xref: /dokuwiki/inc/pageutils.php (revision 8c6be208dcb4734cf5ad7e62ab5df1d42d46ad76) 
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   $check   Existing IDs (title => number)
232 * @return string the title
233 *
234 * @author Andreas Gohr <andi@splitbrain.org>
235 */
236function sectionID($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        // make sure tiles are unique
247        if (!array_key_exists ($title,$check)) {
248            $check[$title] = 0;
249        } else {
250            $title .= ++ $check[$title];
251        }
252    }
253
254    return $title;
255}
256
257/**
258 * Wiki page existence check
259 *
260 * parameters as for wikiFN
261 *
262 * @author Chris Smith <chris@jalakai.co.uk>
263 *
264 * @param string $id page id
265 * @param string|int $rev empty or revision timestamp
266 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
267 * @param bool $date_at
268 * @return bool exists?
269 */
270function page_exists($id,$rev='',$clean=true, $date_at=false) {
271    if($rev !== '' && $date_at) {
272        $pagelog = new PageChangeLog($id);
273        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
274        if($pagelog_rev !== false)
275            $rev = $pagelog_rev;
276    }
277    return file_exists(wikiFN($id,$rev,$clean));
278}
279
280/**
281 * Media existence check
282 *
283 * @param string $id page id
284 * @param string|int $rev empty or revision timestamp
285 * @param bool $clean flag indicating that $id should be cleaned (see mediaFN as well)
286 * @param bool $date_at
287 * @return bool exists?
288 */
289function media_exists($id, $rev = '', $clean = true, $date_at = false)
290{
291    if ($rev !== '' && $date_at) {
292        $changeLog = new MediaChangeLog($id);
293        $changelog_rev = $changeLog->getLastRevisionAt($rev);
294        if ($changelog_rev !== false) {
295            $rev = $changelog_rev;
296        }
297    }
298    return file_exists(mediaFN($id, $rev, $clean));
299}
300
301/**
302 * returns the full path to the datafile specified by ID and optional revision
303 *
304 * The filename is URL encoded to protect Unicode chars
305 *
306 * @param  $raw_id  string   id of wikipage
307 * @param  $rev     int|string   page revision, empty string for current
308 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
309 *                           when $id is guaranteed to have been cleaned already.
310 * @return string full path
311 *
312 * @author Andreas Gohr <andi@splitbrain.org>
313 */
314function wikiFN($raw_id,$rev='',$clean=true){
315    global $conf;
316
317    global $cache_wikifn;
318    $cache = & $cache_wikifn;
319
320    $id = $raw_id;
321
322    if ($clean) $id = cleanID($id);
323    $id = str_replace(':','/',$id);
324
325    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
326        return $cache[$id][$rev];
327    }
328
329    if(empty($rev)){
330        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
331    }else{
332        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
333        if($conf['compression']){
334            //test for extensions here, we want to read both compressions
335            if (file_exists($fn . '.gz')){
336                $fn .= '.gz';
337            }else if(file_exists($fn . '.bz2')){
338                $fn .= '.bz2';
339            }else{
340                //file doesnt exist yet, so we take the configured extension
341                $fn .= '.' . $conf['compression'];
342            }
343        }
344    }
345
346    if (!isset($cache[$id])) { $cache[$id] = array(); }
347    $cache[$id][$rev] = $fn;
348    return $fn;
349}
350
351/**
352 * Returns the full path to the file for locking the page while editing.
353 *
354 * @author Ben Coburn <btcoburn@silicodon.net>
355 *
356 * @param string $id page id
357 * @return string full path
358 */
359function wikiLockFN($id) {
360    global $conf;
361    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
362}
363
364
365/**
366 * returns the full path to the meta file specified by ID and extension
367 *
368 * @author Steven Danz <steven-danz@kc.rr.com>
369 *
370 * @param string $id   page id
371 * @param string $ext  file extension
372 * @return string full path
373 */
374function metaFN($id,$ext){
375    global $conf;
376    $id = cleanID($id);
377    $id = str_replace(':','/',$id);
378    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
379    return $fn;
380}
381
382/**
383 * returns the full path to the media's meta file specified by ID and extension
384 *
385 * @author Kate Arzamastseva <pshns@ukr.net>
386 *
387 * @param string $id   media id
388 * @param string $ext  extension of media
389 * @return string
390 */
391function mediaMetaFN($id,$ext){
392    global $conf;
393    $id = cleanID($id);
394    $id = str_replace(':','/',$id);
395    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
396    return $fn;
397}
398
399/**
400 * returns an array of full paths to all metafiles of a given ID
401 *
402 * @author Esther Brunner <esther@kaffeehaus.ch>
403 * @author Michael Hamann <michael@content-space.de>
404 *
405 * @param string $id page id
406 * @return array
407 */
408function metaFiles($id){
409    $basename = metaFN($id, '');
410    $files    = glob($basename.'.*', GLOB_MARK);
411    // filter files like foo.bar.meta when $id == 'foo'
412    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
413}
414
415/**
416 * returns the full path to the mediafile specified by ID
417 *
418 * The filename is URL encoded to protect Unicode chars
419 *
420 * @author Andreas Gohr <andi@splitbrain.org>
421 * @author Kate Arzamastseva <pshns@ukr.net>
422 *
423 * @param string     $id  media id
424 * @param string|int $rev empty string or revision timestamp
425 * @param bool $clean
426 *
427 * @return string full path
428 */
429function mediaFN($id, $rev='', $clean=true){
430    global $conf;
431    if ($clean) $id = cleanID($id);
432    $id = str_replace(':','/',$id);
433    if(empty($rev)){
434        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
435    }else{
436        $ext = mimetype($id);
437        $name = substr($id,0, -1*strlen($ext[0])-1);
438        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
439    }
440    return $fn;
441}
442
443/**
444 * Returns the full filepath to a localized file if local
445 * version isn't found the english one is returned
446 *
447 * @param  string $id  The id of the local file
448 * @param  string $ext The file extension (usually txt)
449 * @return string full filepath to localized file
450 *
451 * @author Andreas Gohr <andi@splitbrain.org>
452 */
453function localeFN($id,$ext='txt'){
454    global $conf;
455    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
456    if(!file_exists($file)){
457        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
458        if(!file_exists($file)){
459            //fall back to english
460            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
461        }
462    }
463    return $file;
464}
465
466/**
467 * Resolve relative paths in IDs
468 *
469 * Do not call directly use resolve_mediaid or resolve_pageid
470 * instead
471 *
472 * Partyly based on a cleanPath function found at
473 * http://php.net/manual/en/function.realpath.php#57016
474 *
475 * @deprecated 2020-09-30
476 * @param string $ns     namespace which is context of id
477 * @param string $id     relative id
478 * @param bool   $clean  flag indicating that id should be cleaned
479 * @return string
480 */
481function resolve_id($ns,$id,$clean=true){
482    global $conf;
483    dbg_deprecated(\dokuwiki\Utils\Resolver::class.' and its children');
484
485    // some pre cleaning for useslash:
486    if($conf['useslash']) $id = str_replace('/',':',$id);
487
488    // if the id starts with a dot we need to handle the
489    // relative stuff
490    if($id && $id[0] == '.'){
491        // normalize initial dots without a colon
492        $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/','\1\3:',$id);
493        // prepend the current namespace
494        $id = $ns.':'.$id;
495
496        // cleanup relatives
497        $result = array();
498        $pathA  = explode(':', $id);
499        if (!$pathA[0]) $result[] = '';
500        foreach ($pathA AS $key => $dir) {
501            if ($dir == '..') {
502                if (end($result) == '..') {
503                    $result[] = '..';
504                } elseif (!array_pop($result)) {
505                    $result[] = '..';
506                }
507            } elseif ($dir && $dir != '.') {
508                $result[] = $dir;
509            }
510        }
511        if (!end($pathA)) $result[] = '';
512        $id = implode(':', $result);
513    }elseif($ns !== false && strpos($id,':') === false){
514        //if link contains no namespace. add current namespace (if any)
515        $id = $ns.':'.$id;
516    }
517
518    if($clean) $id = cleanID($id);
519    return $id;
520}
521
522/**
523 * Returns a full media id
524 *
525 * @param string $ns namespace which is context of id
526 * @param string &$media (reference) relative media id, updated to resolved id
527 * @param bool &$exists (reference) updated with existance of media
528 * @param int|string $rev
529 * @param bool $date_at
530 * @deprecated 2020-09-30
531 */
532function resolve_mediaid($ns,&$media,&$exists,$rev='',$date_at=false){
533    dbg_deprecated(MediaResolver::class);
534    $resolver = new MediaResolver("$ns:deprecated");
535    $media = $resolver->resolveId($media, $rev, $date_at);
536    $exists = media_exists($media, $rev, false, $date_at);
537}
538
539/**
540 * Returns a full page id
541 *
542 * @deprecated 2020-09-30
543 * @param string $ns namespace which is context of id
544 * @param string &$page (reference) relative page id, updated to resolved id
545 * @param bool &$exists (reference) updated with existance of media
546 * @param string $rev
547 * @param bool $date_at
548 */
549function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false )
550{
551    dbg_deprecated(PageResolver::class);
552    $resolver = new PageResolver("$ns:deprecated");
553    $page = $resolver->resolveId($page, $rev, $date_at);
554    $exists = page_exists($page, $rev, false, $date_at);
555}
556
557/**
558 * Returns the name of a cachefile from given data
559 *
560 * The needed directory is created by this function!
561 *
562 * @author Andreas Gohr <andi@splitbrain.org>
563 *
564 * @param string $data  This data is used to create a unique md5 name
565 * @param string $ext   This is appended to the filename if given
566 * @return string       The filename of the cachefile
567 */
568function getCacheName($data,$ext=''){
569    global $conf;
570    $md5  = md5($data);
571    $file = $conf['cachedir'].'/'.$md5[0].'/'.$md5.$ext;
572    io_makeFileDir($file);
573    return $file;
574}
575
576/**
577 * Checks a pageid against $conf['hidepages']
578 *
579 * @author Andreas Gohr <gohr@cosmocode.de>
580 *
581 * @param string $id page id
582 * @return bool
583 */
584function isHiddenPage($id){
585    $data = array(
586        'id' => $id,
587        'hidden' => false
588    );
589    \dokuwiki\Extension\Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
590    return $data['hidden'];
591}
592
593/**
594 * callback checks if page is hidden
595 *
596 * @param array $data event data    - see isHiddenPage()
597 */
598function _isHiddenPage(&$data) {
599    global $conf;
600    global $ACT;
601
602    if ($data['hidden']) return;
603    if(empty($conf['hidepages'])) return;
604    if($ACT == 'admin') return;
605
606    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
607        $data['hidden'] = true;
608    }
609}
610
611/**
612 * Reverse of isHiddenPage
613 *
614 * @author Andreas Gohr <gohr@cosmocode.de>
615 *
616 * @param string $id page id
617 * @return bool
618 */
619function isVisiblePage($id){
620    return !isHiddenPage($id);
621}
622
623/**
624 * Format an id for output to a user
625 *
626 * Namespaces are denoted by a trailing “:*”. The root namespace is
627 * “*”. Output is escaped.
628 *
629 * @author Adrian Lang <lang@cosmocode.de>
630 *
631 * @param string $id page id
632 * @return string
633 */
634function prettyprint_id($id) {
635    if (!$id || $id === ':') {
636        return '*';
637    }
638    if ((substr($id, -1, 1) === ':')) {
639        $id .= '*';
640    }
641    return hsc($id);
642}
643
644/**
645 * Encode a UTF-8 filename to use on any filesystem
646 *
647 * Uses the 'fnencode' option to determine encoding
648 *
649 * When the second parameter is true the string will
650 * be encoded only if non ASCII characters are detected -
651 * This makes it safe to run it multiple times on the
652 * same string (default is true)
653 *
654 * @author Andreas Gohr <andi@splitbrain.org>
655 * @see    urlencode
656 *
657 * @param string $file file name
658 * @param bool   $safe if true, only encoded when non ASCII characters detected
659 * @return string
660 */
661function utf8_encodeFN($file,$safe=true){
662    global $conf;
663    if($conf['fnencode'] == 'utf-8') return $file;
664
665    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
666        return $file;
667    }
668
669    if($conf['fnencode'] == 'safe'){
670        return SafeFN::encode($file);
671    }
672
673    $file = urlencode($file);
674    $file = str_replace('%2F','/',$file);
675    return $file;
676}
677
678/**
679 * Decode a filename back to UTF-8
680 *
681 * Uses the 'fnencode' option to determine encoding
682 *
683 * @author Andreas Gohr <andi@splitbrain.org>
684 * @see    urldecode
685 *
686 * @param string $file file name
687 * @return string
688 */
689function utf8_decodeFN($file){
690    global $conf;
691    if($conf['fnencode'] == 'utf-8') return $file;
692
693    if($conf['fnencode'] == 'safe'){
694        return SafeFN::decode($file);
695    }
696
697    return urldecode($file);
698}
699
700/**
701 * Find a page in the current namespace (determined from $ID) or any
702 * higher namespace that can be accessed by the current user,
703 * this condition can be overriden by an optional parameter.
704 *
705 * Used for sidebars, but can be used other stuff as well
706 *
707 * @todo   add event hook
708 *
709 * @param  string $page the pagename you're looking for
710 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
711 * @return false|string the full page id of the found page, false if any
712 */
713function page_findnearest($page, $useacl = true){
714    if ((string) $page === '') return false;
715    global $ID;
716
717    $ns = $ID;
718    do {
719        $ns = getNS($ns);
720        $pageid = cleanID("$ns:$page");
721        if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
722            return $pageid;
723        }
724    } while($ns !== false);
725
726    return false;
727}
728