Merge branch 'master' into revisionHandle3
[dokuwiki.git] / inc / pageutils.php
blobe5d0ca390b6c409b9ffbc6c2f32076c030b20f7c
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 || $p === '') {
217 $p = curNS($id);
218 if ($p === false || $p === '') {
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 $check Existing IDs
230 * @return string the title
232 * @author Andreas Gohr <andi@splitbrain.org>
234 function sectionID($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 $suffix=0;
245 $candidateTitle = $title;
246 while(in_array($candidateTitle, $check)){
247 $candidateTitle = $title . ++$suffix;
249 $check []= $candidateTitle;
250 return $candidateTitle;
251 } else {
252 return $title;
257 * Wiki page existence check
259 * parameters as for wikiFN
261 * @author Chris Smith <chris@jalakai.co.uk>
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 $date_at
267 * @return bool exists?
269 function page_exists($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 = $pagelog_rev;
276 return file_exists(wikiFN($id, $rev, $clean));
280 * returns the full path to the datafile specified by ID and optional revision
282 * The filename is URL encoded to protect Unicode chars
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 $id is guaranteed to have been cleaned already.
288 * @return string full path
290 * @author Andreas Gohr <andi@splitbrain.org>
292 function wikiFN($raw_id,$rev='',$clean=true){
293 global $conf;
295 global $cache_wikifn;
296 $cache = & $cache_wikifn;
298 $id = $raw_id;
300 if ($clean) $id = cleanID($id);
301 $id = str_replace(':','/',$id);
303 if (isset($cache[$id]) && isset($cache[$id][$rev])) {
304 return $cache[$id][$rev];
307 if(empty($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 .= '.' . $conf['compression'];
324 if (!isset($cache[$id])) { $cache[$id] = array(); }
325 $cache[$id][$rev] = $fn;
326 return $fn;
330 * Returns the full path to the file for locking the page while editing.
332 * @author Ben Coburn <btcoburn@silicodon.net>
334 * @param string $id page id
335 * @return string full path
337 function wikiLockFN($id) {
338 global $conf;
339 return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
344 * returns the full path to the meta file specified by ID and extension
346 * @author Steven Danz <steven-danz@kc.rr.com>
348 * @param string $id page id
349 * @param string $ext file extension
350 * @return string full path
352 function metaFN($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;
361 * returns the full path to the media's meta file specified by ID and extension
363 * @author Kate Arzamastseva <pshns@ukr.net>
365 * @param string $id media id
366 * @param string $ext extension of media
367 * @return string
369 function mediaMetaFN($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;
378 * returns an array of full paths to all metafiles of a given ID
380 * @author Esther Brunner <esther@kaffeehaus.ch>
381 * @author Michael Hamann <michael@content-space.de>
383 * @param string $id page id
384 * @return array
386 function metaFiles($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();
394 * returns the full path to the mediafile specified by ID
396 * The filename is URL encoded to protect Unicode chars
398 * @author Andreas Gohr <andi@splitbrain.org>
399 * @author Kate Arzamastseva <pshns@ukr.net>
401 * @param string $id media id
402 * @param string|int $rev empty string or revision timestamp
403 * @param bool $clean
405 * @return string full path
407 function mediaFN($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]);
418 return $fn;
422 * Returns the full filepath to a localized file if local
423 * version isn't found the english one is returned
425 * @param string $id The id of the local file
426 * @param string $ext The file extension (usually txt)
427 * @return string full filepath to localized file
429 * @author Andreas Gohr <andi@splitbrain.org>
431 function localeFN($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;
441 return $file;
445 * Resolve relative paths in IDs
447 * Do not call directly use resolve_mediaid or resolve_pageid
448 * instead
450 * Partyly based on a cleanPath function found at
451 * http://php.net/manual/en/function.realpath.php#57016
453 * @author <bart at mediawave dot nl>
455 * @param string $ns namespace which is context of id
456 * @param string $id relative id
457 * @param bool $clean flag indicating that id should be cleaned
458 * @return string
460 function resolve_id($ns,$id,$clean=true){
461 global $conf;
463 // some pre cleaning for useslash:
464 if($conf['useslash']) $id = str_replace('/',':',$id);
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;
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 $result[] = '..';
485 } elseif ($dir && $dir != '.') {
486 $result[] = $dir;
489 if (!end($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;
496 if($clean) $id = cleanID($id);
497 return $id;
501 * Returns a full media id
503 * @author Andreas Gohr <andi@splitbrain.org>
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 $date_at
511 function resolve_mediaid($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;
521 $exists = file_exists(mediaFN($page, $rev));
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) == '/')
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'];
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 = $page.'s';
592 if (page_exists($try, $rev, true, $date_at)) {
593 $page = $try;
594 $exists = true;
597 } else {
598 $exists = true;
602 // now make sure we have a clean page
603 $page = cleanID($page);
605 //add hash if any
606 if (!empty($hash)) $page .= '#'.$hash;
610 * Returns the name of a cachefile from given data
612 * The needed directory is created by this function!
614 * @author Andreas Gohr <andi@splitbrain.org>
616 * @param string $data This data is used to create a unique md5 name
617 * @param string $ext This is appended to the filename if given
618 * @return string The filename of the cachefile
620 function getCacheName($data,$ext=''){
621 global $conf;
622 $md5 = md5($data);
623 $file = $conf['cachedir'].'/'.$md5[0].'/'.$md5.$ext;
624 io_makeFileDir($file);
625 return $file;
629 * Checks a pageid against $conf['hidepages']
631 * @author Andreas Gohr <gohr@cosmocode.de>
633 * @param string $id page id
634 * @return bool
636 function isHiddenPage($id){
637 $data = array(
638 'id' => $id,
639 'hidden' => false
641 \dokuwiki\Extension\Event::createAndTrigger('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
642 return $data['hidden'];
646 * callback checks if page is hidden
648 * @param array $data event data - see isHiddenPage()
650 function _isHiddenPage(&$data) {
651 global $conf;
652 global $ACT;
654 if ($data['hidden']) return;
655 if(empty($conf['hidepages'])) return;
656 if($ACT == 'admin') return;
658 if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
659 $data['hidden'] = true;
664 * Reverse of isHiddenPage
666 * @author Andreas Gohr <gohr@cosmocode.de>
668 * @param string $id page id
669 * @return bool
671 function isVisiblePage($id){
672 return !isHiddenPage($id);
676 * Format an id for output to a user
678 * Namespaces are denoted by a trailing “:*”. The root namespace is
679 * “*”. Output is escaped.
681 * @author Adrian Lang <lang@cosmocode.de>
683 * @param string $id page id
684 * @return string
686 function prettyprint_id($id) {
687 if (!$id || $id === ':') {
688 return '*';
690 if ((substr($id, -1, 1) === ':')) {
691 $id .= '*';
693 return hsc($id);
697 * Encode a UTF-8 filename to use on any filesystem
699 * Uses the 'fnencode' option to determine encoding
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)
706 * @author Andreas Gohr <andi@splitbrain.org>
707 * @see urlencode
709 * @param string $file file name
710 * @param bool $safe if true, only encoded when non ASCII characters detected
711 * @return string
713 function utf8_encodeFN($file,$safe=true){
714 global $conf;
715 if($conf['fnencode'] == 'utf-8') return $file;
717 if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
718 return $file;
721 if($conf['fnencode'] == 'safe'){
722 return SafeFN::encode($file);
725 $file = urlencode($file);
726 $file = str_replace('%2F','/',$file);
727 return $file;
731 * Decode a filename back to UTF-8
733 * Uses the 'fnencode' option to determine encoding
735 * @author Andreas Gohr <andi@splitbrain.org>
736 * @see urldecode
738 * @param string $file file name
739 * @return string
741 function utf8_decodeFN($file){
742 global $conf;
743 if($conf['fnencode'] == 'utf-8') return $file;
745 if($conf['fnencode'] == 'safe'){
746 return SafeFN::decode($file);
749 return urldecode($file);
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.
757 * Used for sidebars, but can be used other stuff as well
759 * @todo add event hook
761 * @param string $page the pagename you're looking for
762 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
763 * @return false|string the full page id of the found page, false if any
765 function page_findnearest($page, $useacl = true){
766 if ((string) $page === '') return false;
767 global $ID;
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 $pageid;
776 } while($ns !== false);
778 return false;