag indicating that id should be cleaned
509 * @return string
510 */
511function resolve_id($ns, $id, $clean = true)
512{
513    global $conf;
514    dbg_deprecated(Resolver::class . ' and its children');
515
516    // some pre cleaning for useslash:
517    if ($conf['useslash']) $id = str_replace('/', ':', $id);
518
519    // if the id starts with a dot we need to handle the
520    // relative stuff
521    if ($id && $id[0] == '.') {
522        // normalize initial dots without a colon
523        $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/', '\1\3:', $id);
524        // prepend the current namespace
525        $id = $ns . ':' . $id;
526
527        // cleanup relatives
528        $result = [];
529        $pathA  = explode(':', $id);
530        if (!$pathA[0]) $result[] = '';
531        foreach ($pathA as $dir) {
532            if ($dir == '..') {
533                if (end($result) == '..') {
534                    $result[] = '..';
535                } elseif (!array_pop($result)) {
536                    $result[] = '..';
537                }
538            } elseif ($dir && $dir != '.') {
539                $result[] = $dir;
540            }
541        }
542        if (!end($pathA)) $result[] = '';
543        $id = implode(':', $result);
544    } elseif ($ns !== false && !str_contains($id, ':')) {
545        //if link contains no namespace. add current namespace (if any)
546        $id = $ns . ':' . $id;
547    }
548
549    if ($clean) $id = cleanID($id);
550    return $id;
551}
552
553/**
554 * Returns a full media id
555 *
556 * @param string $ns namespace which is context of id
557 * @param string &$media (reference) relative media id, updated to resolved id
558 * @param bool &$exists (reference) updated with existance of media
559 * @param int|string $rev
560 * @param bool $date_at
561 * @deprecated 2020-09-30
562 */
563function resolve_mediaid($ns, &$media, &$exists, $rev = '', $date_at = false)
564{
565    dbg_deprecated(MediaResolver::class);
566    $resolver = new MediaResolver("$ns:deprecated");
567    $media = $resolver->resolveId($media, $rev, $date_at);
568    $exists = media_exists($media, $rev, false, $date_at);
569}
570
571/**
572 * Returns a full page id
573 *
574 * @deprecated 2020-09-30
575 * @param string $ns namespace which is context of id
576 * @param string &$page (reference) relative page id, updated to resolved id
577 * @param bool &$exists (reference) updated with existance of media
578 * @param string $rev
579 * @param bool $date_at
580 */
581function resolve_pageid($ns, &$page, &$exists, $rev = '', $date_at = false)
582{
583    dbg_deprecated(PageResolver::class);
584
585    global $ID;
586    if (getNS($ID) == $ns) {
587        $context = $ID; // this is usually the case
588    } else {
589        $context = "$ns:deprecated"; // only used when a different context namespace was given
590    }
591
592    $resolver = new PageResolver($context);
593    $page = $resolver->resolveId($page, $rev, $date_at);
594    $exists = page_exists($page, $rev, false, $date_at);
595}
596
597/**
598 * Returns the name of a cachefile from given data
599 *
600 * The needed directory is created by this function!
601 *
602 * @author Andreas Gohr <andi@splitbrain.org>
603 *
604 * @param string $data  This data is used to create a unique md5 name
605 * @param string $ext   This is appended to the filename if given
606 * @return string       The filename of the cachefile
607 */
608function getCacheName($data, $ext = '')
609{
610    global $conf;
611    $md5  = md5($data);
612    $file = $conf['cachedir'] . '/' . $md5[0] . '/' . $md5 . $ext;
613    io_makeFileDir($file);
614    return $file;
615}
616
617/**
618 * Checks a pageid against $conf['hidepages']
619 *
620 * @author Andreas Gohr <gohr@cosmocode.de>
621 *
622 * @param string $id page id
623 * @return bool
624 */
625function isHiddenPage($id)
626{
627    $data = ['id' => $id, 'hidden' => false];
628    Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
629    return $data['hidden'];
630}
631
632/**
633 * callback checks if page is hidden
634 *
635 * @param array $data event data    - see isHiddenPage()
636 */
637function _isHiddenPage(&$data)
638{
639    global $conf;
640    global $ACT;
641
642    if ($data['hidden']) return;
643    if (empty($conf['hidepages'])) return;
644    if ($ACT == 'admin') return;
645
646    if (preg_match('/' . $conf['hidepages'] . '/ui', ':' . $data['id'])) {
647        $data['hidden'] = true;
648    }
649}
650
651/**
652 * Reverse of isHiddenPage
653 *
654 * @author Andreas Gohr <gohr@cosmocode.de>
655 *
656 * @param string $id page id
657 * @return bool
658 */
659function isVisiblePage($id)
660{
661    return !isHiddenPage($id);
662}
663
664/**
665 * Format an id for output to a user
666 *
667 * Namespaces are denoted by a trailing “:*”. The root namespace is
668 * “*”. Output is escaped.
669 *
670 * @author Adrian Lang <lang@cosmocode.de>
671 *
672 * @param string $id page id
673 * @return string
674 */
675function prettyprint_id($id)
676{
677    if (!$id || $id === ':') {
678        return '*';
679    }
680    if (str_ends_with($id, ':')) {
681        $id .= '*';
682    }
683    return hsc($id);
684}
685
686/**
687 * Encode a UTF-8 filename to use on any filesystem
688 *
689 * Uses the 'fnencode' option to determine encoding
690 *
691 * When the second parameter is true the string will
692 * be encoded only if non ASCII characters are detected -
693 * This makes it safe to run it multiple times on the
694 * same string (default is true)
695 *
696 * @author Andreas Gohr <andi@splitbrain.org>
697 * @see    urlencode
698 *
699 * @param string $file file name
700 * @param bool   $safe if true, only encoded when non ASCII characters detected
701 * @return string
702 */
703function utf8_encodeFN($file, $safe = true)
704{
705    global $conf;
706    if ($conf['fnencode'] == 'utf-8') return $file;
707
708    if ($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#', $file)) {
709        return $file;
710    }
711
712    if ($conf['fnencode'] == 'safe') {
713        return SafeFN::encode($file);
714    }
715
716    $file = urlencode($file);
717    $file = str_replace('%2F', '/', $file);
718    return $file;
719}
720
721/**
722 * Decode a filename back to UTF-8
723 *
724 * Uses the 'fnencode' option to determine encoding
725 *
726 * @author Andreas Gohr <andi@splitbrain.org>
727 * @see    urldecode
728 *
729 * @param string $file file name
730 * @return string
731 */
732function utf8_decodeFN($file)
733{
734    global $conf;
735    if ($conf['fnencode'] == 'utf-8') return $file;
736
737    if ($conf['fnencode'] == 'safe') {
738        return SafeFN::decode($file);
739    }
740
741    return urldecode($file);
742}
743
744/**
745 * Find a page in the current namespace (determined from $ID) or any
746 * higher namespace that can be accessed by the current user,
747 * this condition can be overriden by an optional parameter.
748 *
749 * Used for sidebars, but can be used other stuff as well
750 *
751 * @todo   add event hook
752 *
753 * @param  string $page the pagename you're looking for
754 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
755 * @return false|string the full page id of the found page, false if any
756 */
757function page_findnearest($page, $useacl = true)
758{
759    if ((string) $page === '') return false;
760    global $ID;
761
762    $ns = $ID;
763    do {
764        $ns = getNS($ns);
765        $pageid = cleanID("$ns:$page");
766        if (page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)) {
767            return $pageid;
768        }
769    } while ($ns !== false);
770
771    return false;
772}
773