src/Content/Text/BBCode.php

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