xref: /dokuwiki/inc/pageutils.php (revision 47de339b47c45069e8b9525bc0ef4396bcc60cfd) 
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.utf8_basename($INPUT->server->str('SCRIPT_FILENAME'));
48
49        }elseif($INPUT->server->str('PATH_INFO')){
50            $request = $INPUT->server->str('PATH_INFO');
51        }elseif($INPUT->server->str('SCRIPT_NAME')){
52            $script = $INPUT->server->str('SCRIPT_NAME');
53        }elseif($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')){
54            $script = preg_replace ('/^'.preg_quote($INPUT->server->str('DOCUMENT_ROOT'),'/').'/','',
55                    $INPUT->server->str('SCRIPT_FILENAME'));
56            $script = '/'.$script;
57        }
58
59        //clean script and request (fixes a windows problem)
60        $script  = preg_replace('/\/\/+/','/',$script);
61        $request = preg_replace('/\/\/+/','/',$request);
62
63        //remove script URL and Querystring to gain the id
64        if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
65            $id = preg_replace ('/\?.*/','',$match[1]);
66        }
67        $id = urldecode($id);
68        //strip leading slashes
69        $id = preg_replace('!^/+!','',$id);
70    }
71
72    // Namespace autolinking from URL
73    if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
74        if(page_exists($id.$conf['start'])){
75            // start page inside namespace
76            $id = $id.$conf['start'];
77        }elseif(page_exists($id.noNS(cleanID($id)))){
78            // page named like the NS inside the NS
79            $id = $id.noNS(cleanID($id));
80        }elseif(page_exists($id)){
81            // page like namespace exists
82            $id = substr($id,0,-1);
83        }else{
84            // fall back to default
85            $id = $id.$conf['start'];
86        }
87        if (isset($ACT) && $ACT === 'show') {
88            $urlParameters = $_GET;
89            if (isset($urlParameters['id'])) {
90                unset($urlParameters['id']);
91            }
92            send_redirect(wl($id, $urlParameters, true, '&'));
93        }
94    }
95    if($clean) $id = cleanID($id);
96    if($id === '' && $param=='id') $id = $conf['start'];
97
98    return $id;
99}
100
101/**
102 * Remove unwanted chars from ID
103 *
104 * Cleans a given ID to only use allowed characters. Accented characters are
105 * converted to unaccented ones
106 *
107 * @author Andreas Gohr <andi@splitbrain.org>
108 *
109 * @param  string  $raw_id    The pageid to clean
110 * @param  boolean $ascii     Force ASCII
111 * @return string cleaned id
112 */
113function cleanID($raw_id,$ascii=false){
114    global $conf;
115    static $sepcharpat = null;
116
117    global $cache_cleanid;
118    $cache = & $cache_cleanid;
119
120    // check if it's already in the memory cache
121    if (!$ascii && isset($cache[(string)$raw_id])) {
122        return $cache[(string)$raw_id];
123    }
124
125    $sepchar = $conf['sepchar'];
126    if($sepcharpat == null) // build string only once to save clock cycles
127        $sepcharpat = '#\\'.$sepchar.'+#';
128
129    $id = trim((string)$raw_id);
130    $id = utf8_strtolower($id);
131
132    //alternative namespace seperator
133    if($conf['useslash']){
134        $id = strtr($id,';/','::');
135    }else{
136        $id = strtr($id,';/',':'.$sepchar);
137    }
138
139    if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
140    if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
141
142    //remove specials
143    $id = utf8_stripspecials($id,$sepchar,'\*');
144
145    if($ascii) $id = utf8_strip($id);
146
147    //clean up
148    $id = preg_replace($sepcharpat,$sepchar,$id);
149    $id = preg_replace('#:+#',':',$id);
150    $id = trim($id,':._-');
151    $id = preg_replace('#:[:\._\-]+#',':',$id);
152    $id = preg_replace('#[:\._\-]+:#',':',$id);
153
154    if (!$ascii) $cache[(string)$raw_id] = $id;
155    return($id);
156}
157
158/**
159 * Return namespacepart of a wiki ID
160 *
161 * @author Andreas Gohr <andi@splitbrain.org>
162 *
163 * @param string $id
164 * @return string|false the namespace part or false if the given ID has no namespace (root)
165 */
166function getNS($id){
167    $pos = strrpos((string)$id,':');
168    if($pos!==false){
169        return substr((string)$id,0,$pos);
170    }
171    return false;
172}
173
174/**
175 * Returns the ID without the namespace
176 *
177 * @author Andreas Gohr <andi@splitbrain.org>
178 *
179 * @param string $id
180 * @return string
181 */
182function noNS($id) {
183    $pos = strrpos($id, ':');
184    if ($pos!==false) {
185        return substr($id, $pos+1);
186    } else {
187        return $id;
188    }
189}
190
191/**
192 * Returns the current namespace
193 *
194 * @author Nathan Fritz <fritzn@crown.edu>
195 *
196 * @param string $id
197 * @return string
198 */
199function curNS($id) {
200    return noNS(getNS($id));
201}
202
203/**
204 * Returns the ID without the namespace or current namespace for 'start' pages
205 *
206 * @author Nathan Fritz <fritzn@crown.edu>
207 *
208 * @param string $id
209 * @return string
210 */
211function noNSorNS($id) {
212    global $conf;
213
214    $p = noNS($id);
215    if ($p == $conf['start'] || $p == false) {
216        $p = curNS($id);
217        if ($p == false) {
218            return $conf['start'];
219        }
220    }
221    return $p;
222}
223
224/**
225 * Creates a XHTML valid linkid from a given headline title
226 *
227 * @param string  $title   The headline title
228 * @param array|bool   $check   Existing IDs (title => number)
229 * @return string the title
230 *
231 * @author Andreas Gohr <andi@splitbrain.org>
232 */
233function sectionID($title,&$check) {
234    $title = str_replace(array(':','.'),'',cleanID($title));
235    $new = ltrim($title,'0123456789_-');
236    if(empty($new)){
237        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
238    }else{
239        $title = $new;
240    }
241
242    if(is_array($check)){
243        // make sure tiles are unique
244        if (!array_key_exists ($title,$check)) {
245            $check[$title] = 0;
246        } else {
247            $title .= ++ $check[$title];
248        }
249    }
250
251    return $title;
252}
253
254/**
255 * Wiki page existence check
256 *
257 * parameters as for wikiFN
258 *
259 * @author Chris Smith <chris@jalakai.co.uk>
260 *
261 * @param string $id page id
262 * @param string|int $rev empty or revision timestamp
263 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
264 * @param bool $date_at
265 * @return bool exists?
266 */
267function page_exists($id,$rev='',$clean=true, $date_at=false) {
268    if($rev !== '' && $date_at) {
269        $pagelog = new PageChangeLog($id);
270        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
271        if($pagelog_rev !== false)
272            $rev = $pagelog_rev;
273    }
274    return file_exists(wikiFN($id,$rev,$clean));
275}
276
277/**
278 * returns the full path to the datafile specified by ID and optional revision
279 *
280 * The filename is URL encoded to protect Unicode chars
281 *
282 * @param  $raw_id  string   id of wikipage
283 * @param  $rev     int|string   page revision, empty string for current
284 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
285 *                           when $id is guaranteed to have been cleaned already.
286 * @return string full path
287 *
288 * @author Andreas Gohr <andi@splitbrain.org>
289 */
290function wikiFN($raw_id,$rev='',$clean=true){
291    global $conf;
292
293    global $cache_wikifn;
294    $cache = & $cache_wikifn;
295
296    $id = $raw_id;
297
298    if ($clean) $id = cleanID($id);
299    $id = str_replace(':','/',$id);
300
301    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
302        return $cache[$id][$rev];
303    }
304
305    if(empty($rev)){
306        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
307    }else{
308        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
309        if($conf['compression']){
310            //test for extensions here, we want to read both compressions
311            if (file_exists($fn . '.gz')){
312                $fn .= '.gz';
313            }else if(file_exists($fn . '.bz2')){
314                $fn .= '.bz2';
315            }else{
316                //file doesnt exist yet, so we take the configured extension
317                $fn .= '.' . $conf['compression'];
318            }
319        }
320    }
321
322    if (!isset($cache[$id])) { $cache[$id] = array(); }
323    $cache[$id][$rev] = $fn;
324    return $fn;
325}
326
327/**
328 * Returns the full path to the file for locking the page while editing.
329 *
330 * @author Ben Coburn <btcoburn@silicodon.net>
331 *
332 * @param string $id page id
333 * @return string full path
334 */
335function wikiLockFN($id) {
336    global $conf;
337    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
338}
339
340
341/**
342 * returns the full path to the meta file specified by ID and extension
343 *
344 * @author Steven Danz <steven-danz@kc.rr.com>
345 *
346 * @param string $id   page id
347 * @param string $ext  file extension
348 * @return string full path
349 */
350function metaFN($id,$ext){
351    global $conf;
352    $id = cleanID($id);
353    $id = str_replace(':','/',$id);
354    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
355    return $fn;
356}
357
358/**
359 * returns the full path to the media's meta file specified by ID and extension
360 *
361 * @author Kate Arzamastseva <pshns@ukr.net>
362 *
363 * @param string $id   media id
364 * @param string $ext  extension of media
365 * @return string
366 */
367function mediaMetaFN($id,$ext){
368    global $conf;
369    $id = cleanID($id);
370    $id = str_replace(':','/',$id);
371    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
372    return $fn;
373}
374
375/**
376 * returns an array of full paths to all metafiles of a given ID
377 *
378 * @author Esther Brunner <esther@kaffeehaus.ch>
379 * @author Michael Hamann <michael@content-space.de>
380 *
381 * @param string $id page id
382 * @return array
383 */
384function metaFiles($id){
385    $basename = metaFN($id, '');
386    $files    = glob($basename.'.*', GLOB_MARK);
387    // filter files like foo.bar.meta when $id == 'foo'
388    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
389}
390
391/**
392 * returns the full path to the mediafile specified by ID
393 *
394 * The filename is URL encoded to protect Unicode chars
395 *
396 * @author Andreas Gohr <andi@splitbrain.org>
397 * @author Kate Arzamastseva <pshns@ukr.net>
398 *
399 * @param string     $id  media id
400 * @param string|int $rev empty string or revision timestamp
401 * @param bool $clean
402 *
403 * @return string full path
404 */
405function mediaFN($id, $rev='', $clean=true){
406    global $conf;
407    if ($clean) $id = cleanID($id);
408    $id = str_replace(':','/',$id);
409    if(empty($rev)){
410        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
411    }else{
412        $ext = mimetype($id);
413        $name = substr($id,0, -1*strlen($ext[0])-1);
414        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
415    }
416    return $fn;
417}
418
419/**
420 * Returns the full filepath to a localized file if local
421 * version isn't found the english one is returned
422 *
423 * @param  string $id  The id of the local file
424 * @param  string $ext The file extension (usually txt)
425 * @return string full filepath to localized file
426 *
427 * @author Andreas Gohr <andi@splitbrain.org>
428 */
429function localeFN($id,$ext='txt'){
430    global $conf;
431    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
432    if(!file_exists($file)){
433        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
434        if(!file_exists($file)){
435            //fall back to english
436            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
437        }
438    }
439    return $file;
440}
441
442/**
443 * Resolve relative paths in IDs
444 *
445 * Do not call directly use resolve_mediaid or resolve_pageid
446 * instead
447 *
448 * Partyly based on a cleanPath function found at
449 * http://php.net/manual/en/function.realpath.php#57016
450 *
451 * @author <bart at mediawave dot nl>
452 *
453 * @param string $ns     namespace which is context of id
454 * @param string $id     relative id
455 * @param bool   $clean  flag indicating that id should be cleaned
456 * @return string
457 */
458function resolve_id($ns,$id,$clean=true){
459    global $conf;
460
461    // some pre cleaning for useslash:
462    if($conf['useslash']) $id = str_replace('/',':',$id);
463
464    // if the id starts with a dot we need to handle the
465    // relative stuff
466    if($id && $id{0} == '.'){
467        // normalize initial dots without a colon
468        $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
469        // prepend the current namespace
470        $id = $ns.':'.$id;
471
472        // cleanup relatives
473        $result = array();
474        $pathA  = explode(':', $id);
475        if (!$pathA[0]) $result[] = '';
476        foreach ($pathA AS $key => $dir) {
477            if ($dir == '..') {
478                if (end($result) == '..') {
479                    $result[] = '..';
480                } elseif (!array_pop($result)) {
481                    $result[] = '..';
482                }
483            } elseif ($dir && $dir != '.') {
484                $result[] = $dir;
485            }
486        }
487        if (!end($pathA)) $result[] = '';
488        $id = implode(':', $result);
489    }elseif($ns !== false && strpos($id,':') === false){
490        //if link contains no namespace. add current namespace (if any)
491        $id = $ns.':'.$id;
492    }
493
494    if($clean) $id = cleanID($id);
495    return $id;
496}
497
498/**
499 * Returns a full media id
500 *
501 * @author Andreas Gohr <andi@splitbrain.org>
502 *
503 * @param string $ns namespace which is context of id
504 * @param string &$page (reference) relative media id, updated to resolved id
505 * @param bool &$exists (reference) updated with existance of media
506 * @param int|string $rev
507 * @param bool $date_at
508 */
509function resolve_mediaid($ns,&$page,&$exists,$rev='',$date_at=false){
510    $page   = resolve_id($ns,$page);
511    if($rev !== '' &&  $date_at){
512        $medialog = new MediaChangeLog($page);
513        $medialog_rev = $medialog->getLastRevisionAt($rev);
514        if($medialog_rev !== false) {
515            $rev = $medialog_rev;
516        }
517    }
518
519    $file   = mediaFN($page,$rev);
520    $exists = file_exists($file);
521}
522
523/**
524 * Returns a full page id
525 *
526 * @author Andreas Gohr <andi@splitbrain.org>
527 *
528 * @param string $ns namespace which is context of id
529 * @param string &$page (reference) relative page id, updated to resolved id
530 * @param bool &$exists (reference) updated with existance of media
531 * @param string $rev
532 * @param bool $date_at
533 */
534function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false ){
535    global $conf;
536    global $ID;
537    $exists = false;
538
539    //empty address should point to current page
540    if ($page === "") {
541        $page = $ID;
542    }
543
544    //keep hashlink if exists then clean both parts
545    if (strpos($page,'#')) {
546        list($page,$hash) = explode('#',$page,2);
547    } else {
548        $hash = '';
549    }
550    $hash = cleanID($hash);
551    $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
552
553    // get filename (calls clean itself)
554    if($rev !== '' && $date_at) {
555        $pagelog = new PageChangeLog($page);
556        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
557        if($pagelog_rev !== false)//something found
558           $rev  = $pagelog_rev;
559    }
560    $file = wikiFN($page,$rev);
561
562    // if ends with colon or slash we have a namespace link
563    if(in_array(substr($page,-1), array(':', ';')) ||
564       ($conf['useslash'] && substr($page,-1) == '/')){
565        if(page_exists($page.$conf['start'],$rev,true,$date_at)){
566            // start page inside namespace
567            $page = $page.$conf['start'];
568            $exists = true;
569        }elseif(page_exists($page.noNS(cleanID($page)),$rev,true,$date_at)){
570            // page named like the NS inside the NS
571            $page = $page.noNS(cleanID($page));
572            $exists = true;
573        }elseif(page_exists($page,$rev,true,$date_at)){
574            // page like namespace exists
575            $page = $page;
576            $exists = true;
577        }else{
578            // fall back to default
579            $page = $page.$conf['start'];
580        }
581    }else{
582        //check alternative plural/nonplural form
583        if(!file_exists($file)){
584            if( $conf['autoplural'] ){
585                if(substr($page,-1) == 's'){
586                    $try = substr($page,0,-1);
587                }else{
588                    $try = $page.'s';
589                }
590                if(page_exists($try,$rev,true,$date_at)){
591                    $page   = $try;
592                    $exists = true;
593                }
594            }
595        }else{
596            $exists = true;
597        }
598    }
599
600    // now make sure we have a clean page
601    $page = cleanID($page);
602
603    //add hash if any
604    if(!empty($hash)) $page .= '#'.$hash;
605}
606
607/**
608 * Returns the name of a cachefile from given data
609 *
610 * The needed directory is created by this function!
611 *
612 * @author Andreas Gohr <andi@splitbrain.org>
613 *
614 * @param string $data  This data is used to create a unique md5 name
615 * @param string $ext   This is appended to the filename if given
616 * @return string       The filename of the cachefile
617 */
618function getCacheName($data,$ext=''){
619    global $conf;
620    $md5  = md5($data);
621    $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
622    io_makeFileDir($file);
623    return $file;
624}
625
626/**
627 * Checks a pageid against $conf['hidepages']
628 *
629 * @author Andreas Gohr <gohr@cosmocode.de>
630 *
631 * @param string $id page id
632 * @return bool
633 */
634function isHiddenPage($id){
635    $data = array(
636        'id' => $id,
637        'hidden' => false
638    );
639    trigger_event('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
640    return $data['hidden'];
641}
642
643/**
644 * callback checks if page is hidden
645 *
646 * @param array $data event data    - see isHiddenPage()
647 */
648function _isHiddenPage(&$data) {
649    global $conf;
650    global $ACT;
651
652    if ($data['hidden']) return;
653    if(empty($conf['hidepages'])) return;
654    if($ACT == 'admin') return;
655
656    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
657        $data['hidden'] = true;
658    }
659}
660
661/**
662 * Reverse of isHiddenPage
663 *
664 * @author Andreas Gohr <gohr@cosmocode.de>
665 *
666 * @param string $id page id
667 * @return bool
668 */
669function isVisiblePage($id){
670    return !isHiddenPage($id);
671}
672
673/**
674 * Format an id for output to a user
675 *
676 * Namespaces are denoted by a trailing “:*”. The root namespace is
677 * “*”. Output is escaped.
678 *
679 * @author Adrian Lang <lang@cosmocode.de>
680 *
681 * @param string $id page id
682 * @return string
683 */
684function prettyprint_id($id) {
685    if (!$id || $id === ':') {
686        return '*';
687    }
688    if ((substr($id, -1, 1) === ':')) {
689        $id .= '*';
690    }
691    return hsc($id);
692}
693
694/**
695 * Encode a UTF-8 filename to use on any filesystem
696 *
697 * Uses the 'fnencode' option to determine encoding
698 *
699 * When the second parameter is true the string will
700 * be encoded only if non ASCII characters are detected -
701 * This makes it safe to run it multiple times on the
702 * same string (default is true)
703 *
704 * @author Andreas Gohr <andi@splitbrain.org>
705 * @see    urlencode
706 *
707 * @param string $file file name
708 * @param bool   $safe if true, only encoded when non ASCII characters detected
709 * @return string
710 */
711function utf8_encodeFN($file,$safe=true){
712    global $conf;
713    if($conf['fnencode'] == 'utf-8') return $file;
714
715    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
716        return $file;
717    }
718
719    if($conf['fnencode'] == 'safe'){
720        return SafeFN::encode($file);
721    }
722
723    $file = urlencode($file);
724    $file = str_replace('%2F','/',$file);
725    return $file;
726}
727
728/**
729 * Decode a filename back to UTF-8
730 *
731 * Uses the 'fnencode' option to determine encoding
732 *
733 * @author Andreas Gohr <andi@splitbrain.org>
734 * @see    urldecode
735 *
736 * @param string $file file name
737 * @return string
738 */
739function utf8_decodeFN($file){
740    global $conf;
741    if($conf['fnencode'] == 'utf-8') return $file;
742
743    if($conf['fnencode'] == 'safe'){
744        return SafeFN::decode($file);
745    }
746
747    return urldecode($file);
748}
749
750/**
751 * Find a page in the current namespace (determined from $ID) or any
752 * higher namespace that can be accessed by the current user,
753 * this condition can be overriden by an optional parameter.
754 *
755 * Used for sidebars, but can be used other stuff as well
756 *
757 * @todo   add event hook
758 *
759 * @param  string $page the pagename you're looking for
760 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
761 * @return false|string the full page id of the found page, false if any
762 */
763function page_findnearest($page, $useacl = true){
764    if ((string) $page === '') return false;
765    global $ID;
766
767    $ns = $ID;
768    do {
769        $ns = getNS($ns);
770        $pageid = cleanID("$ns:$page");
771        if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
772            return $pageid;
773        }
774    } while($ns !== false);
775
776    return false;
777}
778