Merge pull request #3024 from splitbrain/cookieupdate
[dokuwiki.git] / inc / pageutils.php
blob8540d227e04617213580dec8f5ccd31a54ba867f
1 <?php
2 /**
3 * Utilities for handling pagenames
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 */
10 use dokuwiki\ChangeLog\MediaChangeLog;
11 use dokuwiki\ChangeLog\PageChangeLog;
13 /**
14 * Fetch the an ID from request
16 * Uses either standard $_REQUEST variable or extracts it from
17 * the full request URI when userewrite is set to 2
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.
22 * @author Andreas Gohr <andi@splitbrain.org>
24 * @param string $param the $_REQUEST variable name, default 'id'
25 * @param bool $clean if true, ID is cleaned
26 * @return string
28 function getID($param='id',$clean=true){
29 /** @var Input $INPUT */
30 global $INPUT;
31 global $conf;
32 global $ACT;
34 $id = $INPUT->str($param);
36 //construct page id from request URI
37 if(empty($id) && $conf['userewrite'] == 2){
38 $request = $INPUT->server->str('REQUEST_URI');
39 $script = '';
41 //get the script URL
42 if($conf['basedir']){
43 $relpath = '';
44 if($param != 'id') {
45 $relpath = 'lib/exe/';
47 $script = $conf['basedir'] . $relpath .
48 \dokuwiki\Utf8\PhpString::basename($INPUT->server->str('SCRIPT_FILENAME'));
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;
60 //clean script and request (fixes a windows problem)
61 $script = preg_replace('/\/\/+/','/',$script);
62 $request = preg_replace('/\/\/+/','/',$request);
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]);
68 $id = urldecode($id);
69 //strip leading slashes
70 $id = preg_replace('!^/+!','',$id);
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'];
88 if (isset($ACT) && $ACT === 'show') {
89 $urlParameters = $_GET;
90 if (isset($urlParameters['id'])) {
91 unset($urlParameters['id']);
93 send_redirect(wl($id, $urlParameters, true, '&'));
96 if($clean) $id = cleanID($id);
97 if($id === '' && $param=='id') $id = $conf['start'];
99 return $id;
103 * Remove unwanted chars from ID
105 * Cleans a given ID to only use allowed characters. Accented characters are
106 * converted to unaccented ones
108 * @author Andreas Gohr <andi@splitbrain.org>
110 * @param string $raw_id The pageid to clean
111 * @param boolean $ascii Force ASCII
112 * @return string cleaned id
114 function cleanID($raw_id,$ascii=false){
115 global $conf;
116 static $sepcharpat = null;
118 global $cache_cleanid;
119 $cache = & $cache_cleanid;
121 // check if it's already in the memory cache
122 if (!$ascii && isset($cache[(string)$raw_id])) {
123 return $cache[(string)$raw_id];
126 $sepchar = $conf['sepchar'];
127 if($sepcharpat == null) // build string only once to save clock cycles
128 $sepcharpat = '#\\'.$sepchar.'+#';
130 $id = trim((string)$raw_id);
131 $id = \dokuwiki\Utf8\PhpString::strtolower($id);
133 //alternative namespace seperator
134 if($conf['useslash']){
135 $id = strtr($id,';/','::');
136 }else{
137 $id = strtr($id,';/',':'.$sepchar);
140 if($conf['deaccent'] == 2 || $ascii) $id = \dokuwiki\Utf8\Clean::romanize($id);
141 if($conf['deaccent'] || $ascii) $id = \dokuwiki\Utf8\Clean::deaccent($id,-1);
143 //remove specials
144 $id = \dokuwiki\Utf8\Clean::stripspecials($id,$sepchar,'\*');
146 if($ascii) $id = \dokuwiki\Utf8\Clean::strip($id);
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);
155 if (!$ascii) $cache[(string)$raw_id] = $id;
156 return($id);
160 * Return namespacepart of a wiki ID
162 * @author Andreas Gohr <andi@splitbrain.org>
164 * @param string $id
165 * @return string|false the namespace part or false if the given ID has no namespace (root)
167 function getNS($id){
168 $pos = strrpos((string)$id,':');
169 if($pos!==false){
170 return substr((string)$id,0,$pos);
172 return false;
176 * Returns the ID without the namespace
178 * @author Andreas Gohr <andi@splitbrain.org>
180 * @param string $id
181 * @return string
183 function noNS($id) {
184 $pos = strrpos($id, ':');
185 if ($pos!==false) {
186 return substr($id, $pos+1);
187 } else {
188 return $id;
193 * Returns the current namespace
195 * @author Nathan Fritz <fritzn@crown.edu>
197 * @param string $id
198 * @return string
200 function curNS($id) {
201 return noNS(getNS($id));
205 * Returns the ID without the namespace or current namespace for 'start' pages
207 * @author Nathan Fritz <fritzn@crown.edu>
209 * @param string $id
210 * @return string
212 function noNSorNS($id) {
213 global $conf;
215 $p = noNS($id);
216 if ($p === $conf['start'] || $p === false) {
217 $p = curNS($id);
218 if ($p === false) {
219 return $conf['start'];
222 return $p;
226 * Creates a XHTML valid linkid from a given headline title
228 * @param string $title The headline title
229 * @param array|bool $check Existing IDs (title => number)
230 * @return string the title
232 * @author Andreas Gohr <andi@splitbrain.org>
234 function sectionID($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 = $new;
243 if(is_array($check)){
244 // make sure tiles are unique
245 if (!array_key_exists ($title,$check)) {
246 $check[$title] = 0;
247 } else {
248 $title .= ++ $check[$title];
252 return $title;
256 * Wiki page existence check
258 * parameters as for wikiFN
260 * @author Chris Smith <chris@jalakai.co.uk>
262 * @param string $id page id
263 * @param string|int $rev empty or revision timestamp
264 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
265 * @param bool $date_at
266 * @return bool exists?
268 function page_exists($id,$rev='',$clean=true, $date_at=false) {
269 if($rev !== '' && $date_at) {
270 $pagelog = new PageChangeLog($id);
271 $pagelog_rev = $pagelog->getLastRevisionAt($rev);
272 if($pagelog_rev !== false)
273 $rev = $pagelog_rev;
275 return file_exists(wikiFN($id,$rev,$clean));
279 * returns the full path to the datafile specified by ID and optional revision
281 * The filename is URL encoded to protect Unicode chars
283 * @param $raw_id string id of wikipage
284 * @param $rev int|string page revision, empty string for current
285 * @param $clean bool flag indicating that $raw_id should be cleaned. Only set to false
286 * when $id is guaranteed to have been cleaned already.
287 * @return string full path
289 * @author Andreas Gohr <andi@splitbrain.org>
291 function wikiFN($raw_id,$rev='',$clean=true){
292 global $conf;
294 global $cache_wikifn;
295 $cache = & $cache_wikifn;
297 $id = $raw_id;
299 if ($clean) $id = cleanID($id);
300 $id = str_replace(':','/',$id);
302 if (isset($cache[$id]) && isset($cache[$id][$rev])) {
303 return $cache[$id][$rev];
306 if(empty($rev)){
307 $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
308 }else{
309 $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
310 if($conf['compression']){
311 //test for extensions here, we want to read both compressions
312 if (file_exists($fn . '.gz')){
313 $fn .= '.gz';
314 }else if(file_exists($fn . '.bz2')){
315 $fn .= '.bz2';
316 }else{
317 //file doesnt exist yet, so we take the configured extension
318 $fn .= '.' . $conf['compression'];
323 if (!isset($cache[$id])) { $cache[$id] = array(); }
324 $cache[$id][$rev] = $fn;
325 return $fn;
329 * Returns the full path to the file for locking the page while editing.
331 * @author Ben Coburn <btcoburn@silicodon.net>
333 * @param string $id page id
334 * @return string full path
336 function wikiLockFN($id) {
337 global $conf;
338 return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
343 * returns the full path to the meta file specified by ID and extension
345 * @author Steven Danz <steven-danz@kc.rr.com>
347 * @param string $id page id
348 * @param string $ext file extension
349 * @return string full path
351 function metaFN($id,$ext){
352 global $conf;
353 $id = cleanID($id);
354 $id = str_replace(':','/',$id);
355 $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
356 return $fn;
360 * returns the full path to the media's meta file specified by ID and extension
362 * @author Kate Arzamastseva <pshns@ukr.net>
364 * @param string $id media id
365 * @param string $ext extension of media
366 * @return string
368 function mediaMetaFN($id,$ext){
369 global $conf;
370 $id = cleanID($id);
371 $id = str_replace(':','/',$id);
372 $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
373 return $fn;
377 * returns an array of full paths to all metafiles of a given ID
379 * @author Esther Brunner <esther@kaffeehaus.ch>
380 * @author Michael Hamann <michael@content-space.de>
382 * @param string $id page id
383 * @return array
385 function metaFiles($id){
386 $basename = metaFN($id, '');
387 $files = glob($basename.'.*', GLOB_MARK);
388 // filter files like foo.bar.meta when $id == 'foo'
389 return $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
393 * returns the full path to the mediafile specified by ID
395 * The filename is URL encoded to protect Unicode chars
397 * @author Andreas Gohr <andi@splitbrain.org>
398 * @author Kate Arzamastseva <pshns@ukr.net>
400 * @param string $id media id
401 * @param string|int $rev empty string or revision timestamp
402 * @param bool $clean
404 * @return string full path
406 function mediaFN($id, $rev='', $clean=true){
407 global $conf;
408 if ($clean) $id = cleanID($id);
409 $id = str_replace(':','/',$id);
410 if(empty($rev)){
411 $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
412 }else{
413 $ext = mimetype($id);
414 $name = substr($id,0, -1*strlen($ext[0])-1);
415 $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
417 return $fn;
421 * Returns the full filepath to a localized file if local
422 * version isn't found the english one is returned
424 * @param string $id The id of the local file
425 * @param string $ext The file extension (usually txt)
426 * @return string full filepath to localized file
428 * @author Andreas Gohr <andi@splitbrain.org>
430 function localeFN($id,$ext='txt'){
431 global $conf;
432 $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
433 if(!file_exists($file)){
434 $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
435 if(!file_exists($file)){
436 //fall back to english
437 $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
440 return $file;
444 * Resolve relative paths in IDs
446 * Do not call directly use resolve_mediaid or resolve_pageid
447 * instead
449 * Partyly based on a cleanPath function found at
450 * http://php.net/manual/en/function.realpath.php#57016
452 * @author <bart at mediawave dot nl>
454 * @param string $ns namespace which is context of id
455 * @param string $id relative id
456 * @param bool $clean flag indicating that id should be cleaned
457 * @return string
459 function resolve_id($ns,$id,$clean=true){
460 global $conf;
462 // some pre cleaning for useslash:
463 if($conf['useslash']) $id = str_replace('/',':',$id);
465 // if the id starts with a dot we need to handle the
466 // relative stuff
467 if($id && $id[0] == '.'){
468 // normalize initial dots without a colon
469 $id = preg_replace('/^((\.+:)*)(\.+)(?=[^:\.])/','\1\3:',$id);
470 // prepend the current namespace
471 $id = $ns.':'.$id;
473 // cleanup relatives
474 $result = array();
475 $pathA = explode(':', $id);
476 if (!$pathA[0]) $result[] = '';
477 foreach ($pathA AS $key => $dir) {
478 if ($dir == '..') {
479 if (end($result) == '..') {
480 $result[] = '..';
481 } elseif (!array_pop($result)) {
482 $result[] = '..';
484 } elseif ($dir && $dir != '.') {
485 $result[] = $dir;
488 if (!end($pathA)) $result[] = '';
489 $id = implode(':', $result);
490 }elseif($ns !== false && strpos($id,':') === false){
491 //if link contains no namespace. add current namespace (if any)
492 $id = $ns.':'.$id;
495 if($clean) $id = cleanID($id);
496 return $id;
500 * Returns a full media id
502 * @author Andreas Gohr <andi@splitbrain.org>
504 * @param string $ns namespace which is context of id
505 * @param string &$page (reference) relative media id, updated to resolved id
506 * @param bool &$exists (reference) updated with existance of media
507 * @param int|string $rev
508 * @param bool $date_at
510 function resolve_mediaid($ns,&$page,&$exists,$rev='',$date_at=false){
511 $page = resolve_id($ns,$page);
512 if($rev !== '' && $date_at){
513 $medialog = new MediaChangeLog($page);
514 $medialog_rev = $medialog->getLastRevisionAt($rev);
515 if($medialog_rev !== false) {
516 $rev = $medialog_rev;
520 $file = mediaFN($page,$rev);
521 $exists = file_exists($file);
525 * Returns a full page id
527 * @author Andreas Gohr <andi@splitbrain.org>
529 * @param string $ns namespace which is context of id
530 * @param string &$page (reference) relative page id, updated to resolved id
531 * @param bool &$exists (reference) updated with existance of media
532 * @param string $rev
533 * @param bool $date_at
535 function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false ){
536 global $conf;
537 global $ID;
538 $exists = false;
540 //empty address should point to current page
541 if ($page === "") {
542 $page = $ID;
545 //keep hashlink if exists then clean both parts
546 if (strpos($page,'#')) {
547 list($page,$hash) = explode('#',$page,2);
548 } else {
549 $hash = '';
551 $hash = cleanID($hash);
552 $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
554 // get filename (calls clean itself)
555 if($rev !== '' && $date_at) {
556 $pagelog = new PageChangeLog($page);
557 $pagelog_rev = $pagelog->getLastRevisionAt($rev);
558 if($pagelog_rev !== false)//something found
559 $rev = $pagelog_rev;
561 $file = wikiFN($page,$rev);
563 // if ends with colon or slash we have a namespace link
564 if(in_array(substr($page,-1), array(':', ';')) ||
565 ($conf['useslash'] && substr($page,-1) == '/')){
566 if(page_exists($page.$conf['start'],$rev,true,$date_at)){
567 // start page inside namespace
568 $page = $page.$conf['start'];
569 $exists = true;
570 }elseif(page_exists($page.noNS(cleanID($page)),$rev,true,$date_at)){
571 // page named like the NS inside the NS
572 $page = $page.noNS(cleanID($page));
573 $exists = true;
574 }elseif(page_exists($page,$rev,true,$date_at)){
575 // page like namespace exists
576 $page = $page;
577 $exists = true;
578 }else{
579 // fall back to default
580 $page = $page.$conf['start'];
582 }else{
583 //check alternative plural/nonplural form
584 if(!file_exists($file)){
585 if( $conf['autoplural'] ){
586 if(substr($page,-1) == 's'){
587 $try = substr($page,0,-1);
588 }else{
589 $try = $page.'s';
591 if(page_exists($try,$rev,true,$date_at)){
592 $page = $try;
593 $exists = true;
596 }else{
597 $exists = true;
601 // now make sure we have a clean page
602 $page = cleanID($page);
604 //add hash if any
605 if(!empty($hash)) $page .= '#'.$hash;
609 * Returns the name of a cachefile from given data
611 * The needed directory is created by this function!
613 * @author Andreas Gohr <andi@splitbrain.org>
615 * @param string $data This data is used to create a unique md5 name
616 * @param string $ext This is appended to the filename if given
617 * @return string The filename of the cachefile
619 function getCacheName($data,$ext=''){
620 global $conf;
621 $md5 = md5($data);
622 $file = $conf['cachedir'].'/'.$md5[0].'/'.$md5.$ext;
623 io_makeFileDir($file);
624 return $file;
628 * Checks a pageid against $conf['hidepages']
630 * @author Andreas Gohr <gohr@cosmocode.de>
632 * @param string $id page id
633 * @return bool
635 function isHiddenPage($id){
636 $data = array(
637 'id' => $id,
638 'hidden' => false
640 \dokuwiki\Extension\Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
641 return $data['hidden'];
645 * callback checks if page is hidden
647 * @param array $data event data - see isHiddenPage()
649 function _isHiddenPage(&$data) {
650 global $conf;
651 global $ACT;
653 if ($data['hidden']) return;
654 if(empty($conf['hidepages'])) return;
655 if($ACT == 'admin') return;
657 if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
658 $data['hidden'] = true;
663 * Reverse of isHiddenPage
665 * @author Andreas Gohr <gohr@cosmocode.de>
667 * @param string $id page id
668 * @return bool
670 function isVisiblePage($id){
671 return !isHiddenPage($id);
675 * Format an id for output to a user
677 * Namespaces are denoted by a trailing “:*”. The root namespace is
678 * “*”. Output is escaped.
680 * @author Adrian Lang <lang@cosmocode.de>
682 * @param string $id page id
683 * @return string
685 function prettyprint_id($id) {
686 if (!$id || $id === ':') {
687 return '*';
689 if ((substr($id, -1, 1) === ':')) {
690 $id .= '*';
692 return hsc($id);
696 * Encode a UTF-8 filename to use on any filesystem
698 * Uses the 'fnencode' option to determine encoding
700 * When the second parameter is true the string will
701 * be encoded only if non ASCII characters are detected -
702 * This makes it safe to run it multiple times on the
703 * same string (default is true)
705 * @author Andreas Gohr <andi@splitbrain.org>
706 * @see urlencode
708 * @param string $file file name
709 * @param bool $safe if true, only encoded when non ASCII characters detected
710 * @return string
712 function utf8_encodeFN($file,$safe=true){
713 global $conf;
714 if($conf['fnencode'] == 'utf-8') return $file;
716 if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
717 return $file;
720 if($conf['fnencode'] == 'safe'){
721 return SafeFN::encode($file);
724 $file = urlencode($file);
725 $file = str_replace('%2F','/',$file);
726 return $file;
730 * Decode a filename back to UTF-8
732 * Uses the 'fnencode' option to determine encoding
734 * @author Andreas Gohr <andi@splitbrain.org>
735 * @see urldecode
737 * @param string $file file name
738 * @return string
740 function utf8_decodeFN($file){
741 global $conf;
742 if($conf['fnencode'] == 'utf-8') return $file;
744 if($conf['fnencode'] == 'safe'){
745 return SafeFN::decode($file);
748 return urldecode($file);
752 * Find a page in the current namespace (determined from $ID) or any
753 * higher namespace that can be accessed by the current user,
754 * this condition can be overriden by an optional parameter.
756 * Used for sidebars, but can be used other stuff as well
758 * @todo add event hook
760 * @param string $page the pagename you're looking for
761 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
762 * @return false|string the full page id of the found page, false if any
764 function page_findnearest($page, $useacl = true){
765 if ((string) $page === '') return false;
766 global $ID;
768 $ns = $ID;
769 do {
770 $ns = getNS($ns);
771 $pageid = cleanID("$ns:$page");
772 if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
773 return $pageid;
775 } while($ns !== false);
777 return false;