xref: /dokuwiki/inc/pageutils.php (revision b9a4556d52a93cf82d1e99f49e49363a1b879906) 
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;
12
13/**
14 * Fetch the an ID from request
15 *
16 * Uses either standard $_REQUEST variable or extracts it from
17 * the full request URI when userewrite is set to 2
18 *
19 * For $param='id' $conf['start'] is returned if no id was found.
20 * If the second parameter is true (default) the ID is cleaned.
21 *
22 * @author Andreas Gohr <andi@splitbrain.org>
23 *
24 * @param string $param  the $_REQUEST variable name, default 'id'
25 * @param bool   $clean  if true, ID is cleaned
26 * @return string
27 */
28function getID($param='id',$clean=true){
29    /** @var Input $INPUT */
30    global $INPUT;
31    global $conf;
32    global $ACT;
33
34    $id = $INPUT->str($param);
35
36    //construct page id from request URI
37    if(empty($id) && $conf['userewrite'] == 2){
38        $request = $INPUT->server->str('REQUEST_URI');
39        $script = '';
40
41        //get the script URL
42        if($conf['basedir']){
43            $relpath = '';
44            if($param != 'id') {
45                $relpath = 'lib/exe/';
46            }
47            $script = $conf['basedir'] . $relpath .
48                \dokuwiki\Utf8\PhpString::basename($INPUT->server->str('SCRIPT_FILENAME'));
49
50        }elseif($INPUT->server->str('PATH_INFO')){
51            $request = $INPUT->server->str('PATH_INFO');
52        }elseif($INPUT->server->str('SCRIPT_NAME')){
53            $script = $INPUT->server->str('SCRIPT_NAME');
54        }elseif($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')){
55            $script = preg_replace ('/^'.preg_quote($INPUT->server->str('DOCUMENT_ROOT'),'/').'/','',
56                    $INPUT->server->str('SCRIPT_FILENAME'));
57            $script = '/'.$script;
58        }
59
60        //clean script and request (fixes a windows problem)
61        $script  = preg_replace('/\/\/+/','/',$script);
62        $request = preg_replace('/\/\/+/','/',$request);
63
64        //remove script URL and Querystring to gain the id
65        if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
66            $id = preg_replace ('/\?.*/','',$match[1]);
67        }
68        $id = urldecode($id);
69        //strip leading slashes
70        $id = preg_replace('!^/+!','',$id);
71    }
72
73    // Namespace autolinking from URL
74    if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
75        if(page_exists($id.$conf['start'])){
76            // start page inside namespace
77            $id = $id.$conf['start'];
78        }elseif(page_exists($id.noNS(cleanID($id)))){
79            // page named like the NS inside the NS
80            $id = $id.noNS(cleanID($id));
81        }elseif(page_exists($id)){
82            // page like namespace exists
83            $id = substr($id,0,-1);
84        }else{
85            // fall back to default
86            $id = $id.$conf['start'];
87        }
88        if (isset($ACT) && $ACT === 'show') {
89            $urlParameters = $_GET;
90            if (isset($urlParameters['id'])) {
91                unset($urlParameters['id']);
92            }
93            send_redirect(wl($id, $urlParameters, true, '&'));
94        }
95    }
96    if($clean) $id = cleanID($id);
97    if($id === '' && $param=='id') $id = $conf['start'];
98
99    return $id;
100}
101
102/**
103 * Remove unwanted chars from ID
104 *
105 * Cleans a given ID to only use allowed characters. Accented characters are
106 * converted to unaccented ones
107 *
108 * @author Andreas Gohr <andi@splitbrain.org>
109 *
110 * @param  string  $raw_id    The pageid to clean
111 * @param  boolean $ascii     Force ASCII
112 * @return string cleaned id
113 */
114function cleanID($raw_id,$ascii=false){
115    global $conf;
116    static $sepcharpat = null;
117
118    global $cache_cleanid;
119    $cache = & $cache_cleanid;
120
121    // check if it's already in the memory cache
122    if (!$ascii && isset($cache[(string)$raw_id])) {
123        return $cache[(string)$raw_id];
124    }
125
126    $sepchar = $conf['sepchar'];
127    if($sepcharpat == null) // build string only once to save clock cycles
128        $sepcharpat = '#\\'.$sepchar.'+#';
129
130    $id = trim((string)$raw_id);
131    $id = \dokuwiki\Utf8\PhpString::strtolower($id);
132
133    //alternative namespace seperator
134    if($conf['useslash']){
135        $id = strtr($id,';/','::');
136    }else{
137        $id = strtr($id,';/',':'.$sepchar);
138    }
139
140    if($conf['deaccent'] == 2 || $ascii) $id = \dokuwiki\Utf8\Clean::romanize($id);
141    if($conf['deaccent'] || $ascii) $id = \dokuwiki\Utf8\Clean::deaccent($id,-1);
142
143    //remove specials
144    $id = \dokuwiki\Utf8\Clean::stripspecials($id,$sepchar,'\*');
145
146    if($ascii) $id = \dokuwiki\Utf8\Clean::strip($id);
147
148    //clean up
149    $id = preg_replace($sepcharpat,$sepchar,$id);
150    $id = preg_replace('#:+#',':',$id);
151    $id = trim($id,':._-');
152    $id = preg_replace('#:[:\._\-]+#',':',$id);
153    $id = preg_replace('#[:\._\-]+:#',':',$id);
154
155    if (!$ascii) $cache[(string)$raw_id] = $id;
156    return($id);
157}
158
159/**
160 * Return namespacepart of a wiki ID
161 *
162 * @author Andreas Gohr <andi@splitbrain.org>
163 *
164 * @param string $id
165 * @return string|false the namespace part or false if the given ID has no namespace (root)
166 */
167function getNS($id){
168    $pos = strrpos((string)$id,':');
169    if($pos!==false){
170        return substr((string)$id,0,$pos);
171    }
172    return false;
173}
174
175/**
176 * Returns the ID without the namespace
177 *
178 * @author Andreas Gohr <andi@splitbrain.org>
179 *
180 * @param string $id
181 * @return string
182 */
183function noNS($id) {
184    $pos = strrpos($id, ':');
185    if ($pos!==false) {
186        return substr($id, $pos+1);
187    } else {
188        return $id;
189    }
190}
191
192/**
193 * Returns the current namespace
194 *
195 * @author Nathan Fritz <fritzn@crown.edu>
196 *
197 * @param string $id
198 * @return string
199 */
200function curNS($id) {
201    return noNS(getNS($id));
202}
203
204/**
205 * Returns the ID without the namespace or current namespace for 'start' pages
206 *
207 * @author Nathan Fritz <fritzn@crown.edu>
208 *
209 * @param string $id
210 * @return string
211 */
212function noNSorNS($id) {
213    global $conf;
214
215    $p = noNS($id);
216    if ($p === $conf['start'] || $p === false || $p === '') {
217        $p = curNS($id);
218        if ($p === false || $p === '') {
219            return $conf['start'];
220        }
221    }
222    return $p;
223}
224
225/**
226 * Creates a XHTML valid linkid from a given headline title
227 *
228 * @param string  $title   The headline title
229 * @param array|bool   $check   Existing IDs
230 * @return string the title
231 *
232 * @author Andreas Gohr <andi@splitbrain.org>
233 */
234function sectionID($title,&$check) {
235    $title = str_replace(array(':','.'),'',cleanID($title));
236    $new = ltrim($title,'0123456789_-');
237    if(empty($new)){
238        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
239    }else{
240        $title = $new;
241    }
242
243    if(is_array($check)){
244        $suffix=0;
245        $candidateTitle = $title;
246        while(in_array($candidateTitle, $check)){
247          $candidateTitle = $title . ++$suffix;
248        }
249        $check []= $candidateTitle;
250        return $candidateTitle;
251    } else {
252      return $title;
253    }
254}
255
256/**
257 * Wiki page existence check
258 *
259 * parameters as for wikiFN
260 *
261 * @author Chris Smith <chris@jalakai.co.uk>
262 *
263 * @param string $id page id
264 * @param string|int $rev empty or revision timestamp
265 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
266 * @param bool $date_at
267 * @return bool exists?
268 */
269function page_exists($id,$rev='',$clean=true, $date_at=false) {
270    if($rev !== '' && $date_at) {
271        $pagelog = new PageChangeLog($id);
272        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
273        if($pagelog_rev !== false)
274            $rev = $pagelog_rev;
275    }
276    return file_exists(wikiFN($id,$rev,$clean));
277}
278
279/**
280 * returns the full path to the datafile specified by ID and optional revision
281 *
282 * The filename is URL encoded to protect Unicode chars
283 *
284 * @param  $raw_id  string   id of wikipage
285 * @param  $rev     int|string   page revision, empty string for current
286 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
287 *                           when $id is guaranteed to have been cleaned already.
288 * @return string full path
289 *
290 * @author Andreas Gohr <andi@splitbrain.org>
291 */
292function wikiFN($raw_id,$rev='',$clean=true){
293    global $conf;
294
295    global $cache_wikifn;
296    $cache = & $cache_wikifn;
297
298    $id = $raw_id;
299
300    if ($clean) $id = cleanID($id);
301    $id = str_replace(':','/',$id);
302
303    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
304        return $cache[$id][$rev];
305    }
306
307    if(empty($rev)){
308        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
309    }else{
310        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
311        if($conf['compression']){
312            //test for extensions here, we want to read both compressions
313            if (file_exists($fn . '.gz')){
314                $fn .= '.gz';
315            }else if(file_exists($fn . '.bz2')){
316                $fn .= '.bz2';
317            }else{
318                //file doesnt exist yet, so we take the configured extension
319                $fn .= '.' . $conf['compression'];
320            }
321        }
322    }
323
324    if (!isset($cache[$id])) { $cache[$id] = array(); }
325    $cache[$id][$rev] = $fn;
326    return $fn;
327}
328
329/**
330 * Returns the full path to the file for locking the page while editing.
331 *
332 * @author Ben Coburn <btcoburn@silicodon.net>
333 *
334 * @param string $id page id
335 * @return string full path
336 */
337function wikiLockFN($id) {
338    global $conf;
339    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
340}
341
342
343/**
344 * returns the full path to the meta file specified by ID and extension
345 *
346 * @author Steven Danz <steven-danz@kc.rr.com>
347 *
348 * @param string $id   page id
349 * @param string $ext  file extension
350 * @return string full path
351 */
352function metaFN($id,$ext){
353    global $conf;
354    $id = cleanID($id);
355    $id = str_replace(':','/',$id);
356    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
357    return $fn;
358}
359
360/**
361 * returns the full path to the media's meta file specified by ID and extension
362 *
363 * @author Kate Arzamastseva <pshns@ukr.net>
364 *
365 * @param string $id   media id
366 * @param string $ext  extension of media
367 * @return string
368 */
369function mediaMetaFN($id,$ext){
370    global $conf;
371    $id = cleanID($id);
372    $id = str_replace(':','/',$id);
373    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
374    return $fn;
375}
376
377/**
378 * returns an array of full paths to all metafiles of a given ID
379 *
380 * @author Esther Brunner <esther@kaffeehaus.ch>
381 * @author Michael Hamann <michael@content-space.de>
382 *
383 * @param string $id page id
384 * @return array
385 */
386function metaFiles($id){
387    $basename = metaFN($id, '');
388    $files    = glob($basename.'.*', GLOB_MARK);
389    // filter files like foo.bar.meta when $id == 'foo'
390    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
391}
392
393/**
394 * returns the full path to the mediafile specified by ID
395 *
396 * The filename is URL encoded to protect Unicode chars
397 *
398 * @author Andreas Gohr <andi@splitbrain.org>
399 * @author Kate Arzamastseva <pshns@ukr.net>
400 *
401 * @param string     $id  media id
402 * @param string|int $rev empty string or revision timestamp
403 * @param bool $clean
404 *
405 * @return string full path
406 */
407function mediaFN($id, $rev='', $clean=true){
408    global $conf;
409    if ($clean) $id = cleanID($id);
410    $id = str_replace(':','/',$id);
411    if(empty($rev)){
412        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
413    }else{
414        $ext = mimetype($id);
415        $name = substr($id,0, -1*strlen($ext[0])-1);
416        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
417    }
418    return $fn;
419}
420
421/**
422 * Returns the full filepath to a localized file if local
423 * version isn't found the english one is returned
424 *
425 * @param  string $id  The id of the local file
426 * @param  string $ext The file extension (usually txt)
427 * @return string full filepath to localized file
428 *
429 * @author Andreas Gohr <andi@splitbrain.org>
430 */
431function localeFN($id,$ext='txt'){
432    global $conf;
433    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
434    if(!file_exists($file)){
435        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
436        if(!file_exists($file)){
437            //fall back to english
438            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
439        }
440    }
441    return $file;
442}
443
444/**
445 * Resolve relative paths in IDs
446 *
447 * Do not call directly use resolve_mediaid or resolve_pageid
448 * instead
449 *
450 * Partyly based on a cleanPath function found at
451 * http://php.net/manual/en/function.realpath.php#57016
452 *
453 * @author <bart at mediawave dot nl>
454 *
455 * @param string $ns     namespace which is context of id
456 * @param string $id     relative id
457 * @param bool   $clean  flag indicating that id should be cleaned
458 * @return string
459 */
460function resolve_id($ns,$id,$clean=true){
461    global $conf;
462
463    // some pre cleaning for useslash:
464    if($conf['useslash']) $id = str_replace('/',':',$id);
465
466    // if the id starts with a dot we need to handle the
467    // relative stuff
468    if($id && $id[0] == '.'){
469        // normalize initial dots without a colon
470        $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/','\1\3:',$id);
471        // prepend the current namespace
472        $id = $ns.':'.$id;
473
474        // cleanup relatives
475        $result = array();
476        $pathA  = explode(':', $id);
477        if (!$pathA[0]) $result[] = '';
478        foreach ($pathA AS $key => $dir) {
479            if ($dir == '..') {
480                if (end($result) == '..') {
481                    $result[] = '..';
482                } elseif (!array_pop($result)) {
483                    $result[] = '..';
484                }
485            } elseif ($dir && $dir != '.') {
486                $result[] = $dir;
487            }
488        }
489        if (!end($pathA)) $result[] = '';
490        $id = implode(':', $result);
491    }elseif($ns !== false && strpos($id,':') === false){
492        //if link contains no namespace. add current namespace (if any)
493        $id = $ns.':'.$id;
494    }
495
496    if($clean) $id = cleanID($id);
497    return $id;
498}
499
500/**
501 * Returns a full media id
502 *
503 * @author Andreas Gohr <andi@splitbrain.org>
504 *
505 * @param string $ns namespace which is context of id
506 * @param string &$page (reference) relative media id, updated to resolved id
507 * @param bool &$exists (reference) updated with existance of media
508 * @param int|string $rev
509 * @param bool $date_at
510 */
511function resolve_mediaid($ns,&$page,&$exists,$rev='',$date_at=false){
512    $page   = resolve_id($ns,$page);
513    if($rev !== '' &&  $date_at){
514        $medialog = new MediaChangeLog($page);
515        $medialog_rev = $medialog->getLastRevisionAt($rev);
516        if($medialog_rev !== false) {
517            $rev = $medialog_rev;
518        }
519    }
520
521    $file   = mediaFN($page,$rev);
522    $exists = file_exists($file);
523}
524
525/**
526 * Returns a full page id
527 *
528 * @author Andreas Gohr <andi@splitbrain.org>
529 *
530 * @param string $ns namespace which is context of id
531 * @param string &$page (reference) relative page id, updated to resolved id
532 * @param bool &$exists (reference) updated with existance of media
533 * @param string $rev
534 * @param bool $date_at
535 */
536function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false ){
537    global $conf;
538    global $ID;
539    $exists = false;
540
541    //empty address should point to current page
542    if ($page === "") {
543        $page = $ID;
544    }
545
546    //keep hashlink if exists then clean both parts
547    if (strpos($page,'#')) {
548        list($page,$hash) = explode('#',$page,2);
549    } else {
550        $hash = '';
551    }
552    $hash = cleanID($hash);
553    $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
554
555    // get filename (calls clean itself)
556    if($rev !== '' && $date_at) {
557        $pagelog = new PageChangeLog($page);
558        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
559        if($pagelog_rev !== false)//something found
560           $rev  = $pagelog_rev;
561    }
562    $file = wikiFN($page,$rev);
563
564    // if ends with colon or slash we have a namespace link
565    if(in_array(substr($page,-1), array(':', ';')) ||
566       ($conf['useslash'] && substr($page,-1) == '/')){
567        if(page_exists($page.$conf['start'],$rev,true,$date_at)){
568            // start page inside namespace
569            $page = $page.$conf['start'];
570            $exists = true;
571        }elseif(page_exists($page.noNS(cleanID($page)),$rev,true,$date_at)){
572            // page named like the NS inside the NS
573            $page = $page.noNS(cleanID($page));
574            $exists = true;
575        }elseif(page_exists($page,$rev,true,$date_at)){
576            // page like namespace exists
577            $page = $page;
578            $exists = true;
579        }else{
580            // fall back to default
581            $page = $page.$conf['start'];
582        }
583    }else{
584        //check alternative plural/nonplural form
585        if(!file_exists($file)){
586            if( $conf['autoplural'] ){
587                if(substr($page,-1) == 's'){
588                    $try = substr($page,0,-1);
589                }else{
590                    $try = $page.'s';
591                }
592                if(page_exists($try,$rev,true,$date_at)){
593                    $page   = $try;
594                    $exists = true;
595                }
596            }
597        }else{
598            $exists = true;
599        }
600    }
601
602    // now make sure we have a clean page
603    $page = cleanID($page);
604
605    //add hash if any
606    if(!empty($hash)) $page .= '#'.$hash;
607}
608
609/**
610 * Returns the name of a cachefile from given data
611 *
612 * The needed directory is created by this function!
613 *
614 * @author Andreas Gohr <andi@splitbrain.org>
615 *
616 * @param string $data  This data is used to create a unique md5 name
617 * @param string $ext   This is appended to the filename if given
618 * @return string       The filename of the cachefile
619 */
620function getCacheName($data,$ext=''){
621    global $conf;
622    $md5  = md5($data);
623    $file = $conf['cachedir'].'/'.$md5[0].'/'.$md5.$ext;
624    io_makeFileDir($file);
625    return $file;
626}
627
628/**
629 * Checks a pageid against $conf['hidepages']
630 *
631 * @author Andreas Gohr <gohr@cosmocode.de>
632 *
633 * @param string $id page id
634 * @return bool
635 */
636function isHiddenPage($id){
637    $data = array(
638        'id' => $id,
639        'hidden' => false
640    );
641    \dokuwiki\Extension\Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
642    return $data['hidden'];
643}
644
645/**
646 * callback checks if page is hidden
647 *
648 * @param array $data event data    - see isHiddenPage()
649 */
650function _isHiddenPage(&$data) {
651    global $conf;
652    global $ACT;
653
654    if ($data['hidden']) return;
655    if(empty($conf['hidepages'])) return;
656    if($ACT == 'admin') return;
657
658    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
659        $data['hidden'] = true;
660    }
661}
662
663/**
664 * Reverse of isHiddenPage
665 *
666 * @author Andreas Gohr <gohr@cosmocode.de>
667 *
668 * @param string $id page id
669 * @return bool
670 */
671function isVisiblePage($id){
672    return !isHiddenPage($id);
673}
674
675/**
676 * Format an id for output to a user
677 *
678 * Namespaces are denoted by a trailing “:*”. The root namespace is
679 * “*”. Output is escaped.
680 *
681 * @author Adrian Lang <lang@cosmocode.de>
682 *
683 * @param string $id page id
684 * @return string
685 */
686function prettyprint_id($id) {
687    if (!$id || $id === ':') {
688        return '*';
689    }
690    if ((substr($id, -1, 1) === ':')) {
691        $id .= '*';
692    }
693    return hsc($id);
694}
695
696/**
697 * Encode a UTF-8 filename to use on any filesystem
698 *
699 * Uses the 'fnencode' option to determine encoding
700 *
701 * When the second parameter is true the string will
702 * be encoded only if non ASCII characters are detected -
703 * This makes it safe to run it multiple times on the
704 * same string (default is true)
705 *
706 * @author Andreas Gohr <andi@splitbrain.org>
707 * @see    urlencode
708 *
709 * @param string $file file name
710 * @param bool   $safe if true, only encoded when non ASCII characters detected
711 * @return string
712 */
713function utf8_encodeFN($file,$safe=true){
714    global $conf;
715    if($conf['fnencode'] == 'utf-8') return $file;
716
717    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
718        return $file;
719    }
720
721    if($conf['fnencode'] == 'safe'){
722        return SafeFN::encode($file);
723    }
724
725    $file = urlencode($file);
726    $file = str_replace('%2F','/',$file);
727    return $file;
728}
729
730/**
731 * Decode a filename back to UTF-8
732 *
733 * Uses the 'fnencode' option to determine encoding
734 *
735 * @author Andreas Gohr <andi@splitbrain.org>
736 * @see    urldecode
737 *
738 * @param string $file file name
739 * @return string
740 */
741function utf8_decodeFN($file){
742    global $conf;
743    if($conf['fnencode'] == 'utf-8') return $file;
744
745    if($conf['fnencode'] == 'safe'){
746        return SafeFN::decode($file);
747    }
748
749    return urldecode($file);
750}
751
752/**
753 * Find a page in the current namespace (determined from $ID) or any
754 * higher namespace that can be accessed by the current user,
755 * this condition can be overriden by an optional parameter.
756 *
757 * Used for sidebars, but can be used other stuff as well
758 *
759 * @todo   add event hook
760 *
761 * @param  string $page the pagename you're looking for
762 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
763 * @return false|string the full page id of the found page, false if any
764 */
765function page_findnearest($page, $useacl = true){
766    if ((string) $page === '') return false;
767    global $ID;
768
769    $ns = $ID;
770    do {
771        $ns = getNS($ns);
772        $pageid = cleanID("$ns:$page");
773        if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
774            return $pageid;
775        }
776    } while($ns !== false);
777
778    return false;
779}
780