xref: /dokuwiki/inc/common.php (revision b24e9c4a65ed9725af53e2f198c6475acd74eebc) 
1<?php
2/**
3 * Common DokuWiki functions
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 */
8
9use dokuwiki\Cache\CacheInstructions;
10use dokuwiki\Cache\CacheRenderer;
11use dokuwiki\ChangeLog\PageChangeLog;
12use dokuwiki\File\PageFile;
13use dokuwiki\Logger;
14use dokuwiki\Subscriptions\PageSubscriptionSender;
15use dokuwiki\Subscriptions\SubscriberManager;
16use dokuwiki\Extension\AuthPlugin;
17use dokuwiki\Extension\Event;
18
19/**
20 * Wrapper around htmlspecialchars()
21 *
22 * @author Andreas Gohr <andi@splitbrain.org>
23 * @see    htmlspecialchars()
24 *
25 * @param string $string the string being converted
26 * @return string converted string
27 */
28function hsc($string) {
29    return htmlspecialchars($string, ENT_QUOTES, 'UTF-8');
30}
31
32/**
33 * Checks if the given input is blank
34 *
35 * This is similar to empty() but will return false for "0".
36 *
37 * Please note: when you pass uninitialized variables, they will implicitly be created
38 * with a NULL value without warning.
39 *
40 * To avoid this it's recommended to guard the call with isset like this:
41 *
42 * (isset($foo) && !blank($foo))
43 * (!isset($foo) || blank($foo))
44 *
45 * @param $in
46 * @param bool $trim Consider a string of whitespace to be blank
47 * @return bool
48 */
49function blank(&$in, $trim = false) {
50    if(is_null($in)) return true;
51    if(is_array($in)) return empty($in);
52    if($in === "\0") return true;
53    if($trim && trim($in) === '') return true;
54    if(strlen($in) > 0) return false;
55    return empty($in);
56}
57
58/**
59 * print a newline terminated string
60 *
61 * You can give an indention as optional parameter
62 *
63 * @author Andreas Gohr <andi@splitbrain.org>
64 *
65 * @param string $string  line of text
66 * @param int    $indent  number of spaces indention
67 */
68function ptln($string, $indent = 0) {
69    echo str_repeat(' ', $indent)."$string\n";
70}
71
72/**
73 * strips control characters (<32) from the given string
74 *
75 * @author Andreas Gohr <andi@splitbrain.org>
76 *
77 * @param string $string being stripped
78 * @return string
79 */
80function stripctl($string) {
81    return preg_replace('/[\x00-\x1F]+/s', '', $string);
82}
83
84/**
85 * Return a secret token to be used for CSRF attack prevention
86 *
87 * @author  Andreas Gohr <andi@splitbrain.org>
88 * @link    http://en.wikipedia.org/wiki/Cross-site_request_forgery
89 * @link    http://christ1an.blogspot.com/2007/04/preventing-csrf-efficiently.html
90 *
91 * @return  string
92 */
93function getSecurityToken() {
94    /** @var Input $INPUT */
95    global $INPUT;
96
97    $user = $INPUT->server->str('REMOTE_USER');
98    $session = session_id();
99
100    // CSRF checks are only for logged in users - do not generate for anonymous
101    if(trim($user) == '' || trim($session) == '') return '';
102    return \dokuwiki\PassHash::hmac('md5', $session.$user, auth_cookiesalt());
103}
104
105/**
106 * Check the secret CSRF token
107 *
108 * @param null|string $token security token or null to read it from request variable
109 * @return bool success if the token matched
110 */
111function checkSecurityToken($token = null) {
112    /** @var Input $INPUT */
113    global $INPUT;
114    if(!$INPUT->server->str('REMOTE_USER')) return true; // no logged in user, no need for a check
115
116    if(is_null($token)) $token = $INPUT->str('sectok');
117    if(getSecurityToken() != $token) {
118        msg('Security Token did not match. Possible CSRF attack.', -1);
119        return false;
120    }
121    return true;
122}
123
124/**
125 * Print a hidden form field with a secret CSRF token
126 *
127 * @author  Andreas Gohr <andi@splitbrain.org>
128 *
129 * @param bool $print  if true print the field, otherwise html of the field is returned
130 * @return string html of hidden form field
131 */
132function formSecurityToken($print = true) {
133    $ret = '<div class="no"><input type="hidden" name="sectok" value="'.getSecurityToken().'" /></div>'."\n";
134    if($print) echo $ret;
135    return $ret;
136}
137
138/**
139 * Determine basic information for a request of $id
140 *
141 * @author Andreas Gohr <andi@splitbrain.org>
142 * @author Chris Smith <chris@jalakai.co.uk>
143 *
144 * @param string $id         pageid
145 * @param bool   $htmlClient add info about whether is mobile browser
146 * @return array with info for a request of $id
147 *
148 */
149function basicinfo($id, $htmlClient=true){
150    global $USERINFO;
151    /* @var Input $INPUT */
152    global $INPUT;
153
154    // set info about manager/admin status.
155    $info = array();
156    $info['isadmin']   = false;
157    $info['ismanager'] = false;
158    if($INPUT->server->has('REMOTE_USER')) {
159        $info['userinfo']   = $USERINFO;
160        $info['perm']       = auth_quickaclcheck($id);
161        $info['client']     = $INPUT->server->str('REMOTE_USER');
162
163        if($info['perm'] == AUTH_ADMIN) {
164            $info['isadmin']   = true;
165            $info['ismanager'] = true;
166        } elseif(auth_ismanager()) {
167            $info['ismanager'] = true;
168        }
169
170        // if some outside auth were used only REMOTE_USER is set
171        if(!$info['userinfo']['name']) {
172            $info['userinfo']['name'] = $INPUT->server->str('REMOTE_USER');
173        }
174
175    } else {
176        $info['perm']       = auth_aclcheck($id, '', null);
177        $info['client']     = clientIP(true);
178    }
179
180    $info['namespace'] = getNS($id);
181
182    // mobile detection
183    if ($htmlClient) {
184        $info['ismobile'] = clientismobile();
185    }
186
187    return $info;
188 }
189
190/**
191 * Return info about the current document as associative
192 * array.
193 *
194 * @author Andreas Gohr <andi@splitbrain.org>
195 *
196 * @return array with info about current document
197 */
198function pageinfo() {
199    global $ID;
200    global $REV;
201    global $RANGE;
202    global $lang;
203    /* @var Input $INPUT */
204    global $INPUT;
205
206    $info = basicinfo($ID);
207
208    // include ID & REV not redundant, as some parts of DokuWiki may temporarily change $ID, e.g. p_wiki_xhtml
209    // FIXME ... perhaps it would be better to ensure the temporary changes weren't necessary
210    $info['id']  = $ID;
211    $info['rev'] = $REV;
212
213    $subManager = new SubscriberManager();
214    $info['subscribed'] = $subManager->userSubscription();
215
216    $info['locked']     = checklock($ID);
217    $info['filepath']   = wikiFN($ID);
218    $info['exists']     = file_exists($info['filepath']);
219    $info['currentrev'] = @filemtime($info['filepath']);
220
221    if ($REV) {
222        //check if current revision was meant
223        if ($info['exists'] && ($info['currentrev'] == $REV)) {
224            $REV = '';
225        } elseif ($RANGE) {
226            //section editing does not work with old revisions!
227            $REV   = '';
228            $RANGE = '';
229            msg($lang['nosecedit'], 0);
230        } else {
231            //really use old revision
232            $info['filepath'] = wikiFN($ID, $REV);
233            $info['exists']   = file_exists($info['filepath']);
234        }
235    }
236    $info['rev'] = $REV;
237    if ($info['exists']) {
238        $info['writable'] = (is_writable($info['filepath']) && $info['perm'] >= AUTH_EDIT);
239    } else {
240        $info['writable'] = ($info['perm'] >= AUTH_CREATE);
241    }
242    $info['editable'] = ($info['writable'] && empty($info['locked']));
243    $info['lastmod']  = @filemtime($info['filepath']);
244
245    //load page meta data
246    $info['meta'] = p_get_metadata($ID);
247
248    //who's the editor
249    $pagelog = new PageChangeLog($ID, 1024);
250    if ($REV) {
251        $revinfo = $pagelog->getRevisionInfo($REV);
252    } else {
253        if (!empty($info['meta']['last_change']) && is_array($info['meta']['last_change'])) {
254            $revinfo = $info['meta']['last_change'];
255        } else {
256            $revinfo = $pagelog->getRevisionInfo($info['lastmod']);
257            // cache most recent changelog line in metadata if missing and still valid
258            if ($revinfo !== false) {
259                $info['meta']['last_change'] = $revinfo;
260                p_set_metadata($ID, array('last_change' => $revinfo));
261            }
262        }
263    }
264    //and check for an external edit
265    if ($revinfo !== false && $revinfo['date'] != $info['lastmod']) {
266        // cached changelog line no longer valid
267        $revinfo                     = false;
268        $info['meta']['last_change'] = $revinfo;
269        p_set_metadata($ID, array('last_change' => $revinfo));
270    }
271
272    if ($revinfo !== false) {
273        $info['ip']   = $revinfo['ip'];
274        $info['user'] = $revinfo['user'];
275        $info['sum']  = $revinfo['sum'];
276        // See also $INFO['meta']['last_change'] which is the most recent log line for page $ID.
277        // Use $INFO['meta']['last_change']['type']===DOKU_CHANGE_TYPE_MINOR_EDIT in place of $info['minor'].
278
279        $info['editor'] = $revinfo['user'] ?: $revinfo['ip'];
280    } else {
281        $info['ip']     = null;
282        $info['user']   = null;
283        $info['sum']    = null;
284        $info['editor'] = null;
285    }
286
287    // draft
288    $draft = new \dokuwiki\Draft($ID, $info['client']);
289    if ($draft->isDraftAvailable()) {
290        $info['draft'] = $draft->getDraftFilename();
291    }
292
293    return $info;
294}
295
296/**
297 * Initialize and/or fill global $JSINFO with some basic info to be given to javascript
298 */
299function jsinfo() {
300    global $JSINFO, $ID, $INFO, $ACT;
301
302    if (!is_array($JSINFO)) {
303        $JSINFO = [];
304    }
305    //export minimal info to JS, plugins can add more
306    $JSINFO['id']                    = $ID;
307    $JSINFO['namespace']             = isset($INFO) ? (string) $INFO['namespace'] : '';
308    $JSINFO['ACT']                   = act_clean($ACT);
309    $JSINFO['useHeadingNavigation']  = (int) useHeading('navigation');
310    $JSINFO['useHeadingContent']     = (int) useHeading('content');
311}
312
313/**
314 * Return information about the current media item as an associative array.
315 *
316 * @return array with info about current media item
317 */
318function mediainfo() {
319    global $NS;
320    global $IMG;
321
322    $info = basicinfo("$NS:*");
323    $info['image'] = $IMG;
324
325    return $info;
326}
327
328/**
329 * Build an string of URL parameters
330 *
331 * @author Andreas Gohr
332 *
333 * @param array  $params    array with key-value pairs
334 * @param string $sep       series of pairs are separated by this character
335 * @return string query string
336 */
337function buildURLparams($params, $sep = '&amp;') {
338    $url = '';
339    $amp = false;
340    foreach($params as $key => $val) {
341        if($amp) $url .= $sep;
342
343        $url .= rawurlencode($key).'=';
344        $url .= rawurlencode((string) $val);
345        $amp = true;
346    }
347    return $url;
348}
349
350/**
351 * Build an string of html tag attributes
352 *
353 * Skips keys starting with '_', values get HTML encoded
354 *
355 * @author Andreas Gohr
356 *
357 * @param array $params           array with (attribute name-attribute value) pairs
358 * @param bool  $skipEmptyStrings skip empty string values?
359 * @return string
360 */
361function buildAttributes($params, $skipEmptyStrings = false) {
362    $url   = '';
363    $white = false;
364    foreach($params as $key => $val) {
365        if($key[0] == '_') continue;
366        if($val === '' && $skipEmptyStrings) continue;
367        if($white) $url .= ' ';
368
369        $url .= $key.'="';
370        $url .= htmlspecialchars($val);
371        $url .= '"';
372        $white = true;
373    }
374    return $url;
375}
376
377/**
378 * This builds the breadcrumb trail and returns it as array
379 *
380 * @author Andreas Gohr <andi@splitbrain.org>
381 *
382 * @return string[] with the data: array(pageid=>name, ... )
383 */
384function breadcrumbs() {
385    // we prepare the breadcrumbs early for quick session closing
386    static $crumbs = null;
387    if($crumbs != null) return $crumbs;
388
389    global $ID;
390    global $ACT;
391    global $conf;
392    global $INFO;
393
394    //first visit?
395    $crumbs = isset($_SESSION[DOKU_COOKIE]['bc']) ? $_SESSION[DOKU_COOKIE]['bc'] : array();
396    //we only save on show and existing visible readable wiki documents
397    $file = wikiFN($ID);
398    if($ACT != 'show' || $INFO['perm'] < AUTH_READ || isHiddenPage($ID) || !file_exists($file)) {
399        $_SESSION[DOKU_COOKIE]['bc'] = $crumbs;
400        return $crumbs;
401    }
402
403    // page names
404    $name = noNSorNS($ID);
405    if(useHeading('navigation')) {
406        // get page title
407        $title = p_get_first_heading($ID, METADATA_RENDER_USING_SIMPLE_CACHE);
408        if($title) {
409            $name = $title;
410        }
411    }
412
413    //remove ID from array
414    if(isset($crumbs[$ID])) {
415        unset($crumbs[$ID]);
416    }
417
418    //add to array
419    $crumbs[$ID] = $name;
420    //reduce size
421    while(count($crumbs) > $conf['breadcrumbs']) {
422        array_shift($crumbs);
423    }
424    //save to session
425    $_SESSION[DOKU_COOKIE]['bc'] = $crumbs;
426    return $crumbs;
427}
428
429/**
430 * Filter for page IDs
431 *
432 * This is run on a ID before it is outputted somewhere
433 * currently used to replace the colon with something else
434 * on Windows (non-IIS) systems and to have proper URL encoding
435 *
436 * See discussions at https://github.com/splitbrain/dokuwiki/pull/84 and
437 * https://github.com/splitbrain/dokuwiki/pull/173 why we use a whitelist of
438 * unaffected servers instead of blacklisting affected servers here.
439 *
440 * Urlencoding is ommitted when the second parameter is false
441 *
442 * @author Andreas Gohr <andi@splitbrain.org>
443 *
444 * @param string $id pageid being filtered
445 * @param bool   $ue apply urlencoding?
446 * @return string
447 */
448function idfilter($id, $ue = true) {
449    global $conf;
450    /* @var Input $INPUT */
451    global $INPUT;
452
453    if($conf['useslash'] && $conf['userewrite']) {
454        $id = strtr($id, ':', '/');
455    } elseif(strtoupper(substr(PHP_OS, 0, 3)) === 'WIN' &&
456        $conf['userewrite'] &&
457        strpos($INPUT->server->str('SERVER_SOFTWARE'), 'Microsoft-IIS') === false
458    ) {
459        $id = strtr($id, ':', ';');
460    }
461    if($ue) {
462        $id = rawurlencode($id);
463        $id = str_replace('%3A', ':', $id); //keep as colon
464        $id = str_replace('%3B', ';', $id); //keep as semicolon
465        $id = str_replace('%2F', '/', $id); //keep as slash
466    }
467    return $id;
468}
469
470/**
471 * This builds a link to a wikipage
472 *
473 * It handles URL rewriting and adds additional parameters
474 *
475 * @author Andreas Gohr <andi@splitbrain.org>
476 *
477 * @param string       $id             page id, defaults to start page
478 * @param string|array $urlParameters  URL parameters, associative array recommended
479 * @param bool         $absolute       request an absolute URL instead of relative
480 * @param string       $separator      parameter separator
481 * @return string
482 */
483function wl($id = '', $urlParameters = '', $absolute = false, $separator = '&amp;') {
484    global $conf;
485    if(is_array($urlParameters)) {
486        if(isset($urlParameters['rev']) && !$urlParameters['rev']) unset($urlParameters['rev']);
487        if(isset($urlParameters['at']) && $conf['date_at_format']) {
488            $urlParameters['at'] = date($conf['date_at_format'], $urlParameters['at']);
489        }
490        $urlParameters = buildURLparams($urlParameters, $separator);
491    } else {
492        $urlParameters = str_replace(',', $separator, $urlParameters);
493    }
494    if($id === '') {
495        $id = $conf['start'];
496    }
497    $id = idfilter($id);
498    if($absolute) {
499        $xlink = DOKU_URL;
500    } else {
501        $xlink = DOKU_BASE;
502    }
503
504    if($conf['userewrite'] == 2) {
505        $xlink .= DOKU_SCRIPT.'/'.$id;
506        if($urlParameters) $xlink .= '?'.$urlParameters;
507    } elseif($conf['userewrite']) {
508        $xlink .= $id;
509        if($urlParameters) $xlink .= '?'.$urlParameters;
510    } elseif($id !== '') {
511        $xlink .= DOKU_SCRIPT.'?id='.$id;
512        if($urlParameters) $xlink .= $separator.$urlParameters;
513    } else {
514        $xlink .= DOKU_SCRIPT;
515        if($urlParameters) $xlink .= '?'.$urlParameters;
516    }
517
518    return $xlink;
519}
520
521/**
522 * This builds a link to an alternate page format
523 *
524 * Handles URL rewriting if enabled. Follows the style of wl().
525 *
526 * @author Ben Coburn <btcoburn@silicodon.net>
527 * @param string       $id             page id, defaults to start page
528 * @param string       $format         the export renderer to use
529 * @param string|array $urlParameters  URL parameters, associative array recommended
530 * @param bool         $abs            request an absolute URL instead of relative
531 * @param string       $sep            parameter separator
532 * @return string
533 */
534function exportlink($id = '', $format = 'raw', $urlParameters = '', $abs = false, $sep = '&amp;') {
535    global $conf;
536    if(is_array($urlParameters)) {
537        $urlParameters = buildURLparams($urlParameters, $sep);
538    } else {
539        $urlParameters = str_replace(',', $sep, $urlParameters);
540    }
541
542    $format = rawurlencode($format);
543    $id     = idfilter($id);
544    if($abs) {
545        $xlink = DOKU_URL;
546    } else {
547        $xlink = DOKU_BASE;
548    }
549
550    if($conf['userewrite'] == 2) {
551        $xlink .= DOKU_SCRIPT.'/'.$id.'?do=export_'.$format;
552        if($urlParameters) $xlink .= $sep.$urlParameters;
553    } elseif($conf['userewrite'] == 1) {
554        $xlink .= '_export/'.$format.'/'.$id;
555        if($urlParameters) $xlink .= '?'.$urlParameters;
556    } else {
557        $xlink .= DOKU_SCRIPT.'?do=export_'.$format.$sep.'id='.$id;
558        if($urlParameters) $xlink .= $sep.$urlParameters;
559    }
560
561    return $xlink;
562}
563
564/**
565 * Build a link to a media file
566 *
567 * Will return a link to the detail page if $direct is false
568 *
569 * The $more parameter should always be given as array, the function then
570 * will strip default parameters to produce even cleaner URLs
571 *
572 * @param string  $id     the media file id or URL
573 * @param mixed   $more   string or array with additional parameters
574 * @param bool    $direct link to detail page if false
575 * @param string  $sep    URL parameter separator
576 * @param bool    $abs    Create an absolute URL
577 * @return string
578 */
579function ml($id = '', $more = '', $direct = true, $sep = '&amp;', $abs = false) {
580    global $conf;
581    $isexternalimage = media_isexternal($id);
582    if(!$isexternalimage) {
583        $id = cleanID($id);
584    }
585
586    if(is_array($more)) {
587        // add token for resized images
588        $w = isset($more['w']) ? $more['w'] : null;
589        $h = isset($more['h']) ? $more['h'] : null;
590        if($w || $h || $isexternalimage){
591            $more['tok'] = media_get_token($id, $w, $h);
592        }
593        // strip defaults for shorter URLs
594        if(isset($more['cache']) && $more['cache'] == 'cache') unset($more['cache']);
595        if(empty($more['w'])) unset($more['w']);
596        if(empty($more['h'])) unset($more['h']);
597        if(isset($more['id']) && $direct) unset($more['id']);
598        if(isset($more['rev']) && !$more['rev']) unset($more['rev']);
599        $more = buildURLparams($more, $sep);
600    } else {
601        $matches = array();
602        if (preg_match_all('/\b(w|h)=(\d*)\b/',$more,$matches,PREG_SET_ORDER) || $isexternalimage){
603            $resize = array('w'=>0, 'h'=>0);
604            foreach ($matches as $match){
605                $resize[$match[1]] = $match[2];
606            }
607            $more .= $more === '' ? '' : $sep;
608            $more .= 'tok='.media_get_token($id,$resize['w'],$resize['h']);
609        }
610        $more = str_replace('cache=cache', '', $more); //skip default
611        $more = str_replace(',,', ',', $more);
612        $more = str_replace(',', $sep, $more);
613    }
614
615    if($abs) {
616        $xlink = DOKU_URL;
617    } else {
618        $xlink = DOKU_BASE;
619    }
620
621    // external URLs are always direct without rewriting
622    if($isexternalimage) {
623        $xlink .= 'lib/exe/fetch.php';
624        $xlink .= '?'.$more;
625        $xlink .= $sep.'media='.rawurlencode($id);
626        return $xlink;
627    }
628
629    $id = idfilter($id);
630
631    // decide on scriptname
632    if($direct) {
633        if($conf['userewrite'] == 1) {
634            $script = '_media';
635        } else {
636            $script = 'lib/exe/fetch.php';
637        }
638    } else {
639        if($conf['userewrite'] == 1) {
640            $script = '_detail';
641        } else {
642            $script = 'lib/exe/detail.php';
643        }
644    }
645
646    // build URL based on rewrite mode
647    if($conf['userewrite']) {
648        $xlink .= $script.'/'.$id;
649        if($more) $xlink .= '?'.$more;
650    } else {
651        if($more) {
652            $xlink .= $script.'?'.$more;
653            $xlink .= $sep.'media='.$id;
654        } else {
655            $xlink .= $script.'?media='.$id;
656        }
657    }
658
659    return $xlink;
660}
661
662/**
663 * Returns the URL to the DokuWiki base script
664 *
665 * Consider using wl() instead, unless you absoutely need the doku.php endpoint
666 *
667 * @author Andreas Gohr <andi@splitbrain.org>
668 *
669 * @return string
670 */
671function script() {
672    return DOKU_BASE.DOKU_SCRIPT;
673}
674
675/**
676 * Spamcheck against wordlist
677 *
678 * Checks the wikitext against a list of blocked expressions
679 * returns true if the text contains any bad words
680 *
681 * Triggers COMMON_WORDBLOCK_BLOCKED
682 *
683 *  Action Plugins can use this event to inspect the blocked data
684 *  and gain information about the user who was blocked.
685 *
686 *  Event data:
687 *    data['matches']  - array of matches
688 *    data['userinfo'] - information about the blocked user
689 *      [ip]           - ip address
690 *      [user]         - username (if logged in)
691 *      [mail]         - mail address (if logged in)
692 *      [name]         - real name (if logged in)
693 *
694 * @author Andreas Gohr <andi@splitbrain.org>
695 * @author Michael Klier <chi@chimeric.de>
696 *
697 * @param  string $text - optional text to check, if not given the globals are used
698 * @return bool         - true if a spam word was found
699 */
700function checkwordblock($text = '') {
701    global $TEXT;
702    global $PRE;
703    global $SUF;
704    global $SUM;
705    global $conf;
706    global $INFO;
707    /* @var Input $INPUT */
708    global $INPUT;
709
710    if(!$conf['usewordblock']) return false;
711
712    if(!$text) $text = "$PRE $TEXT $SUF $SUM";
713
714    // we prepare the text a tiny bit to prevent spammers circumventing URL checks
715    // phpcs:disable Generic.Files.LineLength.TooLong
716    $text = preg_replace(
717        '!(\b)(www\.[\w.:?\-;,]+?\.[\w.:?\-;,]+?[\w/\#~:.?+=&%@\!\-.:?\-;,]+?)([.:?\-;,]*[^\w/\#~:.?+=&%@\!\-.:?\-;,])!i',
718        '\1http://\2 \2\3',
719        $text
720    );
721    // phpcs:enable
722
723    $wordblocks = getWordblocks();
724    // how many lines to read at once (to work around some PCRE limits)
725    if(version_compare(phpversion(), '4.3.0', '<')) {
726        // old versions of PCRE define a maximum of parenthesises even if no
727        // backreferences are used - the maximum is 99
728        // this is very bad performancewise and may even be too high still
729        $chunksize = 40;
730    } else {
731        // read file in chunks of 200 - this should work around the
732        // MAX_PATTERN_SIZE in modern PCRE
733        $chunksize = 200;
734    }
735    while($blocks = array_splice($wordblocks, 0, $chunksize)) {
736        $re = array();
737        // build regexp from blocks
738        foreach($blocks as $block) {
739            $block = preg_replace('/#.*$/', '', $block);
740            $block = trim($block);
741            if(empty($block)) continue;
742            $re[] = $block;
743        }
744        if(count($re) && preg_match('#('.join('|', $re).')#si', $text, $matches)) {
745            // prepare event data
746            $data = array();
747            $data['matches']        = $matches;
748            $data['userinfo']['ip'] = $INPUT->server->str('REMOTE_ADDR');
749            if($INPUT->server->str('REMOTE_USER')) {
750                $data['userinfo']['user'] = $INPUT->server->str('REMOTE_USER');
751                $data['userinfo']['name'] = $INFO['userinfo']['name'];
752                $data['userinfo']['mail'] = $INFO['userinfo']['mail'];
753            }
754            $callback = function () {
755                return true;
756            };
757            return Event::createAndTrigger('COMMON_WORDBLOCK_BLOCKED', $data, $callback, true);
758        }
759    }
760    return false;
761}
762
763/**
764 * Return the IP of the client
765 *
766 * Honours X-Forwarded-For and X-Real-IP Proxy Headers
767 *
768 * It returns a comma separated list of IPs if the above mentioned
769 * headers are set. If the single parameter is set, it tries to return
770 * a routable public address, prefering the ones suplied in the X
771 * headers
772 *
773 * @author Andreas Gohr <andi@splitbrain.org>
774 *
775 * @param  boolean $single If set only a single IP is returned
776 * @return string
777 */
778function clientIP($single = false) {
779    /* @var Input $INPUT */
780    global $INPUT, $conf;
781
782    $ip   = array();
783    $ip[] = $INPUT->server->str('REMOTE_ADDR');
784    if($INPUT->server->str('HTTP_X_FORWARDED_FOR')) {
785        $ip = array_merge($ip, explode(',', str_replace(' ', '', $INPUT->server->str('HTTP_X_FORWARDED_FOR'))));
786    }
787    if($INPUT->server->str('HTTP_X_REAL_IP')) {
788        $ip = array_merge($ip, explode(',', str_replace(' ', '', $INPUT->server->str('HTTP_X_REAL_IP'))));
789    }
790
791    // some IPv4/v6 regexps borrowed from Feyd
792    // see: http://forums.devnetwork.net/viewtopic.php?f=38&t=53479
793    $dec_octet   = '(?:25[0-5]|2[0-4]\d|1\d\d|[1-9]\d|[0-9])';
794    $hex_digit   = '[A-Fa-f0-9]';
795    $h16         = "{$hex_digit}{1,4}";
796    $IPv4Address = "$dec_octet\\.$dec_octet\\.$dec_octet\\.$dec_octet";
797    $ls32        = "(?:$h16:$h16|$IPv4Address)";
798    $IPv6Address =
799        "(?:(?:{$IPv4Address})|(?:".
800            "(?:$h16:){6}$ls32".
801            "|::(?:$h16:){5}$ls32".
802            "|(?:$h16)?::(?:$h16:){4}$ls32".
803            "|(?:(?:$h16:){0,1}$h16)?::(?:$h16:){3}$ls32".
804            "|(?:(?:$h16:){0,2}$h16)?::(?:$h16:){2}$ls32".
805            "|(?:(?:$h16:){0,3}$h16)?::(?:$h16:){1}$ls32".
806            "|(?:(?:$h16:){0,4}$h16)?::$ls32".
807            "|(?:(?:$h16:){0,5}$h16)?::$h16".
808            "|(?:(?:$h16:){0,6}$h16)?::".
809            ")(?:\\/(?:12[0-8]|1[0-1][0-9]|[1-9][0-9]|[0-9]))?)";
810
811    // remove any non-IP stuff
812    $cnt   = count($ip);
813    $match = array();
814    for($i = 0; $i < $cnt; $i++) {
815        if(preg_match("/^$IPv4Address$/", $ip[$i], $match) || preg_match("/^$IPv6Address$/", $ip[$i], $match)) {
816            $ip[$i] = $match[0];
817        } else {
818            $ip[$i] = '';
819        }
820        if(empty($ip[$i])) unset($ip[$i]);
821    }
822    $ip = array_values(array_unique($ip));
823    if(empty($ip) || !$ip[0]) $ip[0] = '0.0.0.0'; // for some strange reason we don't have a IP
824
825    if(!$single) return join(',', $ip);
826
827    // skip trusted local addresses
828    foreach($ip as $i) {
829        if(!empty($conf['trustedproxy']) && preg_match('/'.$conf['trustedproxy'].'/', $i)) {
830            continue;
831        } else {
832            return $i;
833        }
834    }
835
836    // still here? just use the last address
837    // this case all ips in the list are trusted
838    return $ip[count($ip)-1];
839}
840
841/**
842 * Check if the browser is on a mobile device
843 *
844 * Adapted from the example code at url below
845 *
846 * @link http://www.brainhandles.com/2007/10/15/detecting-mobile-browsers/#code
847 *
848 * @deprecated 2018-04-27 you probably want media queries instead anyway
849 * @return bool if true, client is mobile browser; otherwise false
850 */
851function clientismobile() {
852    /* @var Input $INPUT */
853    global $INPUT;
854
855    if($INPUT->server->has('HTTP_X_WAP_PROFILE')) return true;
856
857    if(preg_match('/wap\.|\.wap/i', $INPUT->server->str('HTTP_ACCEPT'))) return true;
858
859    if(!$INPUT->server->has('HTTP_USER_AGENT')) return false;
860
861    $uamatches = join(
862        '|',
863        [
864            'midp', 'j2me', 'avantg', 'docomo', 'novarra', 'palmos', 'palmsource', '240x320', 'opwv',
865            'chtml', 'pda', 'windows ce', 'mmp\/', 'blackberry', 'mib\/', 'symbian', 'wireless', 'nokia',
866            'hand', 'mobi', 'phone', 'cdm', 'up\.b', 'audio', 'SIE\-', 'SEC\-', 'samsung', 'HTC', 'mot\-',
867            'mitsu', 'sagem', 'sony', 'alcatel', 'lg', 'erics', 'vx', 'NEC', 'philips', 'mmm', 'xx',
868            'panasonic', 'sharp', 'wap', 'sch', 'rover', 'pocket', 'benq', 'java', 'pt', 'pg', 'vox',
869            'amoi', 'bird', 'compal', 'kg', 'voda', 'sany', 'kdd', 'dbt', 'sendo', 'sgh', 'gradi', 'jb',
870            '\d\d\di', 'moto'
871        ]
872    );
873
874    if(preg_match("/$uamatches/i", $INPUT->server->str('HTTP_USER_AGENT'))) return true;
875
876    return false;
877}
878
879/**
880 * check if a given link is interwiki link
881 *
882 * @param string $link the link, e.g. "wiki>page"
883 * @return bool
884 */
885function link_isinterwiki($link){
886    if (preg_match('/^[a-zA-Z0-9\.]+>/u',$link)) return true;
887    return false;
888}
889
890/**
891 * Convert one or more comma separated IPs to hostnames
892 *
893 * If $conf['dnslookups'] is disabled it simply returns the input string
894 *
895 * @author Glen Harris <astfgl@iamnota.org>
896 *
897 * @param  string $ips comma separated list of IP addresses
898 * @return string a comma separated list of hostnames
899 */
900function gethostsbyaddrs($ips) {
901    global $conf;
902    if(!$conf['dnslookups']) return $ips;
903
904    $hosts = array();
905    $ips   = explode(',', $ips);
906
907    if(is_array($ips)) {
908        foreach($ips as $ip) {
909            $hosts[] = gethostbyaddr(trim($ip));
910        }
911        return join(',', $hosts);
912    } else {
913        return gethostbyaddr(trim($ips));
914    }
915}
916
917/**
918 * Checks if a given page is currently locked.
919 *
920 * removes stale lockfiles
921 *
922 * @author Andreas Gohr <andi@splitbrain.org>
923 *
924 * @param string $id page id
925 * @return bool page is locked?
926 */
927function checklock($id) {
928    global $conf;
929    /* @var Input $INPUT */
930    global $INPUT;
931
932    $lock = wikiLockFN($id);
933
934    //no lockfile
935    if(!file_exists($lock)) return false;
936
937    //lockfile expired
938    if((time() - filemtime($lock)) > $conf['locktime']) {
939        @unlink($lock);
940        return false;
941    }
942
943    //my own lock
944    @list($ip, $session) = explode("\n", io_readFile($lock));
945    if($ip == $INPUT->server->str('REMOTE_USER') || $ip == clientIP() || (session_id() && $session == session_id())) {
946        return false;
947    }
948
949    return $ip;
950}
951
952/**
953 * Lock a page for editing
954 *
955 * @author Andreas Gohr <andi@splitbrain.org>
956 *
957 * @param string $id page id to lock
958 */
959function lock($id) {
960    global $conf;
961    /* @var Input $INPUT */
962    global $INPUT;
963
964    if($conf['locktime'] == 0) {
965        return;
966    }
967
968    $lock = wikiLockFN($id);
969    if($INPUT->server->str('REMOTE_USER')) {
970        io_saveFile($lock, $INPUT->server->str('REMOTE_USER'));
971    } else {
972        io_saveFile($lock, clientIP()."\n".session_id());
973    }
974}
975
976/**
977 * Unlock a page if it was locked by the user
978 *
979 * @author Andreas Gohr <andi@splitbrain.org>
980 *
981 * @param string $id page id to unlock
982 * @return bool true if a lock was removed
983 */
984function unlock($id) {
985    /* @var Input $INPUT */
986    global $INPUT;
987
988    $lock = wikiLockFN($id);
989    if(file_exists($lock)) {
990        @list($ip, $session) = explode("\n", io_readFile($lock));
991        if($ip == $INPUT->server->str('REMOTE_USER') || $ip == clientIP() || $session == session_id()) {
992            @unlink($lock);
993            return true;
994        }
995    }
996    return false;
997}
998
999/**
1000 * convert line ending to unix format
1001 *
1002 * also makes sure the given text is valid UTF-8
1003 *
1004 * @see    formText() for 2crlf conversion
1005 * @author Andreas Gohr <andi@splitbrain.org>
1006 *
1007 * @param string $text
1008 * @return string
1009 */
1010function cleanText($text) {
1011    $text = preg_replace("/(\015\012)|(\015)/", "\012", $text);
1012
1013    // if the text is not valid UTF-8 we simply assume latin1
1014    // this won't break any worse than it breaks with the wrong encoding
1015    // but might actually fix the problem in many cases
1016    if(!\dokuwiki\Utf8\Clean::isUtf8($text)) $text = utf8_encode($text);
1017
1018    return $text;
1019}
1020
1021/**
1022 * Prepares text for print in Webforms by encoding special chars.
1023 * It also converts line endings to Windows format which is
1024 * pseudo standard for webforms.
1025 *
1026 * @see    cleanText() for 2unix conversion
1027 * @author Andreas Gohr <andi@splitbrain.org>
1028 *
1029 * @param string $text
1030 * @return string
1031 */
1032function formText($text) {
1033    $text = str_replace("\012", "\015\012", $text);
1034    return htmlspecialchars($text);
1035}
1036
1037/**
1038 * Returns the specified local text in raw format
1039 *
1040 * @author Andreas Gohr <andi@splitbrain.org>
1041 *
1042 * @param string $id   page id
1043 * @param string $ext  extension of file being read, default 'txt'
1044 * @return string
1045 */
1046function rawLocale($id, $ext = 'txt') {
1047    return io_readFile(localeFN($id, $ext));
1048}
1049
1050/**
1051 * Returns the raw WikiText
1052 *
1053 * @author Andreas Gohr <andi@splitbrain.org>
1054 *
1055 * @param string $id   page id
1056 * @param string|int $rev  timestamp when a revision of wikitext is desired
1057 * @return string
1058 */
1059function rawWiki($id, $rev = '') {
1060    return io_readWikiPage(wikiFN($id, $rev), $id, $rev);
1061}
1062
1063/**
1064 * Returns the pagetemplate contents for the ID's namespace
1065 *
1066 * @triggers COMMON_PAGETPL_LOAD
1067 * @author Andreas Gohr <andi@splitbrain.org>
1068 *
1069 * @param string $id the id of the page to be created
1070 * @return string parsed pagetemplate content
1071 */
1072function pageTemplate($id) {
1073    global $conf;
1074
1075    if(is_array($id)) $id = $id[0];
1076
1077    // prepare initial event data
1078    $data = array(
1079        'id'        => $id, // the id of the page to be created
1080        'tpl'       => '', // the text used as template
1081        'tplfile'   => '', // the file above text was/should be loaded from
1082        'doreplace' => true // should wildcard replacements be done on the text?
1083    );
1084
1085    $evt = new Event('COMMON_PAGETPL_LOAD', $data);
1086    if($evt->advise_before(true)) {
1087        // the before event might have loaded the content already
1088        if(empty($data['tpl'])) {
1089            // if the before event did not set a template file, try to find one
1090            if(empty($data['tplfile'])) {
1091                $path = dirname(wikiFN($id));
1092                if(file_exists($path.'/_template.txt')) {
1093                    $data['tplfile'] = $path.'/_template.txt';
1094                } else {
1095                    // search upper namespaces for templates
1096                    $len = strlen(rtrim($conf['datadir'], '/'));
1097                    while(strlen($path) >= $len) {
1098                        if(file_exists($path.'/__template.txt')) {
1099                            $data['tplfile'] = $path.'/__template.txt';
1100                            break;
1101                        }
1102                        $path = substr($path, 0, strrpos($path, '/'));
1103                    }
1104                }
1105            }
1106            // load the content
1107            $data['tpl'] = io_readFile($data['tplfile']);
1108        }
1109        if($data['doreplace']) parsePageTemplate($data);
1110    }
1111    $evt->advise_after();
1112    unset($evt);
1113
1114    return $data['tpl'];
1115}
1116
1117/**
1118 * Performs common page template replacements
1119 * This works on data from COMMON_PAGETPL_LOAD
1120 *
1121 * @author Andreas Gohr <andi@splitbrain.org>
1122 *
1123 * @param array $data array with event data
1124 * @return string
1125 */
1126function parsePageTemplate(&$data) {
1127    /**
1128     * @var string $id        the id of the page to be created
1129     * @var string $tpl       the text used as template
1130     * @var string $tplfile   the file above text was/should be loaded from
1131     * @var bool   $doreplace should wildcard replacements be done on the text?
1132     */
1133    extract($data);
1134
1135    global $USERINFO;
1136    global $conf;
1137    /* @var Input $INPUT */
1138    global $INPUT;
1139
1140    // replace placeholders
1141    $file = noNS($id);
1142    $page = strtr($file, $conf['sepchar'], ' ');
1143
1144    $tpl = str_replace(
1145        array(
1146             '@ID@',
1147             '@NS@',
1148             '@CURNS@',
1149             '@!CURNS@',
1150             '@!!CURNS@',
1151             '@!CURNS!@',
1152             '@FILE@',
1153             '@!FILE@',
1154             '@!FILE!@',
1155             '@PAGE@',
1156             '@!PAGE@',
1157             '@!!PAGE@',
1158             '@!PAGE!@',
1159             '@USER@',
1160             '@NAME@',
1161             '@MAIL@',
1162             '@DATE@',
1163        ),
1164        array(
1165             $id,
1166             getNS($id),
1167             curNS($id),
1168             \dokuwiki\Utf8\PhpString::ucfirst(curNS($id)),
1169             \dokuwiki\Utf8\PhpString::ucwords(curNS($id)),
1170             \dokuwiki\Utf8\PhpString::strtoupper(curNS($id)),
1171             $file,
1172             \dokuwiki\Utf8\PhpString::ucfirst($file),
1173             \dokuwiki\Utf8\PhpString::strtoupper($file),
1174             $page,
1175             \dokuwiki\Utf8\PhpString::ucfirst($page),
1176             \dokuwiki\Utf8\PhpString::ucwords($page),
1177             \dokuwiki\Utf8\PhpString::strtoupper($page),
1178             $INPUT->server->str('REMOTE_USER'),
1179             $USERINFO ? $USERINFO['name'] : '',
1180             $USERINFO ? $USERINFO['mail'] : '',
1181             $conf['dformat'],
1182        ), $tpl
1183    );
1184
1185    // we need the callback to work around strftime's char limit
1186    $tpl = preg_replace_callback(
1187        '/%./',
1188        function ($m) {
1189            return strftime($m[0]);
1190        },
1191        $tpl
1192    );
1193    $data['tpl'] = $tpl;
1194    return $tpl;
1195}
1196
1197/**
1198 * Returns the raw Wiki Text in three slices.
1199 *
1200 * The range parameter needs to have the form "from-to"
1201 * and gives the range of the section in bytes - no
1202 * UTF-8 awareness is needed.
1203 * The returned order is prefix, section and suffix.
1204 *
1205 * @author Andreas Gohr <andi@splitbrain.org>
1206 *
1207 * @param string $range in form "from-to"
1208 * @param string $id    page id
1209 * @param string $rev   optional, the revision timestamp
1210 * @return string[] with three slices
1211 */
1212function rawWikiSlices($range, $id, $rev = '') {
1213    $text = io_readWikiPage(wikiFN($id, $rev), $id, $rev);
1214
1215    // Parse range
1216    list($from, $to) = explode('-', $range, 2);
1217    // Make range zero-based, use defaults if marker is missing
1218    $from = !$from ? 0 : ($from - 1);
1219    $to   = !$to ? strlen($text) : ($to - 1);
1220
1221    $slices = array();
1222    $slices[0] = substr($text, 0, $from);
1223    $slices[1] = substr($text, $from, $to - $from);
1224    $slices[2] = substr($text, $to);
1225    return $slices;
1226}
1227
1228/**
1229 * Joins wiki text slices
1230 *
1231 * function to join the text slices.
1232 * When the pretty parameter is set to true it adds additional empty
1233 * lines between sections if needed (used on saving).
1234 *
1235 * @author Andreas Gohr <andi@splitbrain.org>
1236 *
1237 * @param string $pre   prefix
1238 * @param string $text  text in the middle
1239 * @param string $suf   suffix
1240 * @param bool $pretty add additional empty lines between sections
1241 * @return string
1242 */
1243function con($pre, $text, $suf, $pretty = false) {
1244    if($pretty) {
1245        if($pre !== '' && substr($pre, -1) !== "\n" &&
1246            substr($text, 0, 1) !== "\n"
1247        ) {
1248            $pre .= "\n";
1249        }
1250        if($suf !== '' && substr($text, -1) !== "\n" &&
1251            substr($suf, 0, 1) !== "\n"
1252        ) {
1253            $text .= "\n";
1254        }
1255    }
1256
1257    return $pre.$text.$suf;
1258}
1259
1260/**
1261 * Checks if the current page version is newer than the last entry in the page's
1262 * changelog. If so, we assume it has been an external edit and we create an
1263 * attic copy and add a proper changelog line.
1264 *
1265 * This check is only executed when the page is about to be saved again from the
1266 * wiki, triggered in @see saveWikiText()
1267 *
1268 * @param string $id the page ID
1269 * @deprecated YYYY-MM-DD
1270 */
1271function detectExternalEdit($id) {
1272  //dbg_deprecated(\dokuwiki\File\PageFile::class .'::detectExternalEdit()');
1273    (new PageFile($id))->detectExternalEdit();
1274}
1275
1276/**
1277 * Saves a wikitext by calling io_writeWikiPage.
1278 * Also directs changelog and attic updates.
1279 *
1280 * @author Andreas Gohr <andi@splitbrain.org>
1281 * @author Ben Coburn <btcoburn@silicodon.net>
1282 *
1283 * @param string $id       page id
1284 * @param string $text     wikitext being saved
1285 * @param string $summary  summary of text update
1286 * @param bool   $minor    mark this saved version as minor update
1287 */
1288function saveWikiText($id, $text, $summary, $minor = false) {
1289
1290    // get COMMON_WIKIPAGE_SAVE event data
1291    $data = (new PageFile($id))->saveWikiText($text, $summary, $minor);
1292
1293    // send notify mails
1294    notify($id, 'admin', $data['oldRevision'], $data['summary'], $minor, $data['newRevision']);
1295    notify($id, 'subscribers', $data['oldRevision'], $data['summary'], $minor, $data['newRevision']);
1296
1297    // if useheading is enabled, purge the cache of all linking pages
1298    if (useHeading('content')) {
1299        $pages = ft_backlinks($id, true);
1300        foreach ($pages as $page) {
1301            $cache = new CacheRenderer($page, wikiFN($page), 'xhtml');
1302            $cache->removeCache();
1303        }
1304    }
1305}
1306
1307/**
1308 * moves the current version to the attic and returns its revision date
1309 *
1310 * @author Andreas Gohr <andi@splitbrain.org>
1311 *
1312 * @param string $id page id
1313 * @return int|string revision timestamp
1314 * @deprecated YYYY-MM-DD
1315 */
1316function saveOldRevision($id) {
1317  //dbg_deprecated(\dokuwiki\File\PageFile::class .'::saveOldRevision()');
1318    return (new PageFile($id))->saveOldRevision();
1319}
1320
1321/**
1322 * Sends a notify mail on page change or registration
1323 *
1324 * @param string     $id       The changed page
1325 * @param string     $who      Who to notify (admin|subscribers|register)
1326 * @param int|string $rev Old page revision
1327 * @param string     $summary  What changed
1328 * @param boolean    $minor    Is this a minor edit?
1329 * @param string[]   $replace  Additional string substitutions, @KEY@ to be replaced by value
1330 * @param int|string $current_rev  New page revision
1331 * @return bool
1332 *
1333 * @author Andreas Gohr <andi@splitbrain.org>
1334 */
1335function notify($id, $who, $rev = '', $summary = '', $minor = false, $replace = array(), $current_rev = false) {
1336    global $conf;
1337    /* @var Input $INPUT */
1338    global $INPUT;
1339
1340    // decide if there is something to do, eg. whom to mail
1341    if($who == 'admin') {
1342        if(empty($conf['notify'])) return false; //notify enabled?
1343        $tpl = 'mailtext';
1344        $to  = $conf['notify'];
1345    } elseif($who == 'subscribers') {
1346        if(!actionOK('subscribe')) return false; //subscribers enabled?
1347        if($conf['useacl'] && $INPUT->server->str('REMOTE_USER') && $minor) return false; //skip minors
1348        $data = array('id' => $id, 'addresslist' => '', 'self' => false, 'replacements' => $replace);
1349        Event::createAndTrigger(
1350            'COMMON_NOTIFY_ADDRESSLIST', $data,
1351            array(new SubscriberManager(), 'notifyAddresses')
1352        );
1353        $to = $data['addresslist'];
1354        if(empty($to)) return false;
1355        $tpl = 'subscr_single';
1356    } else {
1357        return false; //just to be safe
1358    }
1359
1360    // prepare content
1361    $subscription = new PageSubscriptionSender();
1362    return $subscription->sendPageDiff($to, $tpl, $id, $rev, $summary, $current_rev);
1363}
1364
1365/**
1366 * extracts the query from a search engine referrer
1367 *
1368 * @author Andreas Gohr <andi@splitbrain.org>
1369 * @author Todd Augsburger <todd@rollerorgans.com>
1370 *
1371 * @return array|string
1372 */
1373function getGoogleQuery() {
1374    /* @var Input $INPUT */
1375    global $INPUT;
1376
1377    if(!$INPUT->server->has('HTTP_REFERER')) {
1378        return '';
1379    }
1380    $url = parse_url($INPUT->server->str('HTTP_REFERER'));
1381
1382    // only handle common SEs
1383    if(!preg_match('/(google|bing|yahoo|ask|duckduckgo|babylon|aol|yandex)/',$url['host'])) return '';
1384
1385    $query = array();
1386    parse_str($url['query'], $query);
1387
1388    $q = '';
1389    if(isset($query['q'])){
1390        $q = $query['q'];
1391    }elseif(isset($query['p'])){
1392        $q = $query['p'];
1393    }elseif(isset($query['query'])){
1394        $q = $query['query'];
1395    }
1396    $q = trim($q);
1397
1398    if(!$q) return '';
1399    // ignore if query includes a full URL
1400    if(strpos($q, '//') !== false) return '';
1401    $q = preg_split('/[\s\'"\\\\`()\]\[?:!\.{};,#+*<>\\/]+/', $q, -1, PREG_SPLIT_NO_EMPTY);
1402    return $q;
1403}
1404
1405/**
1406 * Return the human readable size of a file
1407 *
1408 * @param int $size A file size
1409 * @param int $dec A number of decimal places
1410 * @return string human readable size
1411 *
1412 * @author      Martin Benjamin <b.martin@cybernet.ch>
1413 * @author      Aidan Lister <aidan@php.net>
1414 * @version     1.0.0
1415 */
1416function filesize_h($size, $dec = 1) {
1417    $sizes = array('B', 'KB', 'MB', 'GB');
1418    $count = count($sizes);
1419    $i     = 0;
1420
1421    while($size >= 1024 && ($i < $count - 1)) {
1422        $size /= 1024;
1423        $i++;
1424    }
1425
1426    return round($size, $dec)."\xC2\xA0".$sizes[$i]; //non-breaking space
1427}
1428
1429/**
1430 * Return the given timestamp as human readable, fuzzy age
1431 *
1432 * @author Andreas Gohr <gohr@cosmocode.de>
1433 *
1434 * @param int $dt timestamp
1435 * @return string
1436 */
1437function datetime_h($dt) {
1438    global $lang;
1439
1440    $ago = time() - $dt;
1441    if($ago > 24 * 60 * 60 * 30 * 12 * 2) {
1442        return sprintf($lang['years'], round($ago / (24 * 60 * 60 * 30 * 12)));
1443    }
1444    if($ago > 24 * 60 * 60 * 30 * 2) {
1445        return sprintf($lang['months'], round($ago / (24 * 60 * 60 * 30)));
1446    }
1447    if($ago > 24 * 60 * 60 * 7 * 2) {
1448        return sprintf($lang['weeks'], round($ago / (24 * 60 * 60 * 7)));
1449    }
1450    if($ago > 24 * 60 * 60 * 2) {
1451        return sprintf($lang['days'], round($ago / (24 * 60 * 60)));
1452    }
1453    if($ago > 60 * 60 * 2) {
1454        return sprintf($lang['hours'], round($ago / (60 * 60)));
1455    }
1456    if($ago > 60 * 2) {
1457        return sprintf($lang['minutes'], round($ago / (60)));
1458    }
1459    return sprintf($lang['seconds'], $ago);
1460}
1461
1462/**
1463 * Wraps around strftime but provides support for fuzzy dates
1464 *
1465 * The format default to $conf['dformat']. It is passed to
1466 * strftime - %f can be used to get the value from datetime_h()
1467 *
1468 * @see datetime_h
1469 * @author Andreas Gohr <gohr@cosmocode.de>
1470 *
1471 * @param int|null $dt      timestamp when given, null will take current timestamp
1472 * @param string   $format  empty default to $conf['dformat'], or provide format as recognized by strftime()
1473 * @return string
1474 */
1475function dformat($dt = null, $format = '') {
1476    global $conf;
1477
1478    if(is_null($dt)) $dt = time();
1479    $dt = (int) $dt;
1480    if(!$format) $format = $conf['dformat'];
1481
1482    $format = str_replace('%f', datetime_h($dt), $format);
1483    return strftime($format, $dt);
1484}
1485
1486/**
1487 * Formats a timestamp as ISO 8601 date
1488 *
1489 * @author <ungu at terong dot com>
1490 * @link http://php.net/manual/en/function.date.php#54072
1491 *
1492 * @param int $int_date current date in UNIX timestamp
1493 * @return string
1494 */
1495function date_iso8601($int_date) {
1496    $date_mod     = date('Y-m-d\TH:i:s', $int_date);
1497    $pre_timezone = date('O', $int_date);
1498    $time_zone    = substr($pre_timezone, 0, 3).":".substr($pre_timezone, 3, 2);
1499    $date_mod .= $time_zone;
1500    return $date_mod;
1501}
1502
1503/**
1504 * return an obfuscated email address in line with $conf['mailguard'] setting
1505 *
1506 * @author Harry Fuecks <hfuecks@gmail.com>
1507 * @author Christopher Smith <chris@jalakai.co.uk>
1508 *
1509 * @param string $email email address
1510 * @return string
1511 */
1512function obfuscate($email) {
1513    global $conf;
1514
1515    switch($conf['mailguard']) {
1516        case 'visible' :
1517            $obfuscate = array('@' => ' [at] ', '.' => ' [dot] ', '-' => ' [dash] ');
1518            return strtr($email, $obfuscate);
1519
1520        case 'hex' :
1521            return \dokuwiki\Utf8\Conversion::toHtml($email, true);
1522
1523        case 'none' :
1524        default :
1525            return $email;
1526    }
1527}
1528
1529/**
1530 * Removes quoting backslashes
1531 *
1532 * @author Andreas Gohr <andi@splitbrain.org>
1533 *
1534 * @param string $string
1535 * @param string $char backslashed character
1536 * @return string
1537 */
1538function unslash($string, $char = "'") {
1539    return str_replace('\\'.$char, $char, $string);
1540}
1541
1542/**
1543 * Convert php.ini shorthands to byte
1544 *
1545 * On 32 bit systems values >= 2GB will fail!
1546 *
1547 * -1 (infinite size) will be reported as -1
1548 *
1549 * @link   https://www.php.net/manual/en/faq.using.php#faq.using.shorthandbytes
1550 * @param string $value PHP size shorthand
1551 * @return int
1552 */
1553function php_to_byte($value) {
1554    switch (strtoupper(substr($value,-1))) {
1555        case 'G':
1556            $ret = intval(substr($value, 0, -1)) * 1024 * 1024 * 1024;
1557            break;
1558        case 'M':
1559            $ret = intval(substr($value, 0, -1)) * 1024 * 1024;
1560            break;
1561        case 'K':
1562            $ret = intval(substr($value, 0, -1)) * 1024;
1563            break;
1564        default:
1565            $ret = intval($value);
1566            break;
1567    }
1568    return $ret;
1569}
1570
1571/**
1572 * Wrapper around preg_quote adding the default delimiter
1573 *
1574 * @param string $string
1575 * @return string
1576 */
1577function preg_quote_cb($string) {
1578    return preg_quote($string, '/');
1579}
1580
1581/**
1582 * Shorten a given string by removing data from the middle
1583 *
1584 * You can give the string in two parts, the first part $keep
1585 * will never be shortened. The second part $short will be cut
1586 * in the middle to shorten but only if at least $min chars are
1587 * left to display it. Otherwise it will be left off.
1588 *
1589 * @param string $keep   the part to keep
1590 * @param string $short  the part to shorten
1591 * @param int    $max    maximum chars you want for the whole string
1592 * @param int    $min    minimum number of chars to have left for middle shortening
1593 * @param string $char   the shortening character to use
1594 * @return string
1595 */
1596function shorten($keep, $short, $max, $min = 9, $char = '…') {
1597    $max = $max - \dokuwiki\Utf8\PhpString::strlen($keep);
1598    if($max < $min) return $keep;
1599    $len = \dokuwiki\Utf8\PhpString::strlen($short);
1600    if($len <= $max) return $keep.$short;
1601    $half = floor($max / 2);
1602    return $keep .
1603        \dokuwiki\Utf8\PhpString::substr($short, 0, $half - 1) .
1604        $char .
1605        \dokuwiki\Utf8\PhpString::substr($short, $len - $half);
1606}
1607
1608/**
1609 * Return the users real name or e-mail address for use
1610 * in page footer and recent changes pages
1611 *
1612 * @param string|null $username or null when currently logged-in user should be used
1613 * @param bool $textonly true returns only plain text, true allows returning html
1614 * @return string html or plain text(not escaped) of formatted user name
1615 *
1616 * @author Andy Webber <dokuwiki AT andywebber DOT com>
1617 */
1618function editorinfo($username, $textonly = false) {
1619    return userlink($username, $textonly);
1620}
1621
1622/**
1623 * Returns users realname w/o link
1624 *
1625 * @param string|null $username or null when currently logged-in user should be used
1626 * @param bool $textonly true returns only plain text, true allows returning html
1627 * @return string html or plain text(not escaped) of formatted user name
1628 *
1629 * @triggers COMMON_USER_LINK
1630 */
1631function userlink($username = null, $textonly = false) {
1632    global $conf, $INFO;
1633    /** @var AuthPlugin $auth */
1634    global $auth;
1635    /** @var Input $INPUT */
1636    global $INPUT;
1637
1638    // prepare initial event data
1639    $data = array(
1640        'username' => $username, // the unique user name
1641        'name' => '',
1642        'link' => array( //setting 'link' to false disables linking
1643                         'target' => '',
1644                         'pre' => '',
1645                         'suf' => '',
1646                         'style' => '',
1647                         'more' => '',
1648                         'url' => '',
1649                         'title' => '',
1650                         'class' => ''
1651        ),
1652        'userlink' => '', // formatted user name as will be returned
1653        'textonly' => $textonly
1654    );
1655    if($username === null) {
1656        $data['username'] = $username = $INPUT->server->str('REMOTE_USER');
1657        if($textonly){
1658            $data['name'] = $INFO['userinfo']['name']. ' (' . $INPUT->server->str('REMOTE_USER') . ')';
1659        }else {
1660            $data['name'] = '<bdi>' . hsc($INFO['userinfo']['name']) . '</bdi> '.
1661                '(<bdi>' . hsc($INPUT->server->str('REMOTE_USER')) . '</bdi>)';
1662        }
1663    }
1664
1665    $evt = new Event('COMMON_USER_LINK', $data);
1666    if($evt->advise_before(true)) {
1667        if(empty($data['name'])) {
1668            if($auth) $info = $auth->getUserData($username);
1669            if($conf['showuseras'] != 'loginname' && isset($info) && $info) {
1670                switch($conf['showuseras']) {
1671                    case 'username':
1672                    case 'username_link':
1673                        $data['name'] = $textonly ? $info['name'] : hsc($info['name']);
1674                        break;
1675                    case 'email':
1676                    case 'email_link':
1677                        $data['name'] = obfuscate($info['mail']);
1678                        break;
1679                }
1680            } else {
1681                $data['name'] = $textonly ? $data['username'] : hsc($data['username']);
1682            }
1683        }
1684
1685        /** @var Doku_Renderer_xhtml $xhtml_renderer */
1686        static $xhtml_renderer = null;
1687
1688        if(!$data['textonly'] && empty($data['link']['url'])) {
1689
1690            if(in_array($conf['showuseras'], array('email_link', 'username_link'))) {
1691                if(!isset($info)) {
1692                    if($auth) $info = $auth->getUserData($username);
1693                }
1694                if(isset($info) && $info) {
1695                    if($conf['showuseras'] == 'email_link') {
1696                        $data['link']['url'] = 'mailto:' . obfuscate($info['mail']);
1697                    } else {
1698                        if(is_null($xhtml_renderer)) {
1699                            $xhtml_renderer = p_get_renderer('xhtml');
1700                        }
1701                        if(empty($xhtml_renderer->interwiki)) {
1702                            $xhtml_renderer->interwiki = getInterwiki();
1703                        }
1704                        $shortcut = 'user';
1705                        $exists = null;
1706                        $data['link']['url'] = $xhtml_renderer->_resolveInterWiki($shortcut, $username, $exists);
1707                        $data['link']['class'] .= ' interwiki iw_user';
1708                        if($exists !== null) {
1709                            if($exists) {
1710                                $data['link']['class'] .= ' wikilink1';
1711                            } else {
1712                                $data['link']['class'] .= ' wikilink2';
1713                                $data['link']['rel'] = 'nofollow';
1714                            }
1715                        }
1716                    }
1717                } else {
1718                    $data['textonly'] = true;
1719                }
1720
1721            } else {
1722                $data['textonly'] = true;
1723            }
1724        }
1725
1726        if($data['textonly']) {
1727            $data['userlink'] = $data['name'];
1728        } else {
1729            $data['link']['name'] = $data['name'];
1730            if(is_null($xhtml_renderer)) {
1731                $xhtml_renderer = p_get_renderer('xhtml');
1732            }
1733            $data['userlink'] = $xhtml_renderer->_formatLink($data['link']);
1734        }
1735    }
1736    $evt->advise_after();
1737    unset($evt);
1738
1739    return $data['userlink'];
1740}
1741
1742/**
1743 * Returns the path to a image file for the currently chosen license.
1744 * When no image exists, returns an empty string
1745 *
1746 * @author Andreas Gohr <andi@splitbrain.org>
1747 *
1748 * @param  string $type - type of image 'badge' or 'button'
1749 * @return string
1750 */
1751function license_img($type) {
1752    global $license;
1753    global $conf;
1754    if(!$conf['license']) return '';
1755    if(!is_array($license[$conf['license']])) return '';
1756    $try   = array();
1757    $try[] = 'lib/images/license/'.$type.'/'.$conf['license'].'.png';
1758    $try[] = 'lib/images/license/'.$type.'/'.$conf['license'].'.gif';
1759    if(substr($conf['license'], 0, 3) == 'cc-') {
1760        $try[] = 'lib/images/license/'.$type.'/cc.png';
1761    }
1762    foreach($try as $src) {
1763        if(file_exists(DOKU_INC.$src)) return $src;
1764    }
1765    return '';
1766}
1767
1768/**
1769 * Checks if the given amount of memory is available
1770 *
1771 * If the memory_get_usage() function is not available the
1772 * function just assumes $bytes of already allocated memory
1773 *
1774 * @author Filip Oscadal <webmaster@illusionsoftworks.cz>
1775 * @author Andreas Gohr <andi@splitbrain.org>
1776 *
1777 * @param int  $mem    Size of memory you want to allocate in bytes
1778 * @param int  $bytes  already allocated memory (see above)
1779 * @return bool
1780 */
1781function is_mem_available($mem, $bytes = 1048576) {
1782    $limit = trim(ini_get('memory_limit'));
1783    if(empty($limit)) return true; // no limit set!
1784    if($limit == -1) return true; // unlimited
1785
1786    // parse limit to bytes
1787    $limit = php_to_byte($limit);
1788
1789    // get used memory if possible
1790    if(function_exists('memory_get_usage')) {
1791        $used = memory_get_usage();
1792    } else {
1793        $used = $bytes;
1794    }
1795
1796    if($used + $mem > $limit) {
1797        return false;
1798    }
1799
1800    return true;
1801}
1802
1803/**
1804 * Send a HTTP redirect to the browser
1805 *
1806 * Works arround Microsoft IIS cookie sending bug. Exits the script.
1807 *
1808 * @link   http://support.microsoft.com/kb/q176113/
1809 * @author Andreas Gohr <andi@splitbrain.org>
1810 *
1811 * @param string $url url being directed to
1812 */
1813function send_redirect($url) {
1814    $url = stripctl($url); // defend against HTTP Response Splitting
1815
1816    /* @var Input $INPUT */
1817    global $INPUT;
1818
1819    //are there any undisplayed messages? keep them in session for display
1820    global $MSG;
1821    if(isset($MSG) && count($MSG) && !defined('NOSESSION')) {
1822        //reopen session, store data and close session again
1823        @session_start();
1824        $_SESSION[DOKU_COOKIE]['msg'] = $MSG;
1825    }
1826
1827    // always close the session
1828    session_write_close();
1829
1830    // check if running on IIS < 6 with CGI-PHP
1831    if($INPUT->server->has('SERVER_SOFTWARE') && $INPUT->server->has('GATEWAY_INTERFACE') &&
1832        (strpos($INPUT->server->str('GATEWAY_INTERFACE'), 'CGI') !== false) &&
1833        (preg_match('|^Microsoft-IIS/(\d)\.\d$|', trim($INPUT->server->str('SERVER_SOFTWARE')), $matches)) &&
1834        $matches[1] < 6
1835    ) {
1836        header('Refresh: 0;url='.$url);
1837    } else {
1838        header('Location: '.$url);
1839    }
1840
1841    // no exits during unit tests
1842    if(defined('DOKU_UNITTEST')) {
1843        // pass info about the redirect back to the test suite
1844        $testRequest = TestRequest::getRunning();
1845        if($testRequest !== null) {
1846            $testRequest->addData('send_redirect', $url);
1847        }
1848        return;
1849    }
1850
1851    exit;
1852}
1853
1854/**
1855 * Validate a value using a set of valid values
1856 *
1857 * This function checks whether a specified value is set and in the array
1858 * $valid_values. If not, the function returns a default value or, if no
1859 * default is specified, throws an exception.
1860 *
1861 * @param string $param        The name of the parameter
1862 * @param array  $valid_values A set of valid values; Optionally a default may
1863 *                             be marked by the key “default”.
1864 * @param array  $array        The array containing the value (typically $_POST
1865 *                             or $_GET)
1866 * @param string $exc          The text of the raised exception
1867 *
1868 * @throws Exception
1869 * @return mixed
1870 * @author Adrian Lang <lang@cosmocode.de>
1871 */
1872function valid_input_set($param, $valid_values, $array, $exc = '') {
1873    if(isset($array[$param]) && in_array($array[$param], $valid_values)) {
1874        return $array[$param];
1875    } elseif(isset($valid_values['default'])) {
1876        return $valid_values['default'];
1877    } else {
1878        throw new Exception($exc);
1879    }
1880}
1881
1882/**
1883 * Read a preference from the DokuWiki cookie
1884 * (remembering both keys & values are urlencoded)
1885 *
1886 * @param string $pref     preference key
1887 * @param mixed  $default  value returned when preference not found
1888 * @return string preference value
1889 */
1890function get_doku_pref($pref, $default) {
1891    $enc_pref = urlencode($pref);
1892    if(isset($_COOKIE['DOKU_PREFS']) && strpos($_COOKIE['DOKU_PREFS'], $enc_pref) !== false) {
1893        $parts = explode('#', $_COOKIE['DOKU_PREFS']);
1894        $cnt   = count($parts);
1895
1896        // due to #2721 there might be duplicate entries,
1897        // so we read from the end
1898        for($i = $cnt-2; $i >= 0; $i -= 2) {
1899            if($parts[$i] == $enc_pref) {
1900                return urldecode($parts[$i + 1]);
1901            }
1902        }
1903    }
1904    return $default;
1905}
1906
1907/**
1908 * Add a preference to the DokuWiki cookie
1909 * (remembering $_COOKIE['DOKU_PREFS'] is urlencoded)
1910 * Remove it by setting $val to false
1911 *
1912 * @param string $pref  preference key
1913 * @param string $val   preference value
1914 */
1915function set_doku_pref($pref, $val) {
1916    global $conf;
1917    $orig = get_doku_pref($pref, false);
1918    $cookieVal = '';
1919
1920    if($orig !== false && ($orig !== $val)) {
1921        $parts = explode('#', $_COOKIE['DOKU_PREFS']);
1922        $cnt   = count($parts);
1923        // urlencode $pref for the comparison
1924        $enc_pref = rawurlencode($pref);
1925        $seen = false;
1926        for ($i = 0; $i < $cnt; $i += 2) {
1927            if ($parts[$i] == $enc_pref) {
1928                if (!$seen){
1929                    if ($val !== false) {
1930                        $parts[$i + 1] = rawurlencode($val);
1931                    } else {
1932                        unset($parts[$i]);
1933                        unset($parts[$i + 1]);
1934                    }
1935                    $seen = true;
1936                } else {
1937                    // no break because we want to remove duplicate entries
1938                    unset($parts[$i]);
1939                    unset($parts[$i + 1]);
1940                }
1941            }
1942        }
1943        $cookieVal = implode('#', $parts);
1944    } else if ($orig === false && $val !== false) {
1945        $cookieVal = (isset($_COOKIE['DOKU_PREFS']) ? $_COOKIE['DOKU_PREFS'] . '#' : '') .
1946            rawurlencode($pref) . '#' . rawurlencode($val);
1947    }
1948
1949    $cookieDir = empty($conf['cookiedir']) ? DOKU_REL : $conf['cookiedir'];
1950    if(defined('DOKU_UNITTEST')) {
1951        $_COOKIE['DOKU_PREFS'] = $cookieVal;
1952    }else{
1953        setcookie('DOKU_PREFS', $cookieVal, time()+365*24*3600, $cookieDir, '', ($conf['securecookie'] && is_ssl()));
1954    }
1955}
1956
1957/**
1958 * Strips source mapping declarations from given text #601
1959 *
1960 * @param string &$text reference to the CSS or JavaScript code to clean
1961 */
1962function stripsourcemaps(&$text){
1963    $text = preg_replace('/^(\/\/|\/\*)[@#]\s+sourceMappingURL=.*?(\*\/)?$/im', '\\1\\2', $text);
1964}
1965
1966/**
1967 * Returns the contents of a given SVG file for embedding
1968 *
1969 * Inlining SVGs saves on HTTP requests and more importantly allows for styling them through
1970 * CSS. However it should used with small SVGs only. The $maxsize setting ensures only small
1971 * files are embedded.
1972 *
1973 * This strips unneeded headers, comments and newline. The result is not a vaild standalone SVG!
1974 *
1975 * @param string $file full path to the SVG file
1976 * @param int $maxsize maximum allowed size for the SVG to be embedded
1977 * @return string|false the SVG content, false if the file couldn't be loaded
1978 */
1979function inlineSVG($file, $maxsize = 2048) {
1980    $file = trim($file);
1981    if($file === '') return false;
1982    if(!file_exists($file)) return false;
1983    if(filesize($file) > $maxsize) return false;
1984    if(!is_readable($file)) return false;
1985    $content = file_get_contents($file);
1986    $content = preg_replace('/<!--.*?(-->)/s','', $content); // comments
1987    $content = preg_replace('/<\?xml .*?\?>/i', '', $content); // xml header
1988    $content = preg_replace('/<!DOCTYPE .*?>/i', '', $content); // doc type
1989    $content = preg_replace('/>\s+</s', '><', $content); // newlines between tags
1990    $content = trim($content);
1991    if(substr($content, 0, 5) !== '<svg ') return false;
1992    return $content;
1993}
1994
1995//Setup VIM: ex: et ts=2 :
1996