xref: /dokuwiki/inc/pageutils.php (revision dc23fc9a4a335f0a8bc02f7cf3eb80abfd2d0577) 
1<?php
2/**
3 * Utilities for handling pagenames
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 * @todo       Combine similar functions like {wiki,media,meta}FN()
8 */
9
10use dokuwiki\ChangeLog\MediaChangeLog;
11use dokuwiki\ChangeLog\PageChangeLog;
12use dokuwiki\Utils\MediaResolver;
13use dokuwiki\Utils\PageResolver;
14
15/**
16 * Fetch the an ID from request
17 *
18 * Uses either standard $_REQUEST variable or extracts it from
19 * the full request URI when userewrite is set to 2
20 *
21 * For $param='id' $conf['start'] is returned if no id was found.
22 * If the second parameter is true (default) the ID is cleaned.
23 *
24 * @author Andreas Gohr <andi@splitbrain.org>
25 *
26 * @param string $param  the $_REQUEST variable name, default 'id'
27 * @param bool   $clean  if true, ID is cleaned
28 * @return string
29 */
30function getID($param='id',$clean=true){
31    /** @var Input $INPUT */
32    global $INPUT;
33    global $conf;
34    global $ACT;
35
36    $id = $INPUT->str($param);
37
38    //construct page id from request URI
39    if(empty($id) && $conf['userewrite'] == 2){
40        $request = $INPUT->server->str('REQUEST_URI');
41        $script = '';
42
43        //get the script URL
44        if($conf['basedir']){
45            $relpath = '';
46            if($param != 'id') {
47                $relpath = 'lib/exe/';
48            }
49            $script = $conf['basedir'] . $relpath .
50                \dokuwiki\Utf8\PhpString::basename($INPUT->server->str('SCRIPT_FILENAME'));
51
52        }elseif($INPUT->server->str('PATH_INFO')){
53            $request = $INPUT->server->str('PATH_INFO');
54        }elseif($INPUT->server->str('SCRIPT_NAME')){
55            $script = $INPUT->server->str('SCRIPT_NAME');
56        }elseif($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')){
57            $script = preg_replace ('/^'.preg_quote($INPUT->server->str('DOCUMENT_ROOT'),'/').'/','',
58                    $INPUT->server->str('SCRIPT_FILENAME'));
59            $script = '/'.$script;
60        }
61
62        //clean script and request (fixes a windows problem)
63        $script  = preg_replace('/\/\/+/','/',$script);
64        $request = preg_replace('/\/\/+/','/',$request);
65
66        //remove script URL and Querystring to gain the id
67        if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
68            $id = preg_replace ('/\?.*/','',$match[1]);
69        }
70        $id = urldecode($id);
71        //strip leading slashes
72        $id = preg_replace('!^/+!','',$id);
73    }
74
75    // Namespace autolinking from URL
76    if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
77        if(page_exists($id.$conf['start'])){
78            // start page inside namespace
79            $id = $id.$conf['start'];
80        }elseif(page_exists($id.noNS(cleanID($id)))){
81            // page named like the NS inside the NS
82            $id = $id.noNS(cleanID($id));
83        }elseif(page_exists($id)){
84            // page like namespace exists
85            $id = substr($id,0,-1);
86        }else{
87            // fall back to default
88            $id = $id.$conf['start'];
89        }
90        if (isset($ACT) && $ACT === 'show') {
91            $urlParameters = $_GET;
92            if (isset($urlParameters['id'])) {
93                unset($urlParameters['id']);
94            }
95            send_redirect(wl($id, $urlParameters, true, '&'));
96        }
97    }
98    if($clean) $id = cleanID($id);
99    if($id === '' && $param=='id') $id = $conf['start'];
100
101    return $id;
102}
103
104/**
105 * Remove unwanted chars from ID
106 *
107 * Cleans a given ID to only use allowed characters. Accented characters are
108 * converted to unaccented ones
109 *
110 * @author Andreas Gohr <andi@splitbrain.org>
111 *
112 * @param  string  $raw_id    The pageid to clean
113 * @param  boolean $ascii     Force ASCII
114 * @return string cleaned id
115 */
116function cleanID($raw_id,$ascii=false){
117    global $conf;
118    static $sepcharpat = null;
119
120    global $cache_cleanid;
121    $cache = & $cache_cleanid;
122
123    // check if it's already in the memory cache
124    if (!$ascii && isset($cache[(string)$raw_id])) {
125        return $cache[(string)$raw_id];
126    }
127
128    $sepchar = $conf['sepchar'];
129    if($sepcharpat == null) // build string only once to save clock cycles
130        $sepcharpat = '#\\'.$sepchar.'+#';
131
132    $id = trim((string)$raw_id);
133    $id = \dokuwiki\Utf8\PhpString::strtolower($id);
134
135    //alternative namespace seperator
136    if($conf['useslash']){
137        $id = strtr($id,';/','::');
138    }else{
139        $id = strtr($id,';/',':'.$sepchar);
140    }
141
142    if($conf['deaccent'] == 2 || $ascii) $id = \dokuwiki\Utf8\Clean::romanize($id);
143    if($conf['deaccent'] || $ascii) $id = \dokuwiki\Utf8\Clean::deaccent($id,-1);
144
145    //remove specials
146    $id = \dokuwiki\Utf8\Clean::stripspecials($id,$sepchar,'\*');
147
148    if($ascii) $id = \dokuwiki\Utf8\Clean::strip($id);
149
150    //clean up
151    $id = preg_replace($sepcharpat,$sepchar,$id);
152    $id = preg_replace('#:+#',':',$id);
153    $id = trim($id,':._-');
154    $id = preg_replace('#:[:\._\-]+#',':',$id);
155    $id = preg_replace('#[:\._\-]+:#',':',$id);
156
157    if (!$ascii) $cache[(string)$raw_id] = $id;
158    return($id);
159}
160
161/**
162 * Return namespacepart of a wiki ID
163 *
164 * @author Andreas Gohr <andi@splitbrain.org>
165 *
166 * @param string $id
167 * @return string|false the namespace part or false if the given ID has no namespace (root)
168 */
169function getNS($id){
170    $pos = strrpos((string)$id,':');
171    if($pos!==false){
172        return substr((string)$id,0,$pos);
173    }
174    return false;
175}
176
177/**
178 * Returns the ID without the namespace
179 *
180 * @author Andreas Gohr <andi@splitbrain.org>
181 *
182 * @param string $id
183 * @return string
184 */
185function noNS($id) {
186    $pos = strrpos($id, ':');
187    if ($pos!==false) {
188        return substr($id, $pos+1);
189    } else {
190        return $id;
191    }
192}
193
194/**
195 * Returns the current namespace
196 *
197 * @author Nathan Fritz <fritzn@crown.edu>
198 *
199 * @param string $id
200 * @return string
201 */
202function curNS($id) {
203    return noNS(getNS($id));
204}
205
206/**
207 * Returns the ID without the namespace or current namespace for 'start' pages
208 *
209 * @author Nathan Fritz <fritzn@crown.edu>
210 *
211 * @param string $id
212 * @return string
213 */
214function noNSorNS($id) {
215    global $conf;
216
217    $p = noNS($id);
218    if ($p === $conf['start'] || $p === false || $p === '') {
219        $p = curNS($id);
220        if ($p === false || $p === '') {
221            return $conf['start'];
222        }
223    }
224    return $p;
225}
226
227/**
228 * Creates a XHTML valid linkid from a given headline title
229 *
230 * @param string  $title   The headline title
231 * @param array|bool   $check   Existing IDs
232 * @return string the title
233 *
234 * @author Andreas Gohr <andi@splitbrain.org>
235 */
236function sectionID($title,&$check) {
237    $title = str_replace(array(':','.'),'',cleanID($title));
238    $new = ltrim($title,'0123456789_-');
239    if(empty($new)){
240        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
241    }else{
242        $title = $new;
243    }
244
245    if(is_array($check)){
246        $suffix=0;
247        $candidateTitle = $title;
248        while(in_array($candidateTitle, $check)){
249          $candidateTitle = $title . ++$suffix;
250        }
251        $check []= $candidateTitle;
252        return $candidateTitle;
253    } else {
254      return $title;
255    }
256}
257
258/**
259 * Wiki page existence check
260 *
261 * parameters as for wikiFN
262 *
263 * @author Chris Smith <chris@jalakai.co.uk>
264 *
265 * @param string $id page id
266 * @param string|int $rev empty or revision timestamp
267 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
268 * @param bool $date_at
269 * @return bool exists?
270 */
271function page_exists($id,$rev='',$clean=true, $date_at=false) {
272    if($rev !== '' && $date_at) {
273        $pagelog = new PageChangeLog($id);
274        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
275        if($pagelog_rev !== false)
276            $rev = $pagelog_rev;
277    }
278    return file_exists(wikiFN($id,$rev,$clean));
279}
280
281/**
282 * Media existence check
283 *
284 * @param string $id page id
285 * @param string|int $rev empty or revision timestamp
286 * @param bool $clean flag indicating that $id should be cleaned (see mediaFN as well)
287 * @param bool $date_at
288 * @return bool exists?
289 */
290function media_exists($id, $rev = '', $clean = true, $date_at = false)
291{
292    if ($rev !== '' && $date_at) {
293        $changeLog = new MediaChangeLog($id);
294        $changelog_rev = $changeLog->getLastRevisionAt($rev);
295        if ($changelog_rev !== false) {
296            $rev = $changelog_rev;
297        }
298    }
299    return file_exists(mediaFN($id, $rev, $clean));
300}
301
302/**
303 * returns the full path to the datafile specified by ID and optional revision
304 *
305 * The filename is URL encoded to protect Unicode chars
306 *
307 * @param  $raw_id  string   id of wikipage
308 * @param  $rev     int|string   page revision, empty string for current
309 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
310 *                           when $id is guaranteed to have been cleaned already.
311 * @return string full path
312 *
313 * @author Andreas Gohr <andi@splitbrain.org>
314 */
315function wikiFN($raw_id,$rev='',$clean=true){
316    global $conf;
317
318    global $cache_wikifn;
319    $cache = & $cache_wikifn;
320
321    $id = $raw_id;
322
323    if ($clean) $id = cleanID($id);
324    $id = str_replace(':','/',$id);
325
326    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
327        return $cache[$id][$rev];
328    }
329
330    if(empty($rev)){
331        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
332    }else{
333        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
334        if($conf['compression']){
335            //test for extensions here, we want to read both compressions
336            if (file_exists($fn . '.gz')){
337                $fn .= '.gz';
338            }else if(file_exists($fn . '.bz2')){
339                $fn .= '.bz2';
340            }else{
341                //file doesnt exist yet, so we take the configured extension
342                $fn .= '.' . $conf['compression'];
343            }
344        }
345    }
346
347    if (!isset($cache[$id])) { $cache[$id] = array(); }
348    $cache[$id][$rev] = $fn;
349    return $fn;
350}
351
352/**
353 * Returns the full path to the file for locking the page while editing.
354 *
355 * @author Ben Coburn <btcoburn@silicodon.net>
356 *
357 * @param string $id page id
358 * @return string full path
359 */
360function wikiLockFN($id) {
361    global $conf;
362    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
363}
364
365
366/**
367 * returns the full path to the meta file specified by ID and extension
368 *
369 * @author Steven Danz <steven-danz@kc.rr.com>
370 *
371 * @param string $id   page id
372 * @param string $ext  file extension
373 * @return string full path
374 */
375function metaFN($id,$ext){
376    global $conf;
377    $id = cleanID($id);
378    $id = str_replace(':','/',$id);
379    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
380    return $fn;
381}
382
383/**
384 * returns the full path to the media's meta file specified by ID and extension
385 *
386 * @author Kate Arzamastseva <pshns@ukr.net>
387 *
388 * @param string $id   media id
389 * @param string $ext  extension of media
390 * @return string
391 */
392function mediaMetaFN($id,$ext){
393    global $conf;
394    $id = cleanID($id);
395    $id = str_replace(':','/',$id);
396    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
397    return $fn;
398}
399
400/**
401 * returns an array of full paths to all metafiles of a given ID
402 *
403 * @author Esther Brunner <esther@kaffeehaus.ch>
404 * @author Michael Hamann <michael@content-space.de>
405 *
406 * @param string $id page id
407 * @return array
408 */
409function metaFiles($id){
410    $basename = metaFN($id, '');
411    $files    = glob($basename.'.*', GLOB_MARK);
412    // filter files like foo.bar.meta when $id == 'foo'
413    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
414}
415
416/**
417 * returns the full path to the mediafile specified by ID
418 *
419 * The filename is URL encoded to protect Unicode chars
420 *
421 * @author Andreas Gohr <andi@splitbrain.org>
422 * @author Kate Arzamastseva <pshns@ukr.net>
423 *
424 * @param string     $id  media id
425 * @param string|int $rev empty string or revision timestamp
426 * @param bool $clean
427 *
428 * @return string full path
429 */
430function mediaFN($id, $rev='', $clean=true){
431    global $conf;
432    if ($clean) $id = cleanID($id);
433    $id = str_replace(':','/',$id);
434    if(empty($rev)){
435        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
436    }else{
437        $ext = mimetype($id);
438        $name = substr($id,0, -1*strlen($ext[0])-1);
439        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
440    }
441    return $fn;
442}
443
444/**
445 * Returns the full filepath to a localized file if local
446 * version isn't found the english one is returned
447 *
448 * @param  string $id  The id of the local file
449 * @param  string $ext The file extension (usually txt)
450 * @return string full filepath to localized file
451 *
452 * @author Andreas Gohr <andi@splitbrain.org>
453 */
454function localeFN($id,$ext='txt'){
455    global $conf;
456    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
457    if(!file_exists($file)){
458        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
459        if(!file_exists($file)){
460            //fall back to english
461            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
462        }
463    }
464    return $file;
465}
466
467/**
468 * Resolve relative paths in IDs
469 *
470 * Do not call directly use resolve_mediaid or resolve_pageid
471 * instead
472 *
473 * Partyly based on a cleanPath function found at
474 * http://php.net/manual/en/function.realpath.php#57016
475 *
476 * @deprecated 2020-09-30
477 * @param string $ns     namespace which is context of id
478 * @param string $id     relative id
479 * @param bool   $clean  flag indicating that id should be cleaned
480 * @return string
481 */
482function resolve_id($ns,$id,$clean=true){
483    global $conf;
484    dbg_deprecated(\dokuwiki\Utils\Resolver::class.' and its children');
485
486    // some pre cleaning for useslash:
487    if($conf['useslash']) $id = str_replace('/',':',$id);
488
489    // if the id starts with a dot we need to handle the
490    // relative stuff
491    if($id && $id[0] == '.'){
492        // normalize initial dots without a colon
493        $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/','\1\3:',$id);
494        // prepend the current namespace
495        $id = $ns.':'.$id;
496
497        // cleanup relatives
498        $result = array();
499        $pathA  = explode(':', $id);
500        if (!$pathA[0]) $result[] = '';
501        foreach ($pathA AS $key => $dir) {
502            if ($dir == '..') {
503                if (end($result) == '..') {
504                    $result[] = '..';
505                } elseif (!array_pop($result)) {
506                    $result[] = '..';
507                }
508            } elseif ($dir && $dir != '.') {
509                $result[] = $dir;
510            }
511        }
512        if (!end($pathA)) $result[] = '';
513        $id = implode(':', $result);
514    }elseif($ns !== false && strpos($id,':') === false){
515        //if link contains no namespace. add current namespace (if any)
516        $id = $ns.':'.$id;
517    }
518
519    if($clean) $id = cleanID($id);
520    return $id;
521}
522
523/**
524 * Returns a full media id
525 *
526 * @param string $ns namespace which is context of id
527 * @param string &$media (reference) relative media id, updated to resolved id
528 * @param bool &$exists (reference) updated with existance of media
529 * @param int|string $rev
530 * @param bool $date_at
531 * @deprecated 2020-09-30
532 */
533function resolve_mediaid($ns,&$media,&$exists,$rev='',$date_at=false){
534    dbg_deprecated(MediaResolver::class);
535    $resolver = new MediaResolver("$ns:deprecated");
536    $media = $resolver->resolveId($media, $rev, $date_at);
537    $exists = media_exists($media, $rev, false, $date_at);
538}
539
540/**
541 * Returns a full page id
542 *
543 * @deprecated 2020-09-30
544 * @param string $ns namespace which is context of id
545 * @param string &$page (reference) relative page id, updated to resolved id
546 * @param bool &$exists (reference) updated with existance of media
547 * @param string $rev
548 * @param bool $date_at
549 */
550function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false )
551{
552    dbg_deprecated(PageResolver::class);
553    $resolver = new PageResolver("$ns:deprecated");
554    $page = $resolver->resolveId($page, $rev, $date_at);
555    $exists = page_exists($page, $rev, false, $date_at);
556}
557
558/**
559 * Returns the name of a cachefile from given data
560 *
561 * The needed directory is created by this function!
562 *
563 * @author Andreas Gohr <andi@splitbrain.org>
564 *
565 * @param string $data  This data is used to create a unique md5 name
566 * @param string $ext   This is appended to the filename if given
567 * @return string       The filename of the cachefile
568 */
569function getCacheName($data,$ext=''){
570    global $conf;
571    $md5  = md5($data);
572    $file = $conf['cachedir'].'/'.$md5[0].'/'.$md5.$ext;
573    io_makeFileDir($file);
574    return $file;
575}
576
577/**
578 * Checks a pageid against $conf['hidepages']
579 *
580 * @author Andreas Gohr <gohr@cosmocode.de>
581 *
582 * @param string $id page id
583 * @return bool
584 */
585function isHiddenPage($id){
586    $data = array(
587        'id' => $id,
588        'hidden' => false
589    );
590    \dokuwiki\Extension\Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
591    return $data['hidden'];
592}
593
594/**
595 * callback checks if page is hidden
596 *
597 * @param array $data event data    - see isHiddenPage()
598 */
599function _isHiddenPage(&$data) {
600    global $conf;
601    global $ACT;
602
603    if ($data['hidden']) return;
604    if(empty($conf['hidepages'])) return;
605    if($ACT == 'admin') return;
606
607    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
608        $data['hidden'] = true;
609    }
610}
611
612/**
613 * Reverse of isHiddenPage
614 *
615 * @author Andreas Gohr <gohr@cosmocode.de>
616 *
617 * @param string $id page id
618 * @return bool
619 */
620function isVisiblePage($id){
621    return !isHiddenPage($id);
622}
623
624/**
625 * Format an id for output to a user
626 *
627 * Namespaces are denoted by a trailing “:*”. The root namespace is
628 * “*”. Output is escaped.
629 *
630 * @author Adrian Lang <lang@cosmocode.de>
631 *
632 * @param string $id page id
633 * @return string
634 */
635function prettyprint_id($id) {
636    if (!$id || $id === ':') {
637        return '*';
638    }
639    if ((substr($id, -1, 1) === ':')) {
640        $id .= '*';
641    }
642    return hsc($id);
643}
644
645/**
646 * Encode a UTF-8 filename to use on any filesystem
647 *
648 * Uses the 'fnencode' option to determine encoding
649 *
650 * When the second parameter is true the string will
651 * be encoded only if non ASCII characters are detected -
652 * This makes it safe to run it multiple times on the
653 * same string (default is true)
654 *
655 * @author Andreas Gohr <andi@splitbrain.org>
656 * @see    urlencode
657 *
658 * @param string $file file name
659 * @param bool   $safe if true, only encoded when non ASCII characters detected
660 * @return string
661 */
662function utf8_encodeFN($file,$safe=true){
663    global $conf;
664    if($conf['fnencode'] == 'utf-8') return $file;
665
666    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
667        return $file;
668    }
669
670    if($conf['fnencode'] == 'safe'){
671        return SafeFN::encode($file);
672    }
673
674    $file = urlencode($file);
675    $file = str_replace('%2F','/',$file);
676    return $file;
677}
678
679/**
680 * Decode a filename back to UTF-8
681 *
682 * Uses the 'fnencode' option to determine encoding
683 *
684 * @author Andreas Gohr <andi@splitbrain.org>
685 * @see    urldecode
686 *
687 * @param string $file file name
688 * @return string
689 */
690function utf8_decodeFN($file){
691    global $conf;
692    if($conf['fnencode'] == 'utf-8') return $file;
693
694    if($conf['fnencode'] == 'safe'){
695        return SafeFN::decode($file);
696    }
697
698    return urldecode($file);
699}
700
701/**
702 * Find a page in the current namespace (determined from $ID) or any
703 * higher namespace that can be accessed by the current user,
704 * this condition can be overriden by an optional parameter.
705 *
706 * Used for sidebars, but can be used other stuff as well
707 *
708 * @todo   add event hook
709 *
710 * @param  string $page the pagename you're looking for
711 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
712 * @return false|string the full page id of the found page, false if any
713 */
714function page_findnearest($page, $useacl = true){
715    if ((string) $page === '') return false;
716    global $ID;
717
718    $ns = $ID;
719    do {
720        $ns = getNS($ns);
721        $pageid = cleanID("$ns:$page");
722        if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
723            return $pageid;
724        }
725    } while($ns !== false);
726
727    return false;
728}
729