]> git.mxchange.org Git - friendica.git/blob - src/Content/Text/BBCode.php
Merge pull request #11670 from Quix0r/fixes/more-type-hints-004
[friendica.git] / src / Content / Text / BBCode.php
1 <?php
2 /**
3  * @copyright Copyright (C) 2010-2022, the Friendica project
4  *
5  * @license GNU AGPL version 3 or any later version
6  *
7  * This program is free software: you can redistribute it and/or modify
8  * it under the terms of the GNU Affero General Public License as
9  * published by the Free Software Foundation, either version 3 of the
10  * License, or (at your option) any later version.
11  *
12  * This program is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15  * GNU Affero General Public License for more details.
16  *
17  * You should have received a copy of the GNU Affero General Public License
18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
19  *
20  */
21
22 namespace Friendica\Content\Text;
23
24 use DOMDocument;
25 use DOMXPath;
26 use Exception;
27 use Friendica\Content\ContactSelector;
28 use Friendica\Content\Item;
29 use Friendica\Content\OEmbed;
30 use Friendica\Content\PageInfo;
31 use Friendica\Content\Smilies;
32 use Friendica\Core\Hook;
33 use Friendica\Core\Logger;
34 use Friendica\Core\Protocol;
35 use Friendica\Core\Renderer;
36 use Friendica\DI;
37 use Friendica\Model\Contact;
38 use Friendica\Model\Event;
39 use Friendica\Model\Photo;
40 use Friendica\Model\Post;
41 use Friendica\Model\Tag;
42 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
43 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
44 use Friendica\Object\Image;
45 use Friendica\Protocol\Activity;
46 use Friendica\Util\Images;
47 use Friendica\Util\Map;
48 use Friendica\Util\ParseUrl;
49 use Friendica\Util\Proxy;
50 use Friendica\Util\Strings;
51 use Friendica\Util\XML;
52
53 class BBCode
54 {
55         // Update this value to the current date whenever changes are made to BBCode::convert
56         const VERSION = '2021-07-28';
57
58         const INTERNAL = 0;
59         const EXTERNAL = 1;
60         const API = 2;
61         const DIASPORA = 3;
62         const CONNECTORS = 4;
63         const OSTATUS = 7;
64         const TWITTER = 8;
65         const BACKLINK = 8;
66         const ACTIVITYPUB = 9;
67
68         const TOP_ANCHOR = '<br class="top-anchor">';
69         const BOTTOM_ANCHOR = '<br class="button-anchor">';
70         /**
71          * Fetches attachment data that were generated the old way
72          *
73          * @param string $body Message body
74          * @return array
75          *                     'type' -> Message type ('link', 'video', 'photo')
76          *                     'text' -> Text before the shared message
77          *                     'after' -> Text after the shared message
78          *                     'image' -> Preview image of the message
79          *                     'url' -> Url to the attached message
80          *                     'title' -> Title of the attachment
81          *                     'description' -> Description of the attachment
82          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
83          */
84         private static function getOldAttachmentData($body)
85         {
86                 $post = [];
87
88                 // Simplify image codes
89                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
90
91                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
92                         foreach ($attached as $data) {
93                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
94                                         continue;
95                                 }
96
97                                 $post['type'] = substr($data[1], 5);
98
99                                 $pos = strpos($body, $data[0]);
100                                 if ($pos > 0) {
101                                         $post['text'] = trim(substr($body, 0, $pos));
102                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
103                                 } else {
104                                         $post['text'] = trim(str_replace($data[0], '', $body));
105                                         $post['after'] = '';
106                                 }
107
108                                 $attacheddata = $data[2];
109
110                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
111
112                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
113
114                                         if ($picturedata) {
115                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
116                                                         $post['image'] = $matches[1];
117                                                 } else {
118                                                         $post['preview'] = $matches[1];
119                                                 }
120                                         }
121                                 }
122
123                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
124                                         $post['url'] = $matches[1];
125                                         $post['title'] = $matches[2];
126                                 }
127                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
128                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
129                                         $post['url'] = $matches[1];
130                                 }
131
132                                 // Search for description
133                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
134                                         $post['description'] = $matches[1];
135                                 }
136                         }
137                 }
138                 return $post;
139         }
140
141         /**
142          * Fetches attachment data that were generated with the "attachment" element
143          *
144          * @param string $body Message body
145          * @return array
146          *                     'type' -> Message type ('link', 'video', 'photo')
147          *                     'text' -> Text before the shared message
148          *                     'after' -> Text after the shared message
149          *                     'image' -> Preview image of the message
150          *                     'url' -> Url to the attached message
151          *                     'title' -> Title of the attachment
152          *                     'description' -> Description of the attachment
153          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
154          */
155         public static function getAttachmentData($body)
156         {
157                 DI::profiler()->startRecording('rendering');
158                 $data = [
159                         'type'          => '',
160                         'text'          => '',
161                         'after'         => '',
162                         'image'         => null,
163                         'url'           => '',
164                         'author_name'   => '',
165                         'author_url'    => '',
166                         'provider_name' => '',
167                         'provider_url'  => '',
168                         'title'         => '',
169                         'description'   => '',
170                 ];
171
172                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
173                         DI::profiler()->stopRecording();
174                         return self::getOldAttachmentData($body);
175                 }
176
177                 $attributes = $match[2];
178
179                 $data['text'] = trim($match[1]);
180
181                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
182                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
183                         $value = $matches[2] ?? '';
184
185                         if ($value != '') {
186                                 switch ($field) {
187                                         case 'publisher_name':
188                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
189                                                 break;
190                                         case 'publisher_url':
191                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
192                                                 break;
193                                         case 'author_name':
194                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
195                                                 if ($data['provider_name'] == $data['author_name']) {
196                                                         $data['author_name'] = '';
197                                                 }
198                                                 break;
199                                         case 'author_url':
200                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
201                                                 if ($data['provider_url'] == $data['author_url']) {
202                                                         $data['author_url'] = '';
203                                                 }
204                                                 break;
205                                         case 'title':
206                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
207                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
208                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
209                                                 $data['title'] = $value;
210                                         default:
211                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
212                                                 break;
213                                 }
214                         }
215                 }
216
217                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
218                         DI::profiler()->stopRecording();
219                         return [];
220                 }
221
222                 $data['description'] = trim($match[3]);
223
224                 $data['after'] = trim($match[4]);
225
226                 $parts = parse_url($data['url']);
227                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
228                         if (empty($data['provider_name'])) {
229                                 $data['provider_name'] = $parts['host'];
230                         }
231                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
232                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
233
234                                 if (!empty($parts['port'])) {
235                                         $data['provider_url'] .= ':' . $parts['port'];
236                                 }
237                         }
238                 }
239
240                 DI::profiler()->stopRecording();
241                 return $data;
242         }
243
244         public static function getAttachedData($body, $item = [])
245         {
246                 /*
247                 - text:
248                 - type: link, video, photo
249                 - title:
250                 - url:
251                 - image:
252                 - description:
253                 - (thumbnail)
254                 */
255
256                 DI::profiler()->startRecording('rendering');
257                 $has_title = !empty($item['title']);
258                 $plink = $item['plink'] ?? '';
259                 $post = self::getAttachmentData($body);
260
261                 // Get all linked images with alternative image description
262                 if (preg_match_all("/\[img=(http[^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
263                         foreach ($pictures as $picture) {
264                                 if (Photo::isLocal($picture[1])) {
265                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
266                                 } else {
267                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
268                                 }
269                         }
270                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
271                                 $post['image_description'] = $post['images'][0]['description'];
272                         }
273                 }
274
275                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
276                         foreach ($pictures as $picture) {
277                                 if (Photo::isLocal($picture[1])) {
278                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
279                                 } else {
280                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
281                                 }
282                         }
283                 }
284
285                 // if nothing is found, it maybe having an image.
286                 if (!isset($post['type'])) {
287                         // Simplify image codes
288                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
289                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
290                         $post['text'] = $body;
291
292                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
293                                 if ((count($pictures) == 1) && !$has_title) {
294                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
295                                                 // Replace the preview picture with the real picture
296                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
297                                                 $data = ['url' => $url, 'type' => 'photo'];
298                                         } else {
299                                                 // Checking, if the link goes to a picture
300                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
301                                         }
302
303                                         // Workaround:
304                                         // Sometimes photo posts to the own album are not detected at the start.
305                                         // So we seem to cannot use the cache for these cases. That's strange.
306                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
307                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
308                                         }
309
310                                         if ($data['type'] == 'photo') {
311                                                 $post['type'] = 'photo';
312                                                 if (isset($data['images'][0])) {
313                                                         $post['image'] = $data['images'][0]['src'];
314                                                         $post['url'] = $data['url'];
315                                                 } else {
316                                                         $post['image'] = $data['url'];
317                                                 }
318
319                                                 $post['preview'] = $pictures[0][2];
320                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
321                                         } else {
322                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
323                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
324                                                         $post['type'] = 'photo';
325                                                         $post['image'] = $pictures[0][1];
326                                                         $post['preview'] = $pictures[0][2];
327                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
328                                                 }
329                                         }
330                                 } elseif (count($pictures) > 0) {
331                                         if (count($pictures) > 4) {
332                                                 $post['type'] = 'link';
333                                                 $post['url'] = $plink;
334                                         } else {
335                                                 $post['type'] = 'photo';
336                                         }
337
338                                         $post['image'] = $pictures[0][2];
339                                         $post['text'] = $body;
340
341                                         foreach ($pictures as $picture) {
342                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
343                                         }
344                                 }
345                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
346                                 if ($has_title) {
347                                         $post['type'] = 'link';
348                                         $post['url'] = $plink;
349                                 } else {
350                                         $post['type'] = 'photo';
351                                 }
352
353                                 $post['image'] = $pictures[0][1];
354                                 $post['text'] = $body;
355                                 foreach ($pictures as $picture) {
356                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
357                                 }
358                         }
359
360                         // Test for the external links
361                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
362                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
363
364                         $links = array_merge($links1, $links2);
365
366                         // If there is only a single one, then use it.
367                         // This should cover link posts via API.
368                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
369                                 $post['type'] = 'link';
370                                 $post['url'] = $links[0][1];
371                         }
372
373                         // Simplify "video" element
374                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
375
376                         // Now count the number of external media links
377                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
378                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
379                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
380                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
381
382                         // Add them to the other external links
383                         $links = array_merge($links, $links1, $links2, $links3, $links4);
384
385                         // Are there more than one?
386                         if (count($links) > 1) {
387                                 // The post will be the type "text", which means a blog post
388                                 unset($post['type']);
389                                 $post['url'] = $plink;
390                         }
391
392                         if (!isset($post['type'])) {
393                                 $post['type'] = "text";
394                                 $post['text'] = trim($body);
395                         }
396
397                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
398                                 $post['images'] = $post['remote_images'];
399                                 $post['image'] = $post['images'][0]['url'];
400                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
401                                         $post['image_description'] = $post['images'][0]['description'];
402                                 }
403                         }
404                         unset($post['remote_images']);
405                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
406                         $data = ParseUrl::getSiteinfoCached($post['url']);
407
408                         if (isset($data['images'][0])) {
409                                 $post['image'] = $data['images'][0]['src'];
410                         }
411                 }
412
413                 DI::profiler()->stopRecording();
414                 return $post;
415         }
416
417         /**
418          * Remove [attachment] BBCode and replaces it with a regular [url]
419          *
420          * @param string  $body
421          * @param boolean $no_link_desc No link description
422          *
423          * @return string with replaced body
424          */
425         public static function removeAttachment($body, $no_link_desc = false)
426         {
427                 return preg_replace_callback("/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
428                         function ($match) use ($body, $no_link_desc) {
429                                 $attach_data = self::getAttachmentData($match[0]);
430                                 if (empty($attach_data['url'])) {
431                                         return $match[0];
432                                 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
433                                         return '';
434                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
435                                         return " \n[url]" . $attach_data['url'] . "[/url]\n";
436                                 } else {
437                                         return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
438                                 }
439                 }, $body);
440         }
441
442         /**
443          * Converts a BBCode text into plaintext
444          *
445          * @param      $text
446          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
447          *
448          * @return string
449          */
450         public static function toPlaintext($text, $keep_urls = true)
451         {
452                 DI::profiler()->startRecording('rendering');
453                 // Remove pictures in advance to avoid unneeded proxy calls
454                 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
455                 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
456
457                 // Remove attachment
458                 $text = self::removeAttachment($text);
459
460                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
461
462                 DI::profiler()->stopRecording();
463                 return $naked_text;
464         }
465
466         private static function proxyUrl($image, $simplehtml = self::INTERNAL, $uriid = 0, $size = '')
467         {
468                 // Only send proxied pictures to API and for internal display
469                 if (!in_array($simplehtml, [self::INTERNAL, self::API])) {
470                         return $image;
471                 } elseif ($uriid > 0) {
472                         return Post\Link::getByLink($uriid, $image, $size);
473                 } else {
474                         return Proxy::proxifyUrl($image, $size);
475                 }
476         }
477
478         /**
479          * This function changing the visual size (not the real size) of images.
480          * The function does not work for pictures with an alternate text description.
481          * This could only be changed by using some new "img" BBCode format.
482          *
483          * @param string $srctext The body with images
484          * @return string The body with possibly scaled images
485          */
486         public static function scaleExternalImages(string $srctext)
487         {
488                 DI::profiler()->startRecording('rendering');
489                 $s = $srctext;
490
491                 // Simplify image links
492                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
493
494                 $matches = null;
495                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
496                 if ($c) {
497                         foreach ($matches as $mtch) {
498                                 Logger::info('scale_external_image', ['image' => $mtch[1]]);
499
500                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
501                                 if (stristr($mtch[1], $hostname)) {
502                                         continue;
503                                 }
504
505                                 $curlResult = DI::httpClient()->get($mtch[1], HttpClientAccept::IMAGE);
506                                 if (!$curlResult->isSuccess()) {
507                                         continue;
508                                 }
509
510                                 Logger::debug('Got picture', ['Content-Type' => $curlResult->getHeader('Content-Type'), 'url' => $mtch[1]]);
511
512                                 $i = $curlResult->getBody();
513                                 $type = $curlResult->getContentType();
514                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
515
516                                 if ($i) {
517                                         $Image = new Image($i, $type);
518                                         if ($Image->isValid()) {
519                                                 $orig_width = $Image->getWidth();
520                                                 $orig_height = $Image->getHeight();
521
522                                                 if ($orig_width > 640 || $orig_height > 640) {
523                                                         $Image->scaleDown(640);
524                                                         $new_width = $Image->getWidth();
525                                                         $new_height = $Image->getHeight();
526                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
527                                                         $s = str_replace(
528                                                                 $mtch[0],
529                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
530                                                                 . "\n",
531                                                                 $s
532                                                         );
533                                                         Logger::info('New string', ['image' => $s]);
534                                                 }
535                                         }
536                                 }
537                         }
538                 }
539
540                 DI::profiler()->stopRecording();
541                 return $s;
542         }
543
544         /**
545          * Truncates imported message body string length to max_import_size
546          *
547          * The purpose of this function is to apply system message length limits to
548          * imported messages without including any embedded photos in the length
549          *
550          * @param string $body
551          * @return string
552          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
553          */
554         public static function limitBodySize($body)
555         {
556                 DI::profiler()->startRecording('rendering');
557                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
558
559                 // If the length of the body, including the embedded images, is smaller
560                 // than the maximum, then don't waste time looking for the images
561                 if ($maxlen && (strlen($body) > $maxlen)) {
562
563                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
564
565                         $orig_body = $body;
566                         $new_body = '';
567                         $textlen = 0;
568
569                         $img_start = strpos($orig_body, '[img');
570                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
571                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
572                         while (($img_st_close !== false) && ($img_end !== false)) {
573
574                                 $img_st_close++; // make it point to AFTER the closing bracket
575                                 $img_end += $img_start;
576                                 $img_end += strlen('[/img]');
577
578                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
579                                         // This is an embedded image
580
581                                         if (($textlen + $img_start) > $maxlen) {
582                                                 if ($textlen < $maxlen) {
583                                                         Logger::info('the limit happens before an embedded image');
584                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
585                                                         $textlen = $maxlen;
586                                                 }
587                                         } else {
588                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
589                                                 $textlen += $img_start;
590                                         }
591
592                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
593                                 } else {
594
595                                         if (($textlen + $img_end) > $maxlen) {
596                                                 if ($textlen < $maxlen) {
597                                                         Logger::info('the limit happens before the end of a non-embedded image');
598                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
599                                                         $textlen = $maxlen;
600                                                 }
601                                         } else {
602                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
603                                                 $textlen += $img_end;
604                                         }
605                                 }
606                                 $orig_body = substr($orig_body, $img_end);
607
608                                 if ($orig_body === false) {
609                                         // in case the body ends on a closing image tag
610                                         $orig_body = '';
611                                 }
612
613                                 $img_start = strpos($orig_body, '[img');
614                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
615                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
616                         }
617
618                         if (($textlen + strlen($orig_body)) > $maxlen) {
619                                 if ($textlen < $maxlen) {
620                                         Logger::info('the limit happens after the end of the last image');
621                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
622                                 }
623                         } else {
624                                 Logger::info('the text size with embedded images extracted did not violate the limit');
625                                 $new_body = $new_body . $orig_body;
626                         }
627
628                         DI::profiler()->stopRecording();
629                         return $new_body;
630                 } else {
631                         DI::profiler()->stopRecording();
632                         return $body;
633                 }
634         }
635
636         /**
637          * Processes [attachment] tags
638          *
639          * Note: Can produce a [bookmark] tag in the returned string
640          *
641          * @param string  $text
642          * @param integer $simplehtml
643          * @param bool    $tryoembed
644          * @param array   $data
645          * @param int     $uriid
646          * @return string
647          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
648          */
649         public static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true, array $data = [], $uriid = 0)
650         {
651                 DI::profiler()->startRecording('rendering');
652                 $data = $data ?: self::getAttachmentData($text);
653                 if (empty($data) || empty($data['url'])) {
654                         DI::profiler()->stopRecording();
655                         return $text;
656                 }
657
658                 if (isset($data['title'])) {
659                         $data['title'] = strip_tags($data['title']);
660                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
661                 } else {
662                         $data['title'] = null;
663                 }
664
665                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
666                         $data['preview'] = $data['image'];
667                         $data['image'] = '';
668                 }
669
670                 $return = '';
671                 try {
672                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
673                                 $return = OEmbed::getHTML($data['url'], $data['title']);
674                         } else {
675                                 throw new Exception('OEmbed is disabled for this attachment.');
676                         }
677                 } catch (Exception $e) {
678                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
679
680                         if ($simplehtml != self::CONNECTORS) {
681                                 $return = sprintf('<div class="type-%s">', $data['type']);
682                         }
683
684                         if (!empty($data['title']) && !empty($data['url'])) {
685                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
686                                         $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
687                                 } else {
688                                         if (!empty($data['image'])) {
689                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
690                                         } elseif (!empty($data['preview'])) {
691                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br>', $data['url'], self::proxyUrl($data['preview'], $simplehtml, $uriid), $data['title']);
692                                         }
693                                         $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
694                                 }
695                         }
696
697                         if (!empty($data['description']) && $data['description'] != $data['title']) {
698                                 // Sanitize the HTML
699                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
700                         }
701
702                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
703                                 if (!empty($data['author_name'])) {
704                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
705                                 } else {
706                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
707                                 }
708                         }
709
710                         if ($simplehtml != self::CONNECTORS) {
711                                 $return .= '</div>';
712                         }
713                 }
714
715                 DI::profiler()->stopRecording();
716                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
717         }
718
719         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
720         {
721                 DI::profiler()->startRecording('rendering');
722                 $data = self::getAttachmentData($Text);
723
724                 if (!$data) {
725                         DI::profiler()->stopRecording();
726                         return $Text;
727                 } elseif ($nolink) {
728                         DI::profiler()->stopRecording();
729                         return $data['text'] . ($data['after'] ?? '');
730                 }
731
732                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
733                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
734                 if ($plaintext || (($title != '') && strstr($text, $title))) {
735                         $data['title'] = $data['url'];
736                 } elseif (($text != '') && strstr($title, $text)) {
737                         $data['text'] = $data['title'];
738                         $data['title'] = $data['url'];
739                 }
740
741                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
742                         DI::profiler()->stopRecording();
743                         return $data['title'] . $data['after'];
744                 }
745
746                 // If the link already is included in the post, don't add it again
747                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
748                         DI::profiler()->stopRecording();
749                         return $data['text'] . $data['after'];
750                 }
751
752                 $text = $data['text'];
753
754                 if (!empty($data['url']) && !empty($data['title'])) {
755                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
756                 } elseif (!empty($data['url'])) {
757                         $text .= "\n[url]" . $data['url'] . '[/url]';
758                 }
759
760                 DI::profiler()->stopRecording();
761                 return $text . "\n" . $data['after'];
762         }
763
764         /**
765          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
766          *
767          * @param array $match Array with the matching values
768          * @return string reformatted link including HTML codes
769          */
770         private static function convertUrlForActivityPubCallback($match)
771         {
772                 $url = $match[1];
773
774                 if (isset($match[2]) && ($match[1] != $match[2])) {
775                         return $match[0];
776                 }
777
778                 $parts = parse_url($url);
779                 if (!isset($parts['scheme'])) {
780                         return $match[0];
781                 }
782
783                 return self::convertUrlForActivityPub($url);
784         }
785
786         /**
787          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
788          *
789          * @param string $url URL that is about to be reformatted
790          * @return string reformatted link including HTML codes
791          */
792         private static function convertUrlForActivityPub($url)
793         {
794                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
795                 return sprintf($html, $url, self::getStyledURL($url));
796         }
797
798         /**
799          * Converts an URL in a nicer format (without the scheme and possibly shortened)
800          *
801          * @param string $url URL that is about to be reformatted
802          * @return string reformatted link
803          */
804         private static function getStyledURL($url)
805         {
806                 $parts = parse_url($url);
807                 $scheme = $parts['scheme'] . '://';
808                 $styled_url = str_replace($scheme, '', $url);
809
810                 if (strlen($styled_url) > 30) {
811                         $styled_url = substr($styled_url, 0, 30) . "…";
812                 }
813
814                 return $styled_url;
815         }
816
817         /*
818          * [noparse][i]italic[/i][/noparse] turns into
819          * [noparse][ i ]italic[ /i ][/noparse],
820          * to hide them from parser.
821          */
822         private static function escapeNoparseCallback($match)
823         {
824                 $whole_match = $match[0];
825                 $captured = $match[1];
826                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
827                 $new_str = str_replace($captured, $spacefied, $whole_match);
828                 return $new_str;
829         }
830
831         /*
832          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
833          * now turns back and the [noparse] tags are trimed
834          * returning [i]italic[/i]
835          */
836         private static function unescapeNoparseCallback($match)
837         {
838                 $captured = $match[1];
839                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
840                 return $unspacefied;
841         }
842
843         /**
844          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
845          * occurrences
846          *
847          * @param string $text        Text to search
848          * @param string $name        Tag name
849          * @param int    $occurrences Number of first occurrences to skip
850          * @return boolean|array
851          */
852         public static function getTagPosition($text, $name, $occurrences = 0)
853         {
854                 DI::profiler()->startRecording('rendering');
855                 if ($occurrences < 0) {
856                         $occurrences = 0;
857                 }
858
859                 $start_open = -1;
860                 for ($i = 0; $i <= $occurrences; $i++) {
861                         if ($start_open !== false) {
862                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
863                         }
864                 }
865
866                 if ($start_open === false) {
867                         DI::profiler()->stopRecording();
868                         return false;
869                 }
870
871                 $start_equal = strpos($text, '=', $start_open);
872                 $start_close = strpos($text, ']', $start_open);
873
874                 if ($start_close === false) {
875                         DI::profiler()->stopRecording();
876                         return false;
877                 }
878
879                 $start_close++;
880
881                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
882
883                 if ($end_open === false) {
884                         DI::profiler()->stopRecording();
885                         return false;
886                 }
887
888                 $res = [
889                         'start' => [
890                                 'open' => $start_open,
891                                 'close' => $start_close
892                         ],
893                         'end' => [
894                                 'open' => $end_open,
895                                 'close' => $end_open + strlen('[/' . $name . ']')
896                         ],
897                 ];
898
899                 if ($start_equal !== false) {
900                         $res['start']['equal'] = $start_equal + 1;
901                 }
902
903                 DI::profiler()->stopRecording();
904                 return $res;
905         }
906
907         /**
908          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
909          *
910          * @param string $pattern Preg pattern string
911          * @param string $replace Preg replace string
912          * @param string $name    BBCode tag name
913          * @param string $text    Text to search
914          * @return string
915          */
916         public static function pregReplaceInTag($pattern, $replace, $name, $text)
917         {
918                 DI::profiler()->startRecording('rendering');
919                 $occurrences = 0;
920                 $pos = self::getTagPosition($text, $name, $occurrences);
921                 while ($pos !== false && $occurrences++ < 1000) {
922                         $start = substr($text, 0, $pos['start']['open']);
923                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
924                         $end = substr($text, $pos['end']['close']);
925                         if ($end === false) {
926                                 $end = '';
927                         }
928
929                         $subject = preg_replace($pattern, $replace, $subject);
930                         $text = $start . $subject . $end;
931
932                         $pos = self::getTagPosition($text, $name, $occurrences);
933                 }
934
935                 DI::profiler()->stopRecording();
936                 return $text;
937         }
938
939         private static function extractImagesFromItemBody($body)
940         {
941                 $saved_image = [];
942                 $orig_body = $body;
943                 $new_body = '';
944
945                 $cnt = 0;
946                 $img_start = strpos($orig_body, '[img');
947                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
948                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
949                 while (($img_st_close !== false) && ($img_end !== false)) {
950                         $img_st_close++; // make it point to AFTER the closing bracket
951                         $img_end += $img_start;
952
953                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
954                                 // This is an embedded image
955                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
956                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
957
958                                 $cnt++;
959                         } else {
960                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
961                         }
962
963                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
964
965                         if ($orig_body === false) {
966                                 // in case the body ends on a closing image tag
967                                 $orig_body = '';
968                         }
969
970                         $img_start = strpos($orig_body, '[img');
971                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
972                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
973                 }
974
975                 $new_body = $new_body . $orig_body;
976
977                 return ['body' => $new_body, 'images' => $saved_image];
978         }
979
980         private static function interpolateSavedImagesIntoItemBody($uriid, $body, array $images)
981         {
982                 $newbody = $body;
983
984                 $cnt = 0;
985                 foreach ($images as $image) {
986                         // We're depending on the property of 'foreach' (specified on the PHP website) that
987                         // it loops over the array starting from the first element and going sequentially
988                         // to the last element
989                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
990                                 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
991                         $cnt++;
992                 }
993
994                 return $newbody;
995         }
996
997         /**
998          *
999          * @param  string   $text     A BBCode string
1000          * @return array share attributes
1001          */
1002         public static function fetchShareAttributes($text)
1003         {
1004                 DI::profiler()->startRecording('rendering');
1005                 // See Issue https://github.com/friendica/friendica/issues/10454
1006                 // Hashtags in usernames are expanded to links. This here is a quick fix.
1007                 $text = preg_replace('/([@!#])\[url\=.*?\](.*?)\[\/url\]/ism', '$1$2', $text);
1008
1009                 $attributes = [];
1010                 if (!preg_match("/(.*?)\[share(.*?)\](.*)\[\/share\]/ism", $text, $matches)) {
1011                         DI::profiler()->stopRecording();
1012                         return $attributes;
1013                 }
1014
1015                 $attribute_string = $matches[2];
1016                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
1017                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1018                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1019                 }
1020                 DI::profiler()->stopRecording();
1021                 return $attributes;
1022         }
1023
1024         /**
1025          * This function converts a [share] block to text according to a provided callback function whose signature is:
1026          *
1027          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
1028          *
1029          * Where:
1030          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
1031          * data lookup
1032          * - $author_contact is a contact record array
1033          * - $content is the inner content of the [share] block
1034          * - $is_quote_share indicates whether there's any content before the [share] block
1035          * - Return value is the string that should replace the [share] block in the provided text
1036          *
1037          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
1038          *
1039          * @param  string   $text     A BBCode string
1040          * @param  callable $callback
1041          * @return string The BBCode string with all [share] blocks replaced
1042          */
1043         public static function convertShare($text, callable $callback, int $uriid = 0)
1044         {
1045                 DI::profiler()->startRecording('rendering');
1046                 $return = preg_replace_callback(
1047                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
1048                         function ($match) use ($callback, $uriid) {
1049                                 $attribute_string = $match[2];
1050                                 $attributes = [];
1051                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
1052                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1053                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1054                                 }
1055
1056                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
1057                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1058                                 $author_contact['addr'] = ($author_contact['addr'] ?? '');
1059
1060                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1061                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1062                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1063
1064                                 if (!empty($author_contact['id'])) {
1065                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
1066                                 } elseif ($attributes['avatar']) {
1067                                         $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
1068                                 }
1069
1070                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1071
1072                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1073                         },
1074                         $text
1075                 );
1076
1077                 DI::profiler()->stopRecording();
1078                 return $return;
1079         }
1080
1081         /**
1082          * Convert complex IMG and ZMG elements
1083          *
1084          * @param [type] $text
1085          * @param integer $simplehtml
1086          * @param integer $uriid
1087          * @return string
1088          */
1089         private static function convertImages(string $text, int $simplehtml, int $uriid = 0):string
1090         {
1091                 DI::profiler()->startRecording('rendering');
1092                 $return = preg_replace_callback(
1093                         "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
1094                         function ($match) use ($simplehtml, $uriid) {
1095                                 $attribute_string = $match[1];
1096                                 $attributes = [];
1097                                 foreach (['alt', 'width', 'height'] as $field) {
1098                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1099                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1100                                 }
1101
1102                                 $img_str = '<img src="' .
1103                                 self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
1104                                 foreach ($attributes as $key => $value) {
1105                                         if (!empty($value)) {
1106                                                 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
1107                                         }
1108                                 }
1109                                 return $img_str . '>';
1110                         },
1111                         $text
1112                 );
1113
1114                 DI::profiler()->stopRecording();
1115                 return $return;
1116         }
1117
1118         /**
1119          * Default [share] tag conversion callback
1120          *
1121          * Note: Can produce a [bookmark] tag in the output
1122          *
1123          * @see BBCode::convertShare()
1124          * @param array   $attributes     [share] block attribute values
1125          * @param array   $author_contact Contact row of the shared author
1126          * @param string  $content        Inner content of the [share] block
1127          * @param boolean $is_quote_share Whether there is content before the [share] block
1128          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1129          * @return string
1130          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1131          */
1132         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1133         {
1134                 DI::profiler()->startRecording('rendering');
1135                 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
1136
1137                 switch ($simplehtml) {
1138                         case self::API:
1139                                 $text = ($is_quote_share? '<br>' : '') .
1140                                 '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
1141                                 '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
1142                                 break;
1143                         case self::DIASPORA:
1144                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1145                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1146                                 } else {
1147                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1148
1149                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1150                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1151                                         }
1152
1153                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1154
1155                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1156                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1157                                         }
1158                                 }
1159
1160                                 break;
1161                         case self::CONNECTORS:
1162                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1163                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1164                                 $headline .= ':</b></p>' . "\n";
1165
1166                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
1167
1168                                 break;
1169                         case self::OSTATUS:
1170                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1171                                 break;
1172                         case self::ACTIVITYPUB:
1173                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1174                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1175                                 break;
1176                         default:
1177                                 $text = ($is_quote_share? "\n" : '');
1178
1179                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1180                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1181
1182                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1183                                 $text .= Renderer::replaceMacros($tpl, [
1184                                         '$profile'      => $attributes['profile'],
1185                                         '$avatar'       => $attributes['avatar'],
1186                                         '$author'       => $attributes['author'],
1187                                         '$link'         => $attributes['link'],
1188                                         '$link_title'   => DI::l10n()->t('Link to source'),
1189                                         '$posted'       => $attributes['posted'],
1190                                         '$guid'         => $attributes['guid'],
1191                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1192                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1193                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
1194                                 ]);
1195                                 break;
1196                 }
1197
1198                 return $text;
1199         }
1200
1201         private static function removePictureLinksCallback($match)
1202         {
1203                 $cache_key = 'remove:' . $match[1];
1204                 $text = DI::cache()->get($cache_key);
1205
1206                 if (is_null($text)) {
1207                         $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1208                         if ($curlResult->isSuccess()) {
1209                                 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1210                         } else {
1211                                 $mimetype = '';
1212                         }
1213
1214                         if (substr($mimetype, 0, 6) == 'image/') {
1215                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1216                         } else {
1217                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1218
1219                                 // if its not a picture then look if its a page that contains a picture link
1220                                 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1221                                 if (empty($body)) {
1222                                         DI::cache()->set($cache_key, $text);
1223                                         return $text;
1224                                 }
1225
1226                                 $doc = new DOMDocument();
1227                                 @$doc->loadHTML($body);
1228                                 $xpath = new DOMXPath($doc);
1229                                 $list = $xpath->query("//meta[@name]");
1230                                 foreach ($list as $node) {
1231                                         $attr = [];
1232
1233                                         if ($node->attributes->length) {
1234                                                 foreach ($node->attributes as $attribute) {
1235                                                         $attr[$attribute->name] = $attribute->value;
1236                                                 }
1237                                         }
1238
1239                                         if (strtolower($attr['name']) == 'twitter:image') {
1240                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1241                                         }
1242                                 }
1243                         }
1244                         DI::cache()->set($cache_key, $text);
1245                 }
1246
1247                 return $text;
1248         }
1249
1250         /**
1251          * Callback: Expands links from given $match array
1252          *
1253          * @param arrat $match Array with link match
1254          * @return string BBCode
1255          */
1256         private static function expandLinksCallback(array $match): string
1257         {
1258                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1259                         return ($match[1] . '[url]' . $match[2] . '[/url]');
1260                 } else {
1261                         return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1262                 }
1263         }
1264
1265         /**
1266          * Callback: Cleans picture links
1267          *
1268          * @param arrat $match Array with link match
1269          * @return string BBCode
1270          */
1271         private static function cleanPictureLinksCallback(array $match): string
1272         {
1273                 // When the picture link is the own photo path then we can avoid fetching the link
1274                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1275                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1276                         if (!empty($match[3])) {
1277                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1278                         } else {
1279                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1280                         }
1281                         return $text;
1282                 }
1283
1284                 $cache_key = 'clean:' . $match[1];
1285                 $text = DI::cache()->get($cache_key);
1286                 if (!is_null($text)) {
1287                         return $text;
1288                 }
1289
1290                 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1291                 if ($curlResult->isSuccess()) {
1292                         $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1293                 } else {
1294                         $mimetype = '';
1295                 }
1296
1297                 // if its a link to a picture then embed this picture
1298                 if (substr($mimetype, 0, 6) == 'image/') {
1299                         $text = '[img]' . $match[1] . '[/img]';
1300                 } else {
1301                         if (!empty($match[3])) {
1302                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1303                         } else {
1304                                 $text = '[img]' . $match[2] . '[/img]';
1305                         }
1306
1307                         // if its not a picture then look if its a page that contains a picture link
1308                         $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1309                         if (empty($body)) {
1310                                 DI::cache()->set($cache_key, $text);
1311                                 return $text;
1312                         }
1313
1314                         $doc = new DOMDocument();
1315                         @$doc->loadHTML($body);
1316                         $xpath = new DOMXPath($doc);
1317                         $list = $xpath->query("//meta[@name]");
1318                         foreach ($list as $node) {
1319                                 $attr = [];
1320                                 if ($node->attributes->length) {
1321                                         foreach ($node->attributes as $attribute) {
1322                                                 $attr[$attribute->name] = $attribute->value;
1323                                         }
1324                                 }
1325
1326                                 if (strtolower($attr['name']) == "twitter:image") {
1327                                         if (!empty($match[3])) {
1328                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1329                                         } else {
1330                                                 $text = "[img]" . $attr['content'] . "[/img]";
1331                                         }
1332                                 }
1333                         }
1334                 }
1335                 DI::cache()->set($cache_key, $text);
1336
1337                 return $text;
1338         }
1339
1340         /**
1341          * Cleans picture links
1342          *
1343          * @param string $text HTML/BBCode string
1344          * @return string Cleaned HTML/BBCode
1345          */
1346         public static function cleanPictureLinks(string $text): string
1347         {
1348                 DI::profiler()->startRecording('rendering');
1349                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1350                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1351                 DI::profiler()->stopRecording();
1352                 return $return;
1353         }
1354
1355         /**
1356          * Removes links
1357          *
1358          * @param string $text HTML/BBCode string
1359          * @return string Cleaned HTML/BBCode
1360          */
1361         public static function removeLinks(string $bbcode): string
1362         {
1363                 DI::profiler()->startRecording('rendering');
1364                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1365                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1366
1367                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1368                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1369                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1370                 DI::profiler()->stopRecording();
1371                 return $bbcode;
1372         }
1373
1374         /**
1375          * Replace names in mentions with nicknames
1376          *
1377          * @param string $body HTML/BBCode
1378          * @return string Body with replaced mentions
1379          */
1380         public static function setMentionsToNicknames(string $body): string
1381         {
1382                 DI::profiler()->startRecording('rendering');
1383                 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1384                 $body = preg_replace_callback($regexp, ['self', 'mentionCallback'], $body);
1385                 DI::profiler()->stopRecording();
1386                 return $body;
1387         }
1388
1389         /**
1390          * Callback function to replace a Friendica style mention in a mention with the nickname
1391          *
1392          * @param array $match Matching values for the callback
1393          * @return string Replaced mention or empty string
1394          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1395          */
1396         private static function mentionCallback(array $match): string
1397         {
1398                 if (empty($match[2])) {
1399                         return '';
1400                 }
1401
1402                 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1403                 if (empty($data['nick'])) {
1404                         return $match[0];
1405                 }
1406
1407                 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1408         }
1409
1410         /**
1411          * Converts a BBCode message for a given URI-ID to a HTML message
1412          *
1413          * BBcode 2 HTML was written by WAY2WEB.net
1414          * extended to work with Mistpark/Friendica - Mike Macgirvin
1415          *
1416          * Simple HTML values meaning:
1417          * - 0: Friendica display
1418          * - 1: Unused
1419          * - 2: Used for Windows Phone push, Friendica API
1420          * - 3: Used before converting to Markdown in bb2diaspora.php
1421          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1422          * - 5: Unused
1423          * - 6: Unused
1424          * - 7: Used for dfrn, OStatus
1425          * - 8: Used for Twitter, WP backlink text setting
1426          * - 9: ActivityPub
1427          *
1428          * @param int    $uriid
1429          * @param string $text
1430          * @param int    $simple_html
1431          * @return string
1432          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1433          */
1434         public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1435         {
1436                 $try_oembed = ($simple_html == self::INTERNAL);
1437
1438                 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1439         }
1440
1441         /**
1442          * Converts a BBCode message to HTML message
1443          *
1444          * BBcode 2 HTML was written by WAY2WEB.net
1445          * extended to work with Mistpark/Friendica - Mike Macgirvin
1446          *
1447          * Simple HTML values meaning:
1448          * - 0: Friendica display
1449          * - 1: Unused
1450          * - 2: Used for Windows Phone push, Friendica API
1451          * - 3: Used before converting to Markdown in bb2diaspora.php
1452          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1453          * - 5: Unused
1454          * - 6: Unused
1455          * - 7: Used for dfrn, OStatus
1456          * - 8: Used for Twitter, WP backlink text setting
1457          * - 9: ActivityPub
1458          *
1459          * @param string $text
1460          * @param bool   $try_oembed
1461          * @param int    $simple_html
1462          * @param bool   $for_plaintext
1463          * @param int    $uriid
1464          * @return string Converted code or empty string
1465          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1466          */
1467         public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1468         {
1469                 // Accounting for null default column values
1470                 if (is_null($text) || $text === '') {
1471                         return '';
1472                 }
1473
1474                 DI::profiler()->startRecording('rendering');
1475
1476                 Hook::callAll('bbcode', $text);
1477
1478                 $a = DI::app();
1479
1480                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1481                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1482                                 /*
1483                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1484                                  *
1485                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1486                                  * $match[1] = $url
1487                                  * $match[2] = $title or absent
1488                                  */
1489                                 $try_oembed_callback = function ($match)
1490                                 {
1491                                         $url = $match[1];
1492                                         $title = $match[2] ?? null;
1493
1494                                         try {
1495                                                 $return = OEmbed::getHTML($url, $title);
1496                                         } catch (Exception $ex) {
1497                                                 $return = $match[0];
1498                                         }
1499
1500                                         return $return;
1501                                 };
1502
1503                                 // Remove the abstract element. It is a non visible element.
1504                                 $text = self::stripAbstract($text);
1505
1506                                 // Line ending normalisation
1507                                 $text = str_replace("\r\n", "\n", $text);
1508
1509                                 // Move new lines outside of tags
1510                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1511                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1512
1513                                 // Extract the private images which use data urls since preg has issues with
1514                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1515                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1516
1517                                 $extracted = self::extractImagesFromItemBody($text);
1518                                 $text = $extracted['body'];
1519                                 $saved_image = $extracted['images'];
1520
1521                                 // If we find any event code, turn it into an event.
1522                                 // After we're finished processing the bbcode we'll
1523                                 // replace all of the event code with a reformatted version.
1524
1525                                 $ev = Event::fromBBCode($text);
1526
1527                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1528                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1529
1530                                 $text = str_replace("<", "&lt;", $text);
1531                                 $text = str_replace(">", "&gt;", $text);
1532
1533                                 // remove some newlines before the general conversion
1534                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1535                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1536
1537                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1538                                 if (!$try_oembed) {
1539                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1540                                 }
1541
1542                                 // Remove linefeeds inside of the table elements. See issue #6799
1543                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1544                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1545                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1546                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1547                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1548                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1549                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1550                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1551                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1552                                         "[hr]", "[hr]", "[hr]", "[hr]",
1553                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1554                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1555                                 do {
1556                                         $oldtext = $text;
1557                                         $text = str_replace($search, $replace, $text);
1558                                 } while ($oldtext != $text);
1559
1560                                 // Replace these here only once
1561                                 $search = ["\n[table]", "[/table]\n"];
1562                                 $replace = ["[table]", "[/table]"];
1563                                 $text = str_replace($search, $replace, $text);
1564
1565                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1566                                 $text = trim($text, "\n");
1567
1568                                 // removing multiplicated newlines
1569                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1570                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[*]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1571                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1572                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1573                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1574                                         do {
1575                                                 $oldtext = $text;
1576                                                 $text = str_replace($search, $replace, $text);
1577                                         } while ($oldtext != $text);
1578                                 }
1579
1580                                 /// @todo Have a closer look at the different html modes
1581                                 // Handle attached links or videos
1582                                 if (in_array($simple_html, [self::API, self::ACTIVITYPUB])) {
1583                                         $text = self::removeAttachment($text);
1584                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1585                                         $text = self::removeAttachment($text, true);
1586                                 } else {
1587                                         $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1588                                 }
1589
1590                                 $nosmile = strpos($text, '[nosmile]') !== false;
1591                                 $text = str_replace('[nosmile]', '', $text);
1592
1593                                 // Replace non graphical smilies for external posts
1594                                 if (!$nosmile) {
1595                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1596                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1597                                         });
1598                                 }
1599
1600                                 // leave open the posibility of [map=something]
1601                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1602                                 if (strpos($text, '[/map]') !== false) {
1603                                         $text = preg_replace_callback(
1604                                                 "/\[map\](.*?)\[\/map\]/ism",
1605                                                 function ($match) use ($simple_html) {
1606                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1607                                                 },
1608                                                 $text
1609                                         );
1610                                 }
1611
1612                                 if (strpos($text, '[map=') !== false) {
1613                                         $text = preg_replace_callback(
1614                                                 "/\[map=(.*?)\]/ism",
1615                                                 function ($match) use ($simple_html) {
1616                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1617                                                 },
1618                                                 $text
1619                                         );
1620                                 }
1621
1622                                 if (strpos($text, '[map]') !== false) {
1623                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1624                                 }
1625
1626                                 // Check for headers
1627                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1628                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1629                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1630                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1631                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1632                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1633
1634                                 // Check for paragraph
1635                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1636
1637                                 // Check for bold text
1638                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1639
1640                                 // Check for Italics text
1641                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1642
1643                                 // Check for Underline text
1644                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1645
1646                                 // Check for strike-through text
1647                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1648
1649                                 // Check for over-line text
1650                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1651
1652                                 // Check for colored text
1653                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1654
1655                                 // Check for sized text
1656                                 // [size=50] --> font-size: 50px (with the unit).
1657                                 if ($simple_html != self::DIASPORA) {
1658                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1659                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1660                                 } else {
1661                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1662                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1663                                 }
1664
1665
1666                                 // Check for centered text
1667                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1668
1669                                 // Check for list text
1670                                 $text = str_replace("[*]", "<li>", $text);
1671
1672                                 // Check for block-level custom CSS
1673                                 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1674
1675                                 // Check for inline custom CSS
1676                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1677
1678                                 // Mastodon Emoji (internal tag, do not document for users)
1679                                 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1680
1681                                 // Check for CSS classes
1682                                 // @deprecated since 2021.12, left for backward-compatibility reasons
1683                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1684                                 // Add HTML new lines
1685                                 $text = str_replace("\n", '<br>', $text);
1686
1687                                 // handle nested lists
1688                                 $endlessloop = 0;
1689
1690                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1691                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1692                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1693                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1694                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1695                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1696                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1697                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1698                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1699                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1700                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1701                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1702                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1703                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1704                                 }
1705
1706                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1707                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1708                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1709                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1710
1711                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1712                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1713
1714                                 $text = str_replace('[hr]', '<hr />', $text);
1715
1716                                 if (!$for_plaintext) {
1717                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1718                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1719                                         });
1720                                 }
1721
1722                                 // Check for font change text
1723                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1724
1725                                 // Declare the format for [spoiler] layout
1726                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1727
1728                                 // Check for [spoiler] text
1729                                 // handle nested quotes
1730                                 $endlessloop = 0;
1731                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1732                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1733                                 }
1734
1735                                 // Check for [spoiler=Title] text
1736
1737                                 // handle nested quotes
1738                                 $endlessloop = 0;
1739                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1740                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1741                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1742                                                 $text);
1743                                 }
1744
1745                                 // Declare the format for [quote] layout
1746                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1747
1748                                 // Check for [quote] text
1749                                 // handle nested quotes
1750                                 $endlessloop = 0;
1751                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1752                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1753                                 }
1754
1755                                 // Check for [quote=Author] text
1756
1757                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1758
1759                                 // handle nested quotes
1760                                 $endlessloop = 0;
1761                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1762                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1763                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1764                                                 $text);
1765                                 }
1766
1767
1768                                 // [img=widthxheight]image source[/img]
1769                                 $text = preg_replace_callback(
1770                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1771                                         function ($matches) use ($simple_html, $uriid) {
1772                                                 if (strpos($matches[3], "data:image/") === 0) {
1773                                                         return $matches[0];
1774                                                 }
1775
1776                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1777                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1778                                         },
1779                                         $text
1780                                 );
1781
1782                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1783                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1784
1785                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1786                                         function ($matches) use ($simple_html, $uriid) {
1787                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1788                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1789                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1790                                         },
1791                                         $text);
1792
1793                                 // Images
1794                                 // [img]pathtoimage[/img]
1795                                 $text = preg_replace_callback(
1796                                         "/\[img\](.*?)\[\/img\]/ism",
1797                                         function ($matches) use ($simple_html, $uriid) {
1798                                                 if (strpos($matches[1], "data:image/") === 0) {
1799                                                         return $matches[0];
1800                                                 }
1801
1802                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1803                                                 return "[img]" . $matches[1] . "[/img]";
1804                                         },
1805                                         $text
1806                                 );
1807
1808                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1809                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1810
1811                                 $text = self::convertImages($text, $simple_html, $uriid);
1812
1813                                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1814                                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1815                                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $Text);
1816
1817                                 // Simplify "video" element
1818                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1819
1820                                 if ($try_oembed) {
1821                                         // html5 video and audio
1822                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1823                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1824
1825                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1826                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1827
1828                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1829                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1830                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1831                                 } else {
1832                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1833                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1834                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1835                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1836                                 }
1837
1838                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1839                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1840
1841                                 // Youtube extensions
1842                                 if ($try_oembed) {
1843                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1844                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1845                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1846                                 }
1847
1848                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1849                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1850                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1851
1852                                 if ($try_oembed) {
1853                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1854                                 } else {
1855                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1856                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1857                                 }
1858
1859                                 if ($try_oembed) {
1860                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1861                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1862                                 }
1863
1864                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1865                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1866
1867                                 if ($try_oembed) {
1868                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1869                                 } else {
1870                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1871                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1872                                 }
1873
1874                                 // oembed tag
1875                                 $text = OEmbed::BBCode2HTML($text);
1876
1877                                 // Avoid triple linefeeds through oembed
1878                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1879
1880                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1881                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1882                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1883                                 // start which is always required). Allow desc with a missing summary for compatibility.
1884
1885                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1886                                         $sub = Event::getHTML($ev, $simple_html, $uriid);
1887
1888                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1889                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1890                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1891                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1892                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1893                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1894                                 }
1895
1896                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1897                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1898                                         // Emojis are always 4 byte Unicode characters
1899                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1900                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1901                                         }
1902                                 }
1903
1904                                 // Handle mentions and hashtag links
1905                                 if ($simple_html == self::DIASPORA) {
1906                                         // The ! is converted to @ since Diaspora only understands the @
1907                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1908                                                 '@<a href="$2">$3</a>',
1909                                                 $text);
1910                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1911                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1912                                                 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1913                                                 $text);
1914                                         $text = preg_replace("/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1915                                                 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1916                                                 $text);
1917                                 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::API])) {
1918                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1919                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1920                                                 $text);
1921                                 } else {
1922                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1923                                 }
1924
1925                                 if (!$for_plaintext) {
1926                                         if (in_array($simple_html, [self::OSTATUS, self::API, self::ACTIVITYPUB])) {
1927                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1928                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1929                                         }
1930                                 } else {
1931                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1932                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1933                                 }
1934
1935                                 // Bookmarks in red - will be converted to bookmarks in friendica
1936                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1937                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1938                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1939                                                         "[bookmark=$1]$2[/bookmark]", $text);
1940
1941                                 if (in_array($simple_html, [self::OSTATUS, self::TWITTER])) {
1942                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1943                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1944                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1945                                 }
1946
1947                                 // Perform URL Search
1948                                 if ($try_oembed) {
1949                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1950                                 }
1951
1952                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1953
1954                                 // Handle Diaspora posts
1955                                 $text = preg_replace_callback(
1956                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1957                                         function ($match) {
1958                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1959                                         }, $text
1960                                 );
1961
1962                                 $text = preg_replace_callback(
1963                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1964                                         function ($match) {
1965                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1966                                         }, $text
1967                                 );
1968
1969                                 // Server independent link to posts and comments
1970                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1971                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1972                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1973
1974                                 /* Tag conversion
1975                                  * Supports:
1976                                  * - #[url=<anything>]<term>[/url]
1977                                  * - [url=<anything>]#<term>[/url]
1978                                  */
1979                                 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
1980                                         $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1981                                                 if ($simple_html == self::ACTIVITYPUB) {
1982                                                         return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1983                                                                 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1984                                                                 . XML::escape($matches[1]) . '</a>';
1985                                                 } else {
1986                                                         return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1987                                                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1988                                                                 . XML::escape($matches[1]) . '</a>';
1989                                                 }
1990                                         }, $text);
1991                                         return $text;
1992                                 });
1993
1994                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1995                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1996                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1997                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1998                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1999
2000                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
2001                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2002
2003                                 // Red compatibility, though the link can't be authenticated on Friendica
2004                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2005
2006
2007                                 // we may need to restrict this further if it picks up too many strays
2008                                 // link acct:user@host to a webfinger profile redirector
2009
2010                                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
2011
2012                                 // Perform MAIL Search
2013                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
2014                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
2015
2016                                 /// @todo What is the meaning of these lines?
2017                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
2018                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
2019
2020                                 // Currently deactivated, it made problems with " inside of alt texts.
2021                                 //$text = preg_replace('/\&quot\;/', '"', $text);
2022
2023                                 // fix any escaped ampersands that may have been converted into links
2024                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
2025
2026                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
2027                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
2028
2029                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
2030
2031                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
2032                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
2033
2034                                 // sanitize href attributes (only allowlisted protocols URLs)
2035                                 // default value for backward compatibility
2036                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
2037
2038                                 // Always allowed protocol even if config isn't set or not including it
2039                                 $allowed_link_protocols[] = '//';
2040                                 $allowed_link_protocols[] = 'http://';
2041                                 $allowed_link_protocols[] = 'https://';
2042                                 $allowed_link_protocols[] = 'redir/';
2043
2044                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
2045
2046                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
2047                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
2048
2049                                 // Shared content
2050                                 $text = self::convertShare(
2051                                         $text,
2052                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
2053                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
2054                                         }, $uriid
2055                                 );
2056
2057                                 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
2058
2059                                 return $text;
2060                         }); // Escaped noparse, nobb, pre
2061
2062                         // Remove escaping tags and replace new lines that remain
2063                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
2064                                 return str_replace("\n", "<br>", $match[2]);
2065                         }, $text);
2066
2067                         // Additionally, [pre] tags preserve spaces
2068                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
2069                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
2070                         }, $text);
2071
2072                         return $text;
2073                 }); // Escaped code
2074
2075                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
2076                         function ($matches) {
2077                                 if (strpos($matches[2], "\n") !== false) {
2078                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
2079                                 } else {
2080                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
2081                                 }
2082
2083                                 return $return;
2084                         },
2085                         $text
2086                 );
2087
2088                 // Default iframe allowed domains/path
2089                 $allowedIframeDomains = [
2090                         DI::baseUrl()->getHostname()
2091                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
2092                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
2093                         'www.youtube.com/embed/',
2094                         'player.vimeo.com/video/',
2095                 ];
2096
2097                 $allowedIframeDomains = array_merge(
2098                         $allowedIframeDomains,
2099                         DI::config()->get('system', 'allowed_oembed') ?
2100                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
2101                                 : []
2102                 );
2103
2104                 $text = HTML::purify($text, $allowedIframeDomains);
2105                 DI::profiler()->stopRecording();
2106
2107                 return trim($text);
2108         }
2109
2110         /**
2111          * Strips the "abstract" tag from the provided text
2112          *
2113          * @param string $text The text with BBCode
2114          * @return string The same text - but without "abstract" element
2115          */
2116         public static function stripAbstract(string $text): string
2117         {
2118                 DI::profiler()->startRecording('rendering');
2119
2120                 $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
2121                         $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
2122                         $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
2123                         return $text;
2124                 });
2125
2126                 DI::profiler()->stopRecording();
2127                 return $text;
2128         }
2129
2130         /**
2131          * Returns the value of the "abstract" element
2132          *
2133          * @param string $text  The text that maybe contains the element
2134          * @param string $addon The addon for which the abstract is meant for
2135          * @return string The abstract
2136          */
2137         public static function getAbstract(string $text, string $addon = ''): string
2138         {
2139                 DI::profiler()->startRecording('rendering');
2140                 $addon = strtolower($addon);
2141
2142                 $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
2143                         if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
2144                                 return $matches[1];
2145                         }
2146
2147                         if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
2148                                 return $matches[1];
2149                         }
2150
2151                         return '';
2152                 });
2153
2154                 DI::profiler()->stopRecording();
2155                 return $abstract;
2156         }
2157
2158         /**
2159          * Callback function to replace a Friendica style mention in a mention for Diaspora
2160          *
2161          * @param array $match Matching values for the callback
2162          *                     [1] = Mention type (! or @)
2163          *                     [2] = Name
2164          *                     [3] = Address
2165          * @return string Replaced mention
2166          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2167          * @throws \ImagickException
2168          */
2169         private static function bbCodeMention2DiasporaCallback(array $match): string
2170         {
2171                 $contact = Contact::getByURL($match[3], false, ['addr']);
2172                 if (empty($contact['addr'])) {
2173                         return $match[0];
2174                 }
2175
2176                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2177                 return $mention;
2178         }
2179
2180         /**
2181          * Converts a BBCode text into Markdown
2182          *
2183          * This function converts a BBCode item body to be sent to Markdown-enabled
2184          * systems like Diaspora and Libertree
2185          *
2186          * @param string $text
2187          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2188          * @return string
2189          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2190          */
2191         public static function toMarkdown(string $text, bool $for_diaspora = true): string
2192         {
2193                 DI::profiler()->startRecording('rendering');
2194                 $original_text = $text;
2195
2196                 // Since Diaspora is creating a summary for links, this function removes them before posting
2197                 if ($for_diaspora) {
2198                         $text = self::removeShareInformation($text);
2199                 }
2200
2201                 /**
2202                  * Transform #tags, strip off the [url] and replace spaces with underscore
2203                  */
2204                 $url_search_string = "^\[\]";
2205                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2206                         function ($matches) {
2207                                 return '#' . str_replace(' ', '_', $matches[2]);
2208                         },
2209                         $text
2210                 );
2211
2212                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2213                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2214
2215                 // Convert it to HTML - don't try oembed
2216                 if ($for_diaspora) {
2217                         $text = self::convert($text, false, self::DIASPORA);
2218
2219                         // Add all tags that maybe were removed
2220                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2221                                 $tagline = '';
2222                                 foreach ($tags[2] as $tag) {
2223                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2224                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2225                                                 $tagline .= '#' . $tag . ' ';
2226                                         }
2227                                 }
2228                                 $text = $text . " " . $tagline;
2229                         }
2230                 } else {
2231                         $text = self::convert($text, false, self::CONNECTORS);
2232                 }
2233
2234                 // If a link is followed by a quote then there should be a newline before it
2235                 // Maybe we should make this newline at every time before a quote.
2236                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2237
2238                 // Now convert HTML to Markdown
2239                 $text = HTML::toMarkdown($text);
2240
2241                 // Libertree has a problem with escaped hashtags.
2242                 $text = str_replace(['\#'], ['#'], $text);
2243
2244                 // Remove any leading or trailing whitespace, as this will mess up
2245                 // the Diaspora signature verification and cause the item to disappear
2246                 $text = trim($text);
2247
2248                 if ($for_diaspora) {
2249                         $url_search_string = "^\[\]";
2250                         $text = preg_replace_callback(
2251                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2252                                 ['self', 'bbCodeMention2DiasporaCallback'],
2253                                 $text
2254                         );
2255                 }
2256
2257                 Hook::callAll('bb2diaspora', $text);
2258
2259                 DI::profiler()->stopRecording();
2260                 return $text;
2261         }
2262
2263         /**
2264          * Pull out all #hashtags and @person tags from $string.
2265          *
2266          * We also get @person@domain.com - which would make
2267          * the regex quite complicated as tags can also
2268          * end a sentence. So we'll run through our results
2269          * and strip the period from any tags which end with one.
2270          * Returns array of tags found, or empty array.
2271          *
2272          * @param string $string Post content
2273          *
2274          * @return array List of tag and person names
2275          */
2276         public static function getTags(string $string): array
2277         {
2278                 DI::profiler()->startRecording('rendering');
2279                 $ret = [];
2280
2281                 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2282                         // Convert hashtag links to hashtags
2283                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2284
2285                         // Force line feeds at bbtags
2286                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2287
2288                         // ignore anything in a bbtag
2289                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2290
2291                         // Match full names against @tags including the space between first and last
2292                         // We will look these up afterward to see if they are full names or not recognisable.
2293
2294                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2295                                 foreach ($matches[1] as $match) {
2296                                         if (strstr($match, ']')) {
2297                                                 // we might be inside a bbcode color tag - leave it alone
2298                                                 continue;
2299                                         }
2300
2301                                         if (substr($match, -1, 1) === '.') {
2302                                                 $ret[] = substr($match, 0, -1);
2303                                         } else {
2304                                                 $ret[] = $match;
2305                                         }
2306                                 }
2307                         }
2308
2309                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2310                         // and #hash tags.
2311
2312                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2313                                 foreach ($matches[1] as $match) {
2314                                         if (strstr($match, ']')) {
2315                                                 // we might be inside a bbcode color tag - leave it alone
2316                                                 continue;
2317                                         }
2318
2319                                         // try not to catch url fragments
2320                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2321                                                 continue;
2322                                         }
2323
2324                                         $ret[] = $match;
2325                                 }
2326                         }
2327                 });
2328
2329                 DI::profiler()->stopRecording();
2330                 return array_unique($ret);
2331         }
2332
2333         /**
2334          * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2335          *
2336          * @param string $body HTML/BBCode
2337          * @return string body with expanded tags
2338          */
2339         public static function expandTags(string $body): string
2340         {
2341                 return preg_replace_callback("/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2342                         function ($match) {
2343                                 switch ($match[1]) {
2344                                         case '!':
2345                                         case '@':
2346                                                 $contact = Contact::getByURL($match[2]);
2347                                                 if (!empty($contact)) {
2348                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2349                                                 } else {
2350                                                         return $match[1] . $match[2];
2351                                                 }
2352                                                 break;
2353                                         case '#':
2354                                         default:
2355                                                 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2356                                 }
2357                         }, $body);
2358         }
2359
2360         /**
2361          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2362          *
2363          * @param string   $text HTML/BBCode
2364          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2365          * @param callable $callback
2366          * @return string
2367          * @see Strings::performWithEscapedBlocks
2368          */
2369         public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2370         {
2371                 $tagList = array_map('preg_quote', $tagList);
2372
2373                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2374         }
2375
2376         /**
2377          * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2378          *
2379          * @param string $body HTML/BBCode
2380          * @param int $profile_uid Profile user id
2381          * @param string $network Network name
2382          * @return string HTML/BBCode with inserted images
2383          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2384          * @throws \ImagickException
2385          */
2386         public static function setMentions(string $body, $profile_uid = 0, $network = '')
2387         {
2388                 DI::profiler()->startRecording('rendering');
2389                 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2390                         $tags = self::getTags($body);
2391
2392                         $tagged = [];
2393
2394                         foreach ($tags as $tag) {
2395                                 $tag_type = substr($tag, 0, 1);
2396
2397                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2398                                         continue;
2399                                 }
2400
2401                                 /*
2402                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2403                                  * Robert Johnson should be first in the $tags array
2404                                  */
2405                                 foreach ($tagged as $nextTag) {
2406                                         if (stristr($nextTag, $tag . ' ')) {
2407                                                 continue 2;
2408                                         }
2409                                 }
2410
2411                                 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2412                                         $tagged[] = $tag;
2413                                 }
2414                         }
2415
2416                         return $body;
2417                 });
2418
2419                 DI::profiler()->stopRecording();
2420                 return $body;
2421         }
2422
2423         /**
2424          * @param string      $author  Author display name
2425          * @param string      $profile Author profile URL
2426          * @param string      $avatar  Author profile picture URL
2427          * @param string      $link    Post source URL
2428          * @param string      $posted  Post created date
2429          * @param string|null $guid    Post guid (if any)
2430          * @return string
2431          * @TODO Rewrite to handle over whole record array
2432          */
2433         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null): string
2434         {
2435                 DI::profiler()->startRecording('rendering');
2436                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2437                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2438                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2439                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2440                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2441
2442                 if ($guid) {
2443                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2444                 }
2445
2446                 $header  .= "']";
2447
2448                 DI::profiler()->stopRecording();
2449                 return $header;
2450         }
2451
2452         /**
2453          * Returns the BBCode relevant to embed the provided URL in a post body.
2454          * For media type, it will return [img], [video] and [audio] tags.
2455          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2456          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2457          *
2458          * @param string      $url
2459          * @param bool        $tryAttachment
2460          * @param string|null $title
2461          * @param string|null $description
2462          * @param string|null $tags
2463          * @return string
2464          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2465          *@see ParseUrl::getSiteinfoCached
2466          *
2467          */
2468         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2469         {
2470                 DI::profiler()->startRecording('rendering');
2471                 DI::logger()->info($url);
2472
2473                 // If there is already some content information submitted we don't
2474                 // need to parse the url for content.
2475                 if (!empty($title) && !empty($description)) {
2476                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2477
2478                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2479
2480                         $str_tags = '';
2481                         if (!empty($tags)) {
2482                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2483                                 if (count($arr_tags)) {
2484                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2485                                 }
2486                         }
2487
2488                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2489
2490                         DI::logger()->info('(unparsed): returns: ' . $result);
2491
2492                         DI::profiler()->stopRecording();
2493                         return $result;
2494                 }
2495
2496                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2497
2498                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2499                         switch ($siteinfo['type']) {
2500                                 case 'video':
2501                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2502                                         break;
2503                                 case 'audio':
2504                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2505                                         break;
2506                                 default:
2507                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2508                                         break;
2509                         }
2510
2511                         DI::profiler()->stopRecording();
2512                         return $bbcode;
2513                 }
2514
2515                 unset($siteinfo['keywords']);
2516
2517                 // Bypass attachment if parse url for a comment
2518                 if (!$tryAttachment) {
2519                         DI::profiler()->stopRecording();
2520                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2521                 }
2522
2523                 // Format it as BBCode attachment
2524                 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2525                 DI::profiler()->stopRecording();
2526                 return $bbcode;
2527         }
2528 }