xref: /dokuwiki/inc/pageutils.php (revision ab9790ca81397622fcfd58faf56a69b68bd36257) 
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 */
10use dokuwiki\Utf8\PhpString;
11use dokuwiki\Utf8\Clean;
12use dokuwiki\File\Resolver;
13use dokuwiki\Extension\Event;
14use dokuwiki\ChangeLog\MediaChangeLog;
15use dokuwiki\ChangeLog\PageChangeLog;
16use dokuwiki\File\MediaResolver;
17use dokuwiki\File\PageResolver;
18
19/**
20 * Fetch the an ID from request
21 *
22 * Uses either standard $_REQUEST variable or extracts it from
23 * the full request URI when userewrite is set to 2
24 *
25 * For $param='id' $conf['start'] is returned if no id was found.
26 * If the second parameter is true (default) the ID is cleaned.
27 *
28 * @author Andreas Gohr <andi@splitbrain.org>
29 *
30 * @param string $param  the $_REQUEST variable name, default 'id'
31 * @param bool   $clean  if true, ID is cleaned
32 * @return string
33 */
34function getID($param = 'id', $clean = true)
35{
36    /** @var Input $INPUT */
37    global $INPUT;
38    global $conf;
39    global $ACT;
40
41    $id = $INPUT->str($param);
42
43    //construct page id from request URI
44    if (empty($id) && $conf['userewrite'] == 2) {
45        $request = $INPUT->server->str('REQUEST_URI');
46        $script = '';
47
48        //get the script URL
49        if ($conf['basedir']) {
50            $relpath = '';
51            if ($param != 'id') {
52                $relpath = 'lib/exe/';
53            }
54            $script = $conf['basedir'] . $relpath .
55                PhpString::basename($INPUT->server->str('SCRIPT_FILENAME'));
56        } elseif ($INPUT->server->str('PATH_INFO')) {
57            $request = $INPUT->server->str('PATH_INFO');
58        } elseif ($INPUT->server->str('SCRIPT_NAME')) {
59            $script = $INPUT->server->str('SCRIPT_NAME');
60        } elseif ($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')) {
61            $script = preg_replace(
62                '/^' . preg_quote($INPUT->server->str('DOCUMENT_ROOT'), '/') . '/',
63                '',
64                $INPUT->server->str('SCRIPT_FILENAME')
65            );
66            $script = '/' . $script;
67        }
68
69        //clean script and request (fixes a windows problem)
70        $script  = preg_replace('/\/\/+/', '/', $script);
71        $request = preg_replace('/\/\/+/', '/', $request);
72
73        //remove script URL and Querystring to gain the id
74        if (preg_match('/^' . preg_quote($script, '/') . '(.*)/', $request, $match)) {
75            $id = preg_replace('/\?.*/', '', $match[1]);
76        }
77        $id = urldecode($id);
78        //strip leading slashes
79        $id = preg_replace('!^/+!', '', $id);
80    }
81
82    // Namespace autolinking from URL
83    if (substr($id, -1) == ':' || ($conf['useslash'] && substr($id, -1) == '/')) {
84        if (page_exists($id . $conf['start'])) {
85            // start page inside namespace
86            $id .= $conf['start'];
87        } elseif (page_exists($id . noNS(cleanID($id)))) {
88            // page named like the NS inside the NS
89            $id .= noNS(cleanID($id));
90        } elseif (page_exists($id)) {
91            // page like namespace exists
92            $id = substr($id, 0, -1);
93        } else {
94            // fall back to default
95            $id .= $conf['start'];
96        }
97        if (isset($ACT) && $ACT === 'show') {
98            $urlParameters = $_GET;
99            if (isset($urlParameters['id'])) {
100                unset($urlParameters['id']);
101            }
102            send_redirect(wl($id, $urlParameters, true, '&'));
103        }
104    }
105    if ($clean) $id = cleanID($id);
106    if ($id === '' && $param == 'id') $id = $conf['start'];
107
108    return $id;
109}
110
111/**
112 * Remove unwanted chars from ID
113 *
114 * Cleans a given ID to only use allowed characters. Accented characters are
115 * converted to unaccented ones
116 *
117 * @author Andreas Gohr <andi@splitbrain.org>
118 *
119 * @param  string  $raw_id    The pageid to clean
120 * @param  boolean $ascii     Force ASCII
121 * @return string cleaned id
122 */
123function cleanID($raw_id, $ascii = false)
124{
125    global $conf;
126    static $sepcharpat = null;
127
128    global $cache_cleanid;
129    $cache = & $cache_cleanid;
130
131    // check if it's already in the memory cache
132    if (!$ascii && isset($cache[(string)$raw_id])) {
133        return $cache[(string)$raw_id];
134    }
135
136    $sepchar = $conf['sepchar'];
137    if ($sepcharpat == null) // build string only once to save clock cycles
138        $sepcharpat = '#\\' . $sepchar . '+#';
139
140    $id = trim((string)$raw_id);
141    $id = PhpString::strtolower($id);
142
143    //alternative namespace seperator
144    if ($conf['useslash']) {
145        $id = strtr($id, ';/', '::');
146    } else {
147        $id = strtr($id, ';/', ':' . $sepchar);
148    }
149
150    if ($conf['deaccent'] == 2 || $ascii) $id = Clean::romanize($id);
151    if ($conf['deaccent'] || $ascii) $id = Clean::deaccent($id, -1);
152
153    //remove specials
154    $id = Clean::stripspecials($id, $sepchar, '\*');
155
156    if ($ascii) $id = Clean::strip($id);
157
158    //clean up
159    $id = preg_replace($sepcharpat, $sepchar, $id);
160    $id = preg_replace('#:+#', ':', $id);
161    $id = trim($id, ':._-');
162    $id = preg_replace('#:[:\._\-]+#', ':', $id);
163    $id = preg_replace('#[:\._\-]+:#', ':', $id);
164
165    if (!$ascii) $cache[(string)$raw_id] = $id;
166    return($id);
167}
168
169/**
170 * Return namespacepart of a wiki ID
171 *
172 * @author Andreas Gohr <andi@splitbrain.org>
173 *
174 * @param string $id
175 * @return string|false the namespace part or false if the given ID has no namespace (root)
176 */
177function getNS($id)
178{
179    $pos = strrpos((string)$id, ':');
180    if ($pos !== false) {
181        return substr((string)$id, 0, $pos);
182    }
183    return false;
184}
185
186/**
187 * Returns the ID without the namespace
188 *
189 * @author Andreas Gohr <andi@splitbrain.org>
190 *
191 * @param string $id
192 * @return string
193 */
194function noNS($id)
195{
196    $pos = strrpos($id, ':');
197    if ($pos !== false) {
198        return substr($id, $pos + 1);
199    } else {
200        return $id;
201    }
202}
203
204/**
205 * Returns the current namespace
206 *
207 * @author Nathan Fritz <fritzn@crown.edu>
208 *
209 * @param string $id
210 * @return string
211 */
212function curNS($id)
213{
214    return noNS(getNS($id));
215}
216
217/**
218 * Returns the ID without the namespace or current namespace for 'start' pages
219 *
220 * @author Nathan Fritz <fritzn@crown.edu>
221 *
222 * @param string $id
223 * @return string
224 */
225function noNSorNS($id)
226{
227    global $conf;
228
229    $p = noNS($id);
230    if ($p === $conf['start'] || $p === false || $p === '') {
231        $p = curNS($id);
232        if ($p === false || $p === '') {
233            return $conf['start'];
234        }
235    }
236    return $p;
237}
238
239/**
240 * Creates a XHTML valid linkid from a given headline title
241 *
242 * @param string  $title   The headline title
243 * @param array|bool   $check   Existing IDs
244 * @return string the title
245 *
246 * @author Andreas Gohr <andi@splitbrain.org>
247 */
248function sectionID($title, &$check)
249{
250    $title = str_replace([':', '.'], '', cleanID($title));
251    $new = ltrim($title, '0123456789_-');
252    if (empty($new)) {
253        $title = 'section' . preg_replace('/[^0-9]+/', '', $title); //keep numbers from headline
254    } else {
255        $title = $new;
256    }
257
258    if (is_array($check)) {
259        $suffix = 0;
260        $candidateTitle = $title;
261        while (in_array($candidateTitle, $check)) {
262            $candidateTitle = $title . ++$suffix;
263        }
264        $check [] = $candidateTitle;
265        return $candidateTitle;
266    } else {
267        return $title;
268    }
269}
270
271/**
272 * Wiki page existence check
273 *
274 * parameters as for wikiFN
275 *
276 * @author Chris Smith <chris@jalakai.co.uk>
277 *
278 * @param string $id page id
279 * @param string|int $rev empty or revision timestamp
280 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
281 * @param bool $date_at
282 * @return bool exists?
283 */
284function page_exists($id, $rev = '', $clean = true, $date_at = false)
285{
286    $id = (explode('#', $id, 2))[0]; // #3608
287
288    if ($rev !== '' && $date_at) {
289        $pagelog = new PageChangeLog($id);
290        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
291        if ($pagelog_rev !== false)
292            $rev = $pagelog_rev;
293    }
294    return file_exists(wikiFN($id, $rev, $clean));
295}
296
297/**
298 * Media existence check
299 *
300 * @param string $id page id
301 * @param string|int $rev empty or revision timestamp
302 * @param bool $clean flag indicating that $id should be cleaned (see mediaFN as well)
303 * @param bool $date_at
304 * @return bool exists?
305 */
306function media_exists($id, $rev = '', $clean = true, $date_at = false)
307{
308    if ($rev !== '' && $date_at) {
309        $changeLog = new MediaChangeLog($id);
310        $changelog_rev = $changeLog->getLastRevisionAt($rev);
311        if ($changelog_rev !== false) {
312            $rev = $changelog_rev;
313        }
314    }
315    return file_exists(mediaFN($id, $rev, $clean));
316}
317
318/**
319 * returns the full path to the datafile specified by ID and optional revision
320 *
321 * The filename is URL encoded to protect Unicode chars
322 *
323 * @param  $raw_id  string   id of wikipage
324 * @param  $rev     int|string   page revision, empty string for current
325 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
326 *                           when $id is guaranteed to have been cleaned already.
327 * @return string full path
328 *
329 * @author Andreas Gohr <andi@splitbrain.org>
330 */
331function wikiFN($raw_id, $rev = '', $clean = true)
332{
333    global $conf;
334
335    global $cache_wikifn;
336    $cache = & $cache_wikifn;
337
338    $id = $raw_id;
339
340    if ($clean) $id = cleanID($id);
341    $id = str_replace(':', '/', $id);
342
343    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
344        return $cache[$id][$rev];
345    }
346
347    if (empty($rev)) {
348        $fn = $conf['datadir'] . '/' . utf8_encodeFN($id) . '.txt';
349    } else {
350        $fn = $conf['olddir'] . '/' . utf8_encodeFN($id) . '.' . $rev . '.txt';
351        if ($conf['compression']) {
352            //test for extensions here, we want to read both compressions
353            if (file_exists($fn . '.gz')) {
354                $fn .= '.gz';
355            } elseif (file_exists($fn . '.bz2')) {
356                $fn .= '.bz2';
357            } else {
358                //file doesnt exist yet, so we take the configured extension
359                $fn .= '.' . $conf['compression'];
360            }
361        }
362    }
363
364    if (!isset($cache[$id])) {
365        $cache[$id] = [];
366    }
367    $cache[$id][$rev] = $fn;
368    return $fn;
369}
370
371/**
372 * Returns the full path to the file for locking the page while editing.
373 *
374 * @author Ben Coburn <btcoburn@silicodon.net>
375 *
376 * @param string $id page id
377 * @return string full path
378 */
379function wikiLockFN($id)
380{
381    global $conf;
382    return $conf['lockdir'] . '/' . md5(cleanID($id)) . '.lock';
383}
384
385
386/**
387 * returns the full path to the meta file specified by ID and extension
388 *
389 * @author Steven Danz <steven-danz@kc.rr.com>
390 *
391 * @param string $id   page id
392 * @param string $ext  file extension
393 * @return string full path
394 */
395function metaFN($id, $ext)
396{
397    global $conf;
398    $id = cleanID($id);
399    $id = str_replace(':', '/', $id);
400
401    $fn = $conf['metadir'] . '/' . utf8_encodeFN($id) . $ext;
402    return $fn;
403}
404
405/**
406 * returns the full path to the media's meta file specified by ID and extension
407 *
408 * @author Kate Arzamastseva <pshns@ukr.net>
409 *
410 * @param string $id   media id
411 * @param string $ext  extension of media
412 * @return string
413 */
414function mediaMetaFN($id, $ext)
415{
416    global $conf;
417    $id = cleanID($id);
418    $id = str_replace(':', '/', $id);
419
420    $fn = $conf['mediametadir'] . '/' . utf8_encodeFN($id) . $ext;
421    return $fn;
422}
423
424/**
425 * returns an array of full paths to all metafiles of a given ID
426 *
427 * @author Esther Brunner <esther@kaffeehaus.ch>
428 * @author Michael Hamann <michael@content-space.de>
429 *
430 * @param string $id page id
431 * @return array
432 */
433function metaFiles($id)
434{
435    $basename = metaFN($id, '');
436    $files    = glob($basename . '.*', GLOB_MARK);
437    // filter files like foo.bar.meta when $id == 'foo'
438    return    $files ? preg_grep('/^' . preg_quote($basename, '/') . '\.[^.\/]*$/u', $files) : [];
439}
440
441/**
442 * returns the full path to the mediafile specified by ID
443 *
444 * The filename is URL encoded to protect Unicode chars
445 *
446 * @author Andreas Gohr <andi@splitbrain.org>
447 * @author Kate Arzamastseva <pshns@ukr.net>
448 *
449 * @param string     $id  media id
450 * @param string|int $rev empty string or revision timestamp
451 * @param bool $clean
452 *
453 * @return string full path
454 */
455function mediaFN($id, $rev = '', $clean = true)
456{
457    global $conf;
458    if ($clean) $id = cleanID($id);
459    $id = str_replace(':', '/', $id);
460    if (empty($rev)) {
461        $fn = $conf['mediadir'] . '/' . utf8_encodeFN($id);
462    } else {
463        $ext = mimetype($id);
464        $name = substr($id, 0, -1 * strlen($ext[0]) - 1);
465        $fn = $conf['mediaolddir'] . '/' . utf8_encodeFN($name . '.' . ( (int) $rev ) . '.' . $ext[0]);
466    }
467    return $fn;
468}
469
470/**
471 * Returns the full filepath to a localized file if local
472 * version isn't found the english one is returned
473 *
474 * @param  string $id  The id of the local file
475 * @param  string $ext The file extension (usually txt)
476 * @return string full filepath to localized file
477 *
478 * @author Andreas Gohr <andi@splitbrain.org>
479 */
480function localeFN($id, $ext = 'txt')
481{
482    global $conf;
483    $file = DOKU_CONF . 'lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
484    if (!file_exists($file)) {
485        $file = DOKU_INC . 'inc/lang/' . $conf['lang'] . '/' . $id . '.' . $ext;
486        if (!file_exists($file)) {
487            //fall back to english
488            $file = DOKU_INC . 'inc/lang/en/' . $id . '.' . $ext;
489        }
490    }
491    return $file;
492}
493
494/**
495 * Resolve relative paths in IDs
496 *
497 * Do not call directly use resolve_mediaid or resolve_pageid
498 * instead
499 *
500 * Partyly based on a cleanPath function found at
501 * http://php.net/manual/en/function.realpath.php#57016
502 *
503 * @deprecated 2020-09-30
504 * @param string $ns     namespace which is context of id
505 * @param string $id     relative id
506 * @param bool   $clean  flag indicating that id should be cleaned
507 * @return string
508 */
509function resolve_id($ns, $id, $clean = true)
510{
511    global $conf;
512    dbg_deprecated(Resolver::class . ' and its children');
513
514    // some pre cleaning for useslash:
515    if ($conf['useslash']) $id = str_replace('/', ':', $id);
516
517    // if the id starts with a dot we need to handle the
518    // relative stuff
519    if ($id && $id[0] == '.') {
520        // normalize initial dots without a colon
521        $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/', '\1\3:', $id);
522        // prepend the current namespace
523        $id = $ns . ':' . $id;
524
525        // cleanup relatives
526        $result = [];
527        $pathA  = explode(':', $id);
528        if (!$pathA[0]) $result[] = '';
529        foreach ($pathA as $dir) {
530            if ($dir == '..') {
531                if (end($result) == '..') {
532                    $result[] = '..';
533                } elseif (!array_pop($result)) {
534                    $result[] = '..';
535                }
536            } elseif ($dir && $dir != '.') {
537                $result[] = $dir;
538            }
539        }
540        if (!end($pathA)) $result[] = '';
541        $id = implode(':', $result);
542    } elseif ($ns !== false && strpos($id, ':') === false) {
543        //if link contains no namespace. add current namespace (if any)
544        $id = $ns . ':' . $id;
545    }
546
547    if ($clean) $id = cleanID($id);
548    return $id;
549}
550
551/**
552 * Returns a full media id
553 *
554 * @param string $ns namespace which is context of id
555 * @param string &$media (reference) relative media id, updated to resolved id
556 * @param bool &$exists (reference) updated with existance of media
557 * @param int|string $rev
558 * @param bool $date_at
559 * @deprecated 2020-09-30
560 */
561function resolve_mediaid($ns, &$media, &$exists, $rev = '', $date_at = false)
562{
563    dbg_deprecated(MediaResolver::class);
564    $resolver = new MediaResolver("$ns:deprecated");
565    $media = $resolver->resolveId($media, $rev, $date_at);
566    $exists = media_exists($media, $rev, false, $date_at);
567}
568
569/**
570 * Returns a full page id
571 *
572 * @deprecated 2020-09-30
573 * @param string $ns namespace which is context of id
574 * @param string &$page (reference) relative page id, updated to resolved id
575 * @param bool &$exists (reference) updated with existance of media
576 * @param string $rev
577 * @param bool $date_at
578 */
579function resolve_pageid($ns, &$page, &$exists, $rev = '', $date_at = false)
580{
581    dbg_deprecated(PageResolver::class);
582
583    global $ID;
584    if (getNS($ID) == $ns) {
585        $context = $ID; // this is usually the case
586    } else {
587        $context = "$ns:deprecated"; // only used when a different context namespace was given
588    }
589
590    $resolver = new PageResolver($context);
591    $page = $resolver->resolveId($page, $rev, $date_at);
592    $exists = page_exists($page, $rev, false, $date_at);
593}
594
595/**
596 * Returns the name of a cachefile from given data
597 *
598 * The needed directory is created by this function!
599 *
600 * @author Andreas Gohr <andi@splitbrain.org>
601 *
602 * @param string $data  This data is used to create a unique md5 name
603 * @param string $ext   This is appended to the filename if given
604 * @return string       The filename of the cachefile
605 */
606function getCacheName($data, $ext = '')
607{
608    global $conf;
609    $md5  = md5($data);
610    $file = $conf['cachedir'] . '/' . $md5[0] . '/' . $md5 . $ext;
611    io_makeFileDir($file);
612    return $file;
613}
614
615/**
616 * Checks a pageid against $conf['hidepages']
617 *
618 * @author Andreas Gohr <gohr@cosmocode.de>
619 *
620 * @param string $id page id
621 * @return bool
622 */
623function isHiddenPage($id)
624{
625    $data = ['id' => $id, 'hidden' => false];
626    Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
627    return $data['hidden'];
628}
629
630/**
631 * callback checks if page is hidden
632 *
633 * @param array $data event data    - see isHiddenPage()
634 */
635function _isHiddenPage(&$data)
636{
637    global $conf;
638    global $ACT;
639
640    if ($data['hidden']) return;
641    if (empty($conf['hidepages'])) return;
642    if ($ACT == 'admin') return;
643
644    if (preg_match('/' . $conf['hidepages'] . '/ui', ':' . $data['id'])) {
645        $data['hidden'] = true;
646    }
647}
648
649/**
650 * Reverse of isHiddenPage
651 *
652 * @author Andreas Gohr <gohr@cosmocode.de>
653 *
654 * @param string $id page id
655 * @return bool
656 */
657function isVisiblePage($id)
658{
659    return !isHiddenPage($id);
660}
661
662/**
663 * Format an id for output to a user
664 *
665 * Namespaces are denoted by a trailing “:*”. The root namespace is
666 * “*”. Output is escaped.
667 *
668 * @author Adrian Lang <lang@cosmocode.de>
669 *
670 * @param string $id page id
671 * @return string
672 */
673function prettyprint_id($id)
674{
675    if (!$id || $id === ':') {
676        return '*';
677    }
678    if ((substr($id, -1, 1) === ':')) {
679        $id .= '*';
680    }
681    return hsc($id);
682}
683
684/**
685 * Encode a UTF-8 filename to use on any filesystem
686 *
687 * Uses the 'fnencode' option to determine encoding
688 *
689 * When the second parameter is true the string will
690 * be encoded only if non ASCII characters are detected -
691 * This makes it safe to run it multiple times on the
692 * same string (default is true)
693 *
694 * @author Andreas Gohr <andi@splitbrain.org>
695 * @see    urlencode
696 *
697 * @param string $file file name
698 * @param bool   $safe if true, only encoded when non ASCII characters detected
699 * @return string
700 */
701function utf8_encodeFN($file, $safe = true)
702{
703    global $conf;
704    if ($conf['fnencode'] == 'utf-8') return $file;
705
706    if ($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#', $file)) {
707        return $file;
708    }
709
710    if ($conf['fnencode'] == 'safe') {
711        return SafeFN::encode($file);
712    }
713
714    $file = urlencode($file);
715    $file = str_replace('%2F', '/', $file);
716    return $file;
717}
718
719/**
720 * Decode a filename back to UTF-8
721 *
722 * Uses the 'fnencode' option to determine encoding
723 *
724 * @author Andreas Gohr <andi@splitbrain.org>
725 * @see    urldecode
726 *
727 * @param string $file file name
728 * @return string
729 */
730function utf8_decodeFN($file)
731{
732    global $conf;
733    if ($conf['fnencode'] == 'utf-8') return $file;
734
735    if ($conf['fnencode'] == 'safe') {
736        return SafeFN::decode($file);
737    }
738
739    return urldecode($file);
740}
741
742/**
743 * Find a page in the current namespace (determined from $ID) or any
744 * higher namespace that can be accessed by the current user,
745 * this condition can be overriden by an optional parameter.
746 *
747 * Used for sidebars, but can be used other stuff as well
748 *
749 * @todo   add event hook
750 *
751 * @param  string $page the pagename you're looking for
752 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
753 * @return false|string the full page id of the found page, false if any
754 */
755function page_findnearest($page, $useacl = true)
756{
757    if ((string) $page === '') return false;
758    global $ID;
759
760    $ns = $ID;
761    do {
762        $ns = getNS($ns);
763        $pageid = cleanID("$ns:$page");
764        if (page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)) {
765            return $pageid;
766        }
767    } while ($ns !== false);
768
769    return false;
770}
771