src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2020, Friendica
   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\Tag;
  41 use Friendica\Object\Image;
  42 use Friendica\Protocol\Activity;
  43 use Friendica\Util\Images;
  44 use Friendica\Util\Map;
  45 use Friendica\Util\ParseUrl;
  46 use Friendica\Util\Proxy as ProxyUtils;
  47 use Friendica\Util\Strings;
  48 use Friendica\Util\XML;
  49
  50 class BBCode
  51 {
  52         // Update this value to the current date whenever changes are made to BBCode::convert
  53         const VERSION = '2020-12-18-video-embeds';
  54
  55         const INTERNAL = 0;
  56         const API = 2;
  57         const DIASPORA = 3;
  58         const CONNECTORS = 4;
  59         const OSTATUS = 7;
  60         const TWITTER = 8;
  61         const BACKLINK = 8;
  62         const ACTIVITYPUB = 9;
  63
  64         /**
  65          * Fetches attachment data that were generated the old way
  66          *
  67          * @param string $body Message body
  68          * @return array
  69          *                     'type' -> Message type ('link', 'video', 'photo')
  70          *                     'text' -> Text before the shared message
  71          *                     'after' -> Text after the shared message
  72          *                     'image' -> Preview image of the message
  73          *                     'url' -> Url to the attached message
  74          *                     'title' -> Title of the attachment
  75          *                     'description' -> Description of the attachment
  76          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  77          */
  78         private static function getOldAttachmentData($body)
  79         {
  80                 $post = [];
  81
  82                 // Simplify image codes
  83                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  84
  85                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  86                         foreach ($attached as $data) {
  87                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  88                                         continue;
  89                                 }
  90
  91                                 $post['type'] = substr($data[1], 5);
  92
  93                                 $pos = strpos($body, $data[0]);
  94                                 if ($pos > 0) {
  95                                         $post['text'] = trim(substr($body, 0, $pos));
  96                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  97                                 } else {
  98                                         $post['text'] = trim(str_replace($data[0], '', $body));
  99                                         $post['after'] = '';
 100                                 }
 101
 102                                 $attacheddata = $data[2];
 103
 104                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 105
 106                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 107
 108                                         if ($picturedata) {
 109                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 110                                                         $post['image'] = $matches[1];
 111                                                 } else {
 112                                                         $post['preview'] = $matches[1];
 113                                                 }
 114                                         }
 115                                 }
 116
 117                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 118                                         $post['url'] = $matches[1];
 119                                         $post['title'] = $matches[2];
 120                                 }
 121                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 122                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 123                                         $post['url'] = $matches[1];
 124                                 }
 125
 126                                 // Search for description
 127                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 128                                         $post['description'] = $matches[1];
 129                                 }
 130                         }
 131                 }
 132                 return $post;
 133         }
 134
 135         /**
 136          * Fetches attachment data that were generated with the "attachment" element
 137          *
 138          * @param string $body Message body
 139          * @return array
 140          *                     'type' -> Message type ('link', 'video', 'photo')
 141          *                     'text' -> Text before the shared message
 142          *                     'after' -> Text after the shared message
 143          *                     'image' -> Preview image of the message
 144          *                     'url' -> Url to the attached message
 145          *                     'title' -> Title of the attachment
 146          *                     'description' -> Description of the attachment
 147          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 148          */
 149         public static function getAttachmentData($body)
 150         {
 151                 $data = [
 152                         'type'          => '',
 153                         'text'          => '',
 154                         'after'         => '',
 155                         'image'         => null,
 156                         'url'           => '',
 157                         'provider_name' => '',
 158                         'provider_url'  => '',
 159                         'title'         => '',
 160                         'description'   => '',
 161                 ];
 162
 163                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 164                         return self::getOldAttachmentData($body);
 165                 }
 166
 167                 $attributes = $match[2];
 168
 169                 $data['text'] = trim($match[1]);
 170
 171                 $type = '';
 172                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 173                 if (!empty($matches[1])) {
 174                         $type = strtolower($matches[1]);
 175                 }
 176
 177                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 178                 if (!empty($matches[1])) {
 179                         $type = strtolower($matches[1]);
 180                 }
 181
 182                 if ($type == '') {
 183                         return [];
 184                 }
 185
 186                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 187                         return [];
 188                 }
 189
 190                 if ($type != '') {
 191                         $data['type'] = $type;
 192                 }
 193
 194                 $url = '';
 195                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 196                 if (!empty($matches[1])) {
 197                         $url = $matches[1];
 198                 }
 199
 200                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 201                 if (!empty($matches[1])) {
 202                         $url = $matches[1];
 203                 }
 204
 205                 if ($url != '') {
 206                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 207                 }
 208
 209                 $title = '';
 210                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 211                 if (!empty($matches[1])) {
 212                         $title = $matches[1];
 213                 }
 214
 215                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 216                 if (!empty($matches[1])) {
 217                         $title = $matches[1];
 218                 }
 219
 220                 if ($title != '') {
 221                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 222                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 223                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 224                         $data['title'] = $title;
 225                 }
 226
 227                 $image = '';
 228                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 229                 if (!empty($matches[1])) {
 230                         $image = $matches[1];
 231                 }
 232
 233                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 234                 if (!empty($matches[1])) {
 235                         $image = $matches[1];
 236                 }
 237
 238                 if ($image != '') {
 239                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 240                 }
 241
 242                 $preview = '';
 243                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 244                 if (!empty($matches[1])) {
 245                         $preview = $matches[1];
 246                 }
 247
 248                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 249                 if (!empty($matches[1])) {
 250                         $preview = $matches[1];
 251                 }
 252
 253                 if ($preview != '') {
 254                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 255                 }
 256
 257                 $data['description'] = trim($match[3]);
 258
 259                 $data['after'] = trim($match[4]);
 260
 261                 $parts = parse_url($data['url']);
 262                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 263                         $data['provider_name'] = $parts['host'];
 264                         $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 265
 266                         if (!empty($parts['port'])) {
 267                                 $data['provider_url'] .= ':' . $parts['port'];
 268                         }
 269                 }
 270
 271                 return $data;
 272         }
 273
 274         public static function getAttachedData($body, $item = [])
 275         {
 276                 /*
 277                 - text:
 278                 - type: link, video, photo
 279                 - title:
 280                 - url:
 281                 - image:
 282                 - description:
 283                 - (thumbnail)
 284                 */
 285
 286                 $has_title = !empty($item['title']);
 287                 $plink = $item['plink'] ?? '';
 288                 $post = self::getAttachmentData($body);
 289
 290                 // Get all linked images with alternative image description
 291                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 292                         foreach ($pictures as $picture) {
 293                                 if (Photo::isLocal($picture[1])) {
 294                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 295                                 }
 296                         }
 297                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 298                                 $post['image_description'] = $post['images'][0]['description'];
 299                         }
 300                 }
 301
 302                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 303                         foreach ($pictures as $picture) {
 304                                 if (Photo::isLocal($picture[1])) {
 305                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 306                                 }
 307                         }
 308                 }
 309
 310                 // if nothing is found, it maybe having an image.
 311                 if (!isset($post['type'])) {
 312                         // Simplify image codes
 313                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 314                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 315                         $post['text'] = $body;
 316
 317                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 318                                 if ((count($pictures) == 1) && !$has_title) {
 319                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 320                                                 // Replace the preview picture with the real picture
 321                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 322                                                 $data = ['url' => $url, 'type' => 'photo'];
 323                                         } else {
 324                                                 // Checking, if the link goes to a picture
 325                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 326                                         }
 327
 328                                         // Workaround:
 329                                         // Sometimes photo posts to the own album are not detected at the start.
 330                                         // So we seem to cannot use the cache for these cases. That's strange.
 331                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 332                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 333                                         }
 334
 335                                         if ($data['type'] == 'photo') {
 336                                                 $post['type'] = 'photo';
 337                                                 if (isset($data['images'][0])) {
 338                                                         $post['image'] = $data['images'][0]['src'];
 339                                                         $post['url'] = $data['url'];
 340                                                 } else {
 341                                                         $post['image'] = $data['url'];
 342                                                 }
 343
 344                                                 $post['preview'] = $pictures[0][2];
 345                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 346                                         } else {
 347                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 348                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 349                                                         $post['type'] = 'photo';
 350                                                         $post['image'] = $pictures[0][1];
 351                                                         $post['preview'] = $pictures[0][2];
 352                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 353                                                 }
 354                                         }
 355                                 } elseif (count($pictures) > 0) {
 356                                         $post['type'] = 'link';
 357                                         $post['url'] = $plink;
 358                                         $post['image'] = $pictures[0][2];
 359                                         $post['text'] = $body;
 360
 361                                         foreach ($pictures as $picture) {
 362                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 363                                         }
 364                                 }
 365                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 366                                 if ((count($pictures) == 1) && !$has_title) {
 367                                         $post['type'] = 'photo';
 368                                         $post['image'] = $pictures[0][1];
 369                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 370                                 } elseif (count($pictures) > 0) {
 371                                         $post['type'] = 'link';
 372                                         $post['url'] = $plink;
 373                                         $post['image'] = $pictures[0][1];
 374                                         $post['text'] = $body;
 375
 376                                         foreach ($pictures as $picture) {
 377                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 378                                         }
 379                                 }
 380                         }
 381
 382                         // Test for the external links
 383                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 384                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 385
 386                         $links = array_merge($links1, $links2);
 387
 388                         // If there is only a single one, then use it.
 389                         // This should cover link posts via API.
 390                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 391                                 $post['type'] = 'link';
 392                                 $post['url'] = $links[0][1];
 393                         }
 394
 395                         // Simplify "video" element
 396                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 397
 398                         // Now count the number of external media links
 399                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 400                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 401                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 402                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 403
 404                         // Add them to the other external links
 405                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 406
 407                         // Are there more than one?
 408                         if (count($links) > 1) {
 409                                 // The post will be the type "text", which means a blog post
 410                                 unset($post['type']);
 411                                 $post['url'] = $plink;
 412                         }
 413
 414                         if (!isset($post['type'])) {
 415                                 $post['type'] = "text";
 416                                 $post['text'] = trim($body);
 417                         }
 418                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 419                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 420
 421                         if (isset($data['images'][0])) {
 422                                 $post['image'] = $data['images'][0]['src'];
 423                         }
 424                 }
 425
 426                 return $post;
 427         }
 428
 429         /**
 430          * Remove [attachment] BBCode and replaces it with a regular [url]
 431          *
 432          * @param string  $body
 433          * @param boolean $no_link_desc No link description
 434          *
 435          * @return string with replaced body
 436          */
 437         public static function removeAttachment($body, $no_link_desc = false)
 438         {
 439                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 440                         function ($match) use ($no_link_desc) {
 441                                 $attach_data = self::getAttachmentData($match[0]);
 442                                 if (empty($attach_data['url'])) {
 443                                         return $match[0];
 444                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 445                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 446                                 } else {
 447                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 448                                 }
 449                 }, $body);
 450         }
 451
 452         /**
 453          * Converts a BBCode text into plaintext
 454          *
 455          * @param      $text
 456          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 457          *
 458          * @return string
 459          */
 460         public static function toPlaintext($text, $keep_urls = true)
 461         {
 462                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 463
 464                 return $naked_text;
 465         }
 466
 467         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 468         {
 469                 // Only send proxied pictures to API and for internal display
 470                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 471                         return ProxyUtils::proxifyUrl($image);
 472                 } else {
 473                         return $image;
 474                 }
 475         }
 476
 477         /**
 478          * This function changing the visual size (not the real size) of images.
 479          * The function does not work for pictures with an alternate text description.
 480          * This could only be changed by using some new "img" BBCode format.
 481          *
 482          * @param string $srctext The body with images
 483          * @return string The body with possibly scaled images
 484          */
 485         public static function scaleExternalImages(string $srctext)
 486         {
 487                 $s = $srctext;
 488
 489                 // Simplify image links
 490                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 491
 492                 $matches = null;
 493                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 494                 if ($c) {
 495                         foreach ($matches as $mtch) {
 496                                 Logger::log('scale_external_image: ' . $mtch[1]);
 497
 498                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 499                                 if (stristr($mtch[1], $hostname)) {
 500                                         continue;
 501                                 }
 502
 503                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 504                                 if (!$curlResult->isSuccess()) {
 505                                         continue;
 506                                 }
 507
 508                                 $i = $curlResult->getBody();
 509                                 $type = $curlResult->getContentType();
 510                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 511
 512                                 if ($i) {
 513                                         $Image = new Image($i, $type);
 514                                         if ($Image->isValid()) {
 515                                                 $orig_width = $Image->getWidth();
 516                                                 $orig_height = $Image->getHeight();
 517
 518                                                 if ($orig_width > 640 || $orig_height > 640) {
 519                                                         $Image->scaleDown(640);
 520                                                         $new_width = $Image->getWidth();
 521                                                         $new_height = $Image->getHeight();
 522                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 523                                                         $s = str_replace(
 524                                                                 $mtch[0],
 525                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 526                                                                 . "\n",
 527                                                                 $s
 528                                                         );
 529                                                         Logger::info('New string', ['image' => $s]);
 530                                                 }
 531                                         }
 532                                 }
 533                         }
 534                 }
 535
 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                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 552
 553                 // If the length of the body, including the embedded images, is smaller
 554                 // than the maximum, then don't waste time looking for the images
 555                 if ($maxlen && (strlen($body) > $maxlen)) {
 556
 557                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 558
 559                         $orig_body = $body;
 560                         $new_body = '';
 561                         $textlen = 0;
 562
 563                         $img_start = strpos($orig_body, '[img');
 564                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 565                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 566                         while (($img_st_close !== false) && ($img_end !== false)) {
 567
 568                                 $img_st_close++; // make it point to AFTER the closing bracket
 569                                 $img_end += $img_start;
 570                                 $img_end += strlen('[/img]');
 571
 572                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 573                                         // This is an embedded image
 574
 575                                         if (($textlen + $img_start) > $maxlen) {
 576                                                 if ($textlen < $maxlen) {
 577                                                         Logger::info('the limit happens before an embedded image');
 578                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 579                                                         $textlen = $maxlen;
 580                                                 }
 581                                         } else {
 582                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 583                                                 $textlen += $img_start;
 584                                         }
 585
 586                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 587                                 } else {
 588
 589                                         if (($textlen + $img_end) > $maxlen) {
 590                                                 if ($textlen < $maxlen) {
 591                                                         Logger::info('the limit happens before the end of a non-embedded image');
 592                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 593                                                         $textlen = $maxlen;
 594                                                 }
 595                                         } else {
 596                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 597                                                 $textlen += $img_end;
 598                                         }
 599                                 }
 600                                 $orig_body = substr($orig_body, $img_end);
 601
 602                                 if ($orig_body === false) {
 603                                         // in case the body ends on a closing image tag
 604                                         $orig_body = '';
 605                                 }
 606
 607                                 $img_start = strpos($orig_body, '[img');
 608                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 609                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 610                         }
 611
 612                         if (($textlen + strlen($orig_body)) > $maxlen) {
 613                                 if ($textlen < $maxlen) {
 614                                         Logger::info('the limit happens after the end of the last image');
 615                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 616                                 }
 617                         } else {
 618                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 619                                 $new_body = $new_body . $orig_body;
 620                         }
 621
 622                         return $new_body;
 623                 } else {
 624                         return $body;
 625                 }
 626         }
 627
 628         /**
 629          * Processes [attachment] tags
 630          *
 631          * Note: Can produce a [bookmark] tag in the returned string
 632          *
 633          * @param string  $text
 634          * @param integer $simplehtml
 635          * @param bool    $tryoembed
 636          * @return string
 637          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 638          */
 639         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 640         {
 641                 $data = self::getAttachmentData($text);
 642                 if (empty($data) || empty($data['url'])) {
 643                         return $text;
 644                 }
 645
 646                 if (isset($data['title'])) {
 647                         $data['title'] = strip_tags($data['title']);
 648                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 649                 } else {
 650                         $data['title'] = null;
 651                 }
 652
 653                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 654                         $data['preview'] = $data['image'];
 655                         $data['image'] = '';
 656                 }
 657
 658                 $return = '';
 659                 try {
 660                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 661                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 662                         } else {
 663                                 throw new Exception('OEmbed is disabled for this attachment.');
 664                         }
 665                 } catch (Exception $e) {
 666                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 667
 668                         if ($simplehtml != self::CONNECTORS) {
 669                                 $return = sprintf('<div class="type-%s">', $data['type']);
 670                         }
 671
 672                         if (!empty($data['title']) && !empty($data['url'])) {
 673                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 674                                         $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), $data['title']);
 675                                 } else {
 676                                         if (!empty($data['image'])) {
 677                                                 $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), $data['title']);
 678                                         } elseif (!empty($data['preview'])) {
 679                                                 $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), $data['title']);
 680                                         }
 681                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 682                                 }
 683                         }
 684
 685                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 686                                 // Sanitize the HTML by converting it to BBCode
 687                                 $bbcode = HTML::toBBCode($data['description']);
 688                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 689                         }
 690
 691                         if (!empty($data['url'])) {
 692                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 693                         }
 694
 695                         if ($simplehtml != self::CONNECTORS) {
 696                                 $return .= '</div>';
 697                         }
 698                 }
 699
 700                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 701         }
 702
 703         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 704         {
 705                 $data = self::getAttachmentData($Text);
 706
 707                 if (!$data) {
 708                         return $Text;
 709                 } elseif ($nolink) {
 710                         return $data['text'] . ($data['after'] ?? '');
 711                 }
 712
 713                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 714                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 715                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 716                         $data['title'] = $data['url'];
 717                 } elseif (($text != '') && strstr($title, $text)) {
 718                         $data['text'] = $data['title'];
 719                         $data['title'] = $data['url'];
 720                 }
 721
 722                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 723                         return $data['title'] . $data['after'];
 724                 }
 725
 726                 // If the link already is included in the post, don't add it again
 727                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 728                         return $data['text'] . $data['after'];
 729                 }
 730
 731                 $text = $data['text'];
 732
 733                 if (!empty($data['url']) && !empty($data['title'])) {
 734                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 735                 } elseif (!empty($data['url'])) {
 736                         $text .= "\n[url]" . $data['url'] . '[/url]';
 737                 }
 738
 739                 return $text . "\n" . $data['after'];
 740         }
 741
 742         /**
 743          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 744          *
 745          * @param array $match Array with the matching values
 746          * @return string reformatted link including HTML codes
 747          */
 748         private static function convertUrlForActivityPubCallback($match)
 749         {
 750                 $url = $match[1];
 751
 752                 if (isset($match[2]) && ($match[1] != $match[2])) {
 753                         return $match[0];
 754                 }
 755
 756                 $parts = parse_url($url);
 757                 if (!isset($parts['scheme'])) {
 758                         return $match[0];
 759                 }
 760
 761                 return self::convertUrlForActivityPub($url);
 762         }
 763
 764         /**
 765          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 766          *
 767          * @param string $url URL that is about to be reformatted
 768          * @return string reformatted link including HTML codes
 769          */
 770         private static function convertUrlForActivityPub($url)
 771         {
 772                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 773                 return sprintf($html, $url, self::getStyledURL($url));
 774         }
 775
 776         /**
 777          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 778          *
 779          * @param string $url URL that is about to be reformatted
 780          * @return string reformatted link
 781          */
 782         private static function getStyledURL($url)
 783         {
 784                 $parts = parse_url($url);
 785                 $scheme = $parts['scheme'] . '://';
 786                 $styled_url = str_replace($scheme, '', $url);
 787
 788                 if (strlen($styled_url) > 30) {
 789                         $styled_url = substr($styled_url, 0, 30) . "…";
 790                 }
 791
 792                 return $styled_url;
 793         }
 794
 795         /*
 796          * [noparse][i]italic[/i][/noparse] turns into
 797          * [noparse][ i ]italic[ /i ][/noparse],
 798          * to hide them from parser.
 799          */
 800         private static function escapeNoparseCallback($match)
 801         {
 802                 $whole_match = $match[0];
 803                 $captured = $match[1];
 804                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 805                 $new_str = str_replace($captured, $spacefied, $whole_match);
 806                 return $new_str;
 807         }
 808
 809         /*
 810          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 811          * now turns back and the [noparse] tags are trimed
 812          * returning [i]italic[/i]
 813          */
 814         private static function unescapeNoparseCallback($match)
 815         {
 816                 $captured = $match[1];
 817                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 818                 return $unspacefied;
 819         }
 820
 821         /**
 822          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 823          * occurrences
 824          *
 825          * @param string $text        Text to search
 826          * @param string $name        Tag name
 827          * @param int    $occurrences Number of first occurrences to skip
 828          * @return boolean|array
 829          */
 830         public static function getTagPosition($text, $name, $occurrences = 0)
 831         {
 832                 if ($occurrences < 0) {
 833                         $occurrences = 0;
 834                 }
 835
 836                 $start_open = -1;
 837                 for ($i = 0; $i <= $occurrences; $i++) {
 838                         if ($start_open !== false) {
 839                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 840                         }
 841                 }
 842
 843                 if ($start_open === false) {
 844                         return false;
 845                 }
 846
 847                 $start_equal = strpos($text, '=', $start_open);
 848                 $start_close = strpos($text, ']', $start_open);
 849
 850                 if ($start_close === false) {
 851                         return false;
 852                 }
 853
 854                 $start_close++;
 855
 856                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 857
 858                 if ($end_open === false) {
 859                         return false;
 860                 }
 861
 862                 $res = [
 863                         'start' => [
 864                                 'open' => $start_open,
 865                                 'close' => $start_close
 866                         ],
 867                         'end' => [
 868                                 'open' => $end_open,
 869                                 'close' => $end_open + strlen('[/' . $name . ']')
 870                         ],
 871                 ];
 872
 873                 if ($start_equal !== false) {
 874                         $res['start']['equal'] = $start_equal + 1;
 875                 }
 876
 877                 return $res;
 878         }
 879
 880         /**
 881          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 882          *
 883          * @param string $pattern Preg pattern string
 884          * @param string $replace Preg replace string
 885          * @param string $name    BBCode tag name
 886          * @param string $text    Text to search
 887          * @return string
 888          */
 889         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 890         {
 891                 $occurrences = 0;
 892                 $pos = self::getTagPosition($text, $name, $occurrences);
 893                 while ($pos !== false && $occurrences++ < 1000) {
 894                         $start = substr($text, 0, $pos['start']['open']);
 895                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 896                         $end = substr($text, $pos['end']['close']);
 897                         if ($end === false) {
 898                                 $end = '';
 899                         }
 900
 901                         $subject = preg_replace($pattern, $replace, $subject);
 902                         $text = $start . $subject . $end;
 903
 904                         $pos = self::getTagPosition($text, $name, $occurrences);
 905                 }
 906
 907                 return $text;
 908         }
 909
 910         private static function extractImagesFromItemBody($body)
 911         {
 912                 $saved_image = [];
 913                 $orig_body = $body;
 914                 $new_body = '';
 915
 916                 $cnt = 0;
 917                 $img_start = strpos($orig_body, '[img');
 918                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 919                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 920                 while (($img_st_close !== false) && ($img_end !== false)) {
 921                         $img_st_close++; // make it point to AFTER the closing bracket
 922                         $img_end += $img_start;
 923
 924                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 925                                 // This is an embedded image
 926                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 927                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 928
 929                                 $cnt++;
 930                         } else {
 931                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 932                         }
 933
 934                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 935
 936                         if ($orig_body === false) {
 937                                 // in case the body ends on a closing image tag
 938                                 $orig_body = '';
 939                         }
 940
 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                 }
 945
 946                 $new_body = $new_body . $orig_body;
 947
 948                 return ['body' => $new_body, 'images' => $saved_image];
 949         }
 950
 951         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 952         {
 953                 $newbody = $body;
 954
 955                 $cnt = 0;
 956                 foreach ($images as $image) {
 957                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 958                         // it loops over the array starting from the first element and going sequentially
 959                         // to the last element
 960                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 961                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 962                         $cnt++;
 963                 }
 964
 965                 return $newbody;
 966         }
 967
 968         /**
 969          * This function converts a [share] block to text according to a provided callback function whose signature is:
 970          *
 971          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 972          *
 973          * Where:
 974          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 975          * data lookup
 976          * - $author_contact is a contact record array
 977          * - $content is the inner content of the [share] block
 978          * - $is_quote_share indicates whether there's any content before the [share] block
 979          * - Return value is the string that should replace the [share] block in the provided text
 980          *
 981          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 982          *
 983          * @param  string   $text     A BBCode string
 984          * @param  callable $callback
 985          * @return string The BBCode string with all [share] blocks replaced
 986          */
 987         public static function convertShare($text, callable $callback)
 988         {
 989                 $return = preg_replace_callback(
 990                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 991                         function ($match) use ($callback) {
 992                                 $attribute_string = $match[2];
 993                                 $attributes = [];
 994                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 995                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 996                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 997                                 }
 998
 999                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
1000                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1001                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1002
1003                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1004                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1005                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1006
1007                                 if ($attributes['avatar']) {
1008                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1009                                 }
1010
1011                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1012
1013                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1014                         },
1015                         $text
1016                 );
1017
1018                 return $return;
1019         }
1020
1021         /**
1022          * Default [share] tag conversion callback
1023          *
1024          * Note: Can produce a [bookmark] tag in the output
1025          *
1026          * @see BBCode::convertShare()
1027          * @param array   $attributes     [share] block attribute values
1028          * @param array   $author_contact Contact row of the shared author
1029          * @param string  $content        Inner content of the [share] block
1030          * @param boolean $is_quote_share Whether there is content before the [share] block
1031          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1032          * @return string
1033          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1034          */
1035         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1036         {
1037                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1038
1039                 switch ($simplehtml) {
1040                         case self::API:
1041                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1042                                 break;
1043                         case self::DIASPORA:
1044                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1045                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1046                                 } else {
1047                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1048
1049                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1050                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1051                                         }
1052
1053                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1054
1055                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1056                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1057                                         }
1058                                 }
1059
1060                                 break;
1061                         case self::CONNECTORS:
1062                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1063                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1064                                 $headline .= ':</b></p>' . "\n";
1065
1066                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1067
1068                                 break;
1069                         case self::OSTATUS:
1070                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1071                                 break;
1072                         case self::ACTIVITYPUB:
1073                                 $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>';
1074                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1075                                 break;
1076                         default:
1077                                 $text = ($is_quote_share? "\n" : '');
1078
1079                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1080                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1081
1082                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1083                                 $text .= Renderer::replaceMacros($tpl, [
1084                                         '$profile'      => $attributes['profile'],
1085                                         '$avatar'       => $attributes['avatar'],
1086                                         '$author'       => $attributes['author'],
1087                                         '$link'         => $attributes['link'],
1088                                         '$link_title'   => DI::l10n()->t('link to source'),
1089                                         '$posted'       => $attributes['posted'],
1090                                         '$guid'         => $attributes['guid'],
1091                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1092                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1093                                         '$content'      => self::setMentions(trim($content), 0, $network),
1094                                 ]);
1095                                 break;
1096                 }
1097
1098                 return $text;
1099         }
1100
1101         private static function removePictureLinksCallback($match)
1102         {
1103                 $cache_key = 'remove:' . $match[1];
1104                 $text = DI::cache()->get($cache_key);
1105
1106                 if (is_null($text)) {
1107                         $a = DI::app();
1108
1109                         $stamp1 = microtime(true);
1110
1111                         $ch = @curl_init($match[1]);
1112                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1113                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1114                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1115                         @curl_exec($ch);
1116                         $curl_info = @curl_getinfo($ch);
1117
1118                         DI::profiler()->saveTimestamp($stamp1, "network");
1119
1120                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1121                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1122                         } else {
1123                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1124
1125                                 // if its not a picture then look if its a page that contains a picture link
1126                                 $body = DI::httpRequest()->fetch($match[1]);
1127
1128                                 $doc = new DOMDocument();
1129                                 @$doc->loadHTML($body);
1130                                 $xpath = new DOMXPath($doc);
1131                                 $list = $xpath->query("//meta[@name]");
1132                                 foreach ($list as $node) {
1133                                         $attr = [];
1134
1135                                         if ($node->attributes->length) {
1136                                                 foreach ($node->attributes as $attribute) {
1137                                                         $attr[$attribute->name] = $attribute->value;
1138                                                 }
1139                                         }
1140
1141                                         if (strtolower($attr['name']) == 'twitter:image') {
1142                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1143                                         }
1144                                 }
1145                         }
1146                         DI::cache()->set($cache_key, $text);
1147                 }
1148
1149                 return $text;
1150         }
1151
1152         private static function expandLinksCallback($match)
1153         {
1154                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1155                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1156                 } else {
1157                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1158                 }
1159         }
1160
1161         private static function cleanPictureLinksCallback($match)
1162         {
1163                 $a = DI::app();
1164
1165                 // When the picture link is the own photo path then we can avoid fetching the link
1166                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1167                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1168                         if (!empty($match[3])) {
1169                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1170                         } else {
1171                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1172                         }
1173                         return $text;
1174                 }
1175
1176                 $cache_key = 'clean:' . $match[1];
1177                 $text = DI::cache()->get($cache_key);
1178                 if (!is_null($text)) {
1179                         return $text;
1180                 }
1181
1182                 // Only fetch the header, not the content
1183                 $stamp1 = microtime(true);
1184
1185                 $ch = @curl_init($match[1]);
1186                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1187                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1188                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1189                 @curl_exec($ch);
1190                 $curl_info = @curl_getinfo($ch);
1191
1192                 DI::profiler()->saveTimestamp($stamp1, "network");
1193
1194                 // if its a link to a picture then embed this picture
1195                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1196                         $text = '[img]' . $match[1] . '[/img]';
1197                 } else {
1198                         if (!empty($match[3])) {
1199                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1200                         } else {
1201                                 $text = '[img]' . $match[2] . '[/img]';
1202                         }
1203
1204                         // if its not a picture then look if its a page that contains a picture link
1205                         $body = DI::httpRequest()->fetch($match[1]);
1206
1207                         $doc = new DOMDocument();
1208                         @$doc->loadHTML($body);
1209                         $xpath = new DOMXPath($doc);
1210                         $list = $xpath->query("//meta[@name]");
1211                         foreach ($list as $node) {
1212                                 $attr = [];
1213                                 if ($node->attributes->length) {
1214                                         foreach ($node->attributes as $attribute) {
1215                                                 $attr[$attribute->name] = $attribute->value;
1216                                         }
1217                                 }
1218
1219                                 if (strtolower($attr['name']) == "twitter:image") {
1220                                         if (!empty($match[3])) {
1221                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1222                                         } else {
1223                                                 $text = "[img]" . $attr['content'] . "[/img]";
1224                                         }
1225                                 }
1226                         }
1227                 }
1228                 DI::cache()->set($cache_key, $text);
1229
1230                 return $text;
1231         }
1232
1233         public static function cleanPictureLinks($text)
1234         {
1235                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1236                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1237                 return $return;
1238         }
1239
1240         public static function removeLinks(string $bbcode)
1241         {
1242                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1243                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1244
1245                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1246                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1247                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1248                 return $bbcode;
1249         }
1250
1251         /**
1252          * Converts a BBCode message to HTML message
1253          *
1254          * BBcode 2 HTML was written by WAY2WEB.net
1255          * extended to work with Mistpark/Friendica - Mike Macgirvin
1256          *
1257          * Simple HTML values meaning:
1258          * - 0: Friendica display
1259          * - 1: Unused
1260          * - 2: Used for Windows Phone push, Friendica API
1261          * - 3: Used before converting to Markdown in bb2diaspora.php
1262          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1263          * - 5: Unused
1264          * - 6: Unused
1265          * - 7: Used for dfrn, OStatus
1266          * - 8: Used for Twitter, WP backlink text setting
1267          * - 9: ActivityPub
1268          *
1269          * @param string $text
1270          * @param bool   $try_oembed
1271          * @param int    $simple_html
1272          * @param bool   $for_plaintext
1273          * @return string
1274          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1275          */
1276         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1277         {
1278                 // Accounting for null default column values
1279                 if (is_null($text) || $text === '') {
1280                         return '';
1281                 }
1282
1283                 Hook::callAll('bbcode', $text);
1284
1285                 $a = DI::app();
1286
1287                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1288                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1289                                 /*
1290                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1291                                  *
1292                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1293                                  * $match[1] = $url
1294                                  * $match[2] = $title or absent
1295                                  */
1296                                 $try_oembed_callback = function ($match)
1297                                 {
1298                                         $url = $match[1];
1299                                         $title = $match[2] ?? null;
1300
1301                                         try {
1302                                                 $return = OEmbed::getHTML($url, $title);
1303                                         } catch (Exception $ex) {
1304                                                 $return = $match[0];
1305                                         }
1306
1307                                         return $return;
1308                                 };
1309
1310                                 // Remove the abstract element. It is a non visible element.
1311                                 $text = self::stripAbstract($text);
1312
1313                                 // Line ending normalisation
1314                                 $text = str_replace("\r\n", "\n", $text);
1315
1316                                 // Move new lines outside of tags
1317                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1318                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1319
1320                                 // Extract the private images which use data urls since preg has issues with
1321                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1322                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1323
1324                                 $extracted = self::extractImagesFromItemBody($text);
1325                                 $text = $extracted['body'];
1326                                 $saved_image = $extracted['images'];
1327
1328                                 // If we find any event code, turn it into an event.
1329                                 // After we're finished processing the bbcode we'll
1330                                 // replace all of the event code with a reformatted version.
1331
1332                                 $ev = Event::fromBBCode($text);
1333
1334                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1335                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1336
1337                                 $text = str_replace("<", "&lt;", $text);
1338                                 $text = str_replace(">", "&gt;", $text);
1339
1340                                 // remove some newlines before the general conversion
1341                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1342                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1343
1344                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1345                                 if (!$try_oembed) {
1346                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1347                                 }
1348
1349                                 // Remove linefeeds inside of the table elements. See issue #6799
1350                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1351                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1352                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1353                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1354                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1355                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1356                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1357                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1358                                 do {
1359                                         $oldtext = $text;
1360                                         $text = str_replace($search, $replace, $text);
1361                                 } while ($oldtext != $text);
1362
1363                                 // Replace these here only once
1364                                 $search = ["\n[table]", "[/table]\n"];
1365                                 $replace = ["[table]", "[/table]"];
1366                                 $text = str_replace($search, $replace, $text);
1367
1368                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1369                                 $text = trim($text, "\n");
1370
1371                                 // removing multiplicated newlines
1372                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1373                                         $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",
1374                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1375                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1376                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1377                                         do {
1378                                                 $oldtext = $text;
1379                                                 $text = str_replace($search, $replace, $text);
1380                                         } while ($oldtext != $text);
1381                                 }
1382
1383                                 // Add HTML new lines
1384                                 $text = str_replace("\n", '<br>', $text);
1385
1386                                 /// @todo Have a closer look at the different html modes
1387                                 // Handle attached links or videos
1388                                 if ($simple_html == self::ACTIVITYPUB) {
1389                                         $text = self::removeAttachment($text);
1390                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1391                                         $text = self::removeAttachment($text, true);
1392                                 } else {
1393                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1394                                 }
1395
1396                                 $nosmile = strpos($text, '[nosmile]') !== false;
1397                                 $text = str_replace('[nosmile]', '', $text);
1398
1399                                 // Replace non graphical smilies for external posts
1400                                 if (!$nosmile && !$for_plaintext) {
1401                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) {
1402                                                 return Smilies::replace($text);
1403                                         });
1404                                 }
1405
1406                                 // leave open the posibility of [map=something]
1407                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1408                                 if (strpos($text, '[/map]') !== false) {
1409                                         $text = preg_replace_callback(
1410                                                 "/\[map\](.*?)\[\/map\]/ism",
1411                                                 function ($match) use ($simple_html) {
1412                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1413                                                 },
1414                                                 $text
1415                                         );
1416                                 }
1417
1418                                 if (strpos($text, '[map=') !== false) {
1419                                         $text = preg_replace_callback(
1420                                                 "/\[map=(.*?)\]/ism",
1421                                                 function ($match) use ($simple_html) {
1422                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1423                                                 },
1424                                                 $text
1425                                         );
1426                                 }
1427
1428                                 if (strpos($text, '[map]') !== false) {
1429                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1430                                 }
1431
1432                                 // Check for headers
1433                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1434                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1435                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1436                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1437                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1438                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1439
1440                                 // Check for paragraph
1441                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1442
1443                                 // Check for bold text
1444                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1445
1446                                 // Check for Italics text
1447                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1448
1449                                 // Check for Underline text
1450                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1451
1452                                 // Check for strike-through text
1453                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1454
1455                                 // Check for over-line text
1456                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1457
1458                                 // Check for colored text
1459                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1460
1461                                 // Check for sized text
1462                                 // [size=50] --> font-size: 50px (with the unit).
1463                                 if ($simple_html != self::DIASPORA) {
1464                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1465                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1466                                 } else {
1467                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1468                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1469                                 }
1470
1471
1472                                 // Check for centered text
1473                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1474
1475                                 // Check for list text
1476                                 $text = str_replace("[*]", "<li>", $text);
1477
1478                                 // Check for style sheet commands
1479                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1480
1481                                 // Check for CSS classes
1482                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1483
1484                                 // handle nested lists
1485                                 $endlessloop = 0;
1486
1487                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1488                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1489                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1490                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1491                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1492                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1493                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1494                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1495                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1496                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1497                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1498                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1499                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1500                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1501                                 }
1502
1503                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1504                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1505                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1506                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1507
1508                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1509                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1510
1511                                 $text = str_replace('[hr]', '<hr />', $text);
1512
1513                                 if (!$for_plaintext) {
1514                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1515                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1516                                         });
1517                                 }
1518
1519                                 // Check for font change text
1520                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1521
1522                                 // Declare the format for [spoiler] layout
1523                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1524
1525                                 // Check for [spoiler] text
1526                                 // handle nested quotes
1527                                 $endlessloop = 0;
1528                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1529                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1530                                 }
1531
1532                                 // Check for [spoiler=Title] text
1533
1534                                 // handle nested quotes
1535                                 $endlessloop = 0;
1536                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1537                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1538                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1539                                                 $text);
1540                                 }
1541
1542                                 // Declare the format for [quote] layout
1543                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1544
1545                                 // Check for [quote] text
1546                                 // handle nested quotes
1547                                 $endlessloop = 0;
1548                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1549                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1550                                 }
1551
1552                                 // Check for [quote=Author] text
1553
1554                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1555
1556                                 // handle nested quotes
1557                                 $endlessloop = 0;
1558                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1559                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1560                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1561                                                 $text);
1562                                 }
1563
1564
1565                                 // [img=widthxheight]image source[/img]
1566                                 $text = preg_replace_callback(
1567                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1568                                         function ($matches) use ($simple_html) {
1569                                                 if (strpos($matches[3], "data:image/") === 0) {
1570                                                         return $matches[0];
1571                                                 }
1572
1573                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1574                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1575                                         },
1576                                         $text
1577                                 );
1578
1579                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1580                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1581
1582                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1583                                         function ($matches) use ($simple_html) {
1584                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1585                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1586                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1587                                         },
1588                                         $text);
1589
1590                                 // Images
1591                                 // [img]pathtoimage[/img]
1592                                 $text = preg_replace_callback(
1593                                         "/\[img\](.*?)\[\/img\]/ism",
1594                                         function ($matches) use ($simple_html) {
1595                                                 if (strpos($matches[1], "data:image/") === 0) {
1596                                                         return $matches[0];
1597                                                 }
1598
1599                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1600                                                 return "[img]" . $matches[1] . "[/img]";
1601                                         },
1602                                         $text
1603                                 );
1604
1605                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1606                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1607
1608                                 $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);
1609                                 $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>', $text);
1610                                 //$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);
1611
1612                                 // Simplify "video" element
1613                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1614
1615                                 if ($try_oembed) {
1616                                         // html5 video and audio
1617                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1618                                                 '<video src="$1" controls width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1619
1620                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1621                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1622
1623                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1624                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1625                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1626                                 } else {
1627                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1628                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1629                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1630                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1631                                 }
1632
1633                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1634                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1635
1636                                 // Youtube extensions
1637                                 if ($try_oembed) {
1638                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1639                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1640                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1641                                 }
1642
1643                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1644                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1645                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1646
1647                                 if ($try_oembed) {
1648                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1649                                 } else {
1650                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1651                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1652                                 }
1653
1654                                 if ($try_oembed) {
1655                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1656                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1657                                 }
1658
1659                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1660                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1661
1662                                 if ($try_oembed) {
1663                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1664                                 } else {
1665                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1666                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1667                                 }
1668
1669                                 // oembed tag
1670                                 $text = OEmbed::BBCode2HTML($text);
1671
1672                                 // Avoid triple linefeeds through oembed
1673                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1674
1675                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1676                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1677                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1678                                 // start which is always required). Allow desc with a missing summary for compatibility.
1679
1680                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1681                                         $sub = Event::getHTML($ev, $simple_html);
1682
1683                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1684                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1685                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1686                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1687                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1688                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1689                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1690                                 }
1691
1692                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1693                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1694                                         // Emojis are always 4 byte Unicode characters
1695                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1696                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1697                                         }
1698                                 }
1699
1700                                 if (!$for_plaintext) {
1701                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1702                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1703                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1704                                         }
1705                                 } else {
1706                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1707                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1708                                 }
1709
1710                                 // Remove all hashtag addresses
1711                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1712                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1713                                 } elseif ($simple_html == self::DIASPORA) {
1714                                         // The ! is converted to @ since Diaspora only understands the @
1715                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1716                                                 '@<a href="$2">$3</a>',
1717                                                 $text);
1718                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1719                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1720                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1721                                                 $text);
1722                                 } elseif (!$simple_html) {
1723                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1724                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1725                                                 $text);
1726                                 }
1727
1728                                 // Bookmarks in red - will be converted to bookmarks in friendica
1729                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1730                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1731                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1732                                                         "[bookmark=$1]$2[/bookmark]", $text);
1733
1734                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1735                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1736                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1737                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1738                                 }
1739
1740                                 // Perform URL Search
1741                                 if ($try_oembed) {
1742                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1743                                 }
1744
1745                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1746
1747                                 // Handle Diaspora posts
1748                                 $text = preg_replace_callback(
1749                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1750                                         function ($match) {
1751                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1752                                         }, $text
1753                                 );
1754
1755                                 $text = preg_replace_callback(
1756                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1757                                         function ($match) {
1758                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1759                                         }, $text
1760                                 );
1761
1762                                 // Server independent link to posts and comments
1763                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1764                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1765                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1766
1767                                 /* Tag conversion
1768                                  * Supports:
1769                                  * - #[url=<anything>]<term>[/url]
1770                                  * - [url=<anything>]#<term>[/url]
1771                                  */
1772                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1773                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1774                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1775                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1776                                                         . XML::escape($matches[1]) . '</a>';
1777                                         } else {
1778                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1779                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1780                                                         . XML::escape($matches[1]) . '</a>';
1781                                         }
1782                                 }, $text);
1783
1784                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1785                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1786                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1787                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1788                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1789
1790                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1791                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1792
1793                                 // Red compatibility, though the link can't be authenticated on Friendica
1794                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1795
1796
1797                                 // we may need to restrict this further if it picks up too many strays
1798                                 // link acct:user@host to a webfinger profile redirector
1799
1800                                 $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);
1801
1802                                 // Perform MAIL Search
1803                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1804                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1805
1806                                 /// @todo What is the meaning of these lines?
1807                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1808                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1809
1810                                 // Currently deactivated, it made problems with " inside of alt texts.
1811                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1812
1813                                 // fix any escaped ampersands that may have been converted into links
1814                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1815
1816                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1817                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1818
1819                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1820
1821                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1822                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1823
1824                                 // sanitize href attributes (only allowlisted protocols URLs)
1825                                 // default value for backward compatibility
1826                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1827
1828                                 // Always allowed protocol even if config isn't set or not including it
1829                                 $allowed_link_protocols[] = '//';
1830                                 $allowed_link_protocols[] = 'http://';
1831                                 $allowed_link_protocols[] = 'https://';
1832                                 $allowed_link_protocols[] = 'redir/';
1833
1834                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1835
1836                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1837                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1838
1839                                 // Shared content
1840                                 $text = self::convertShare(
1841                                         $text,
1842                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1843                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1844                                         }
1845                                 );
1846
1847                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1848
1849                                 return $text;
1850                         }); // Escaped noparse, nobb, pre
1851
1852                         // Remove escaping tags and replace new lines that remain
1853                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1854                                 return str_replace("\n", "<br>", $match[2]);
1855                         }, $text);
1856
1857                         // Additionally, [pre] tags preserve spaces
1858                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1859                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1860                         }, $text);
1861
1862                         return $text;
1863                 }); // Escaped code
1864
1865                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1866                         function ($matches) {
1867                                 if (strpos($matches[2], "\n") !== false) {
1868                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1869                                 } else {
1870                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1871                                 }
1872
1873                                 return $return;
1874                         },
1875                         $text
1876                 );
1877
1878                 // Default iframe allowed domains/path
1879                 $allowedIframeDomains = [
1880                         DI::baseUrl()->getHostname()
1881                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1882                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1883                         'www.youtube.com/embed/',
1884                         'player.vimeo.com/video/',
1885                 ];
1886
1887                 $text = HTML::purify($text, $allowedIframeDomains);
1888
1889                 return $text;
1890         }
1891
1892         /**
1893          * Strips the "abstract" tag from the provided text
1894          *
1895          * @param string $text The text with BBCode
1896          * @return string The same text - but without "abstract" element
1897          */
1898         public static function stripAbstract($text)
1899         {
1900                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1901                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1902
1903                 return $text;
1904         }
1905
1906         /**
1907          * Returns the value of the "abstract" element
1908          *
1909          * @param string $text The text that maybe contains the element
1910          * @param string $addon The addon for which the abstract is meant for
1911          * @return string The abstract
1912          */
1913         public static function getAbstract($text, $addon = '')
1914         {
1915                 $abstract = '';
1916                 $abstracts = [];
1917                 $addon = strtolower($addon);
1918
1919                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1920                         foreach ($results AS $result) {
1921                                 $abstracts[strtolower($result[1])] = $result[2];
1922                         }
1923                 }
1924
1925                 if (isset($abstracts[$addon])) {
1926                         $abstract = $abstracts[$addon];
1927                 }
1928
1929                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1930                         $abstract = $result[1];
1931                 }
1932
1933                 return $abstract;
1934         }
1935
1936         /**
1937          * Callback function to replace a Friendica style mention in a mention for Diaspora
1938          *
1939          * @param array $match Matching values for the callback
1940          *                     [1] = Mention type (! or @)
1941          *                     [2] = Name
1942          *                     [3] = Address
1943          * @return string Replaced mention
1944          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1945          * @throws \ImagickException
1946          */
1947         private static function bbCodeMention2DiasporaCallback($match)
1948         {
1949                 $contact = Contact::getByURL($match[3], false, ['addr']);
1950                 if (empty($contact['addr'])) {
1951                         return $match[0];
1952                 }
1953
1954                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1955                 return $mention;
1956         }
1957
1958         /**
1959          * Converts a BBCode text into Markdown
1960          *
1961          * This function converts a BBCode item body to be sent to Markdown-enabled
1962          * systems like Diaspora and Libertree
1963          *
1964          * @param string $text
1965          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1966          * @return string
1967          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1968          */
1969         public static function toMarkdown($text, $for_diaspora = true)
1970         {
1971                 $original_text = $text;
1972
1973                 // Since Diaspora is creating a summary for links, this function removes them before posting
1974                 if ($for_diaspora) {
1975                         $text = self::removeShareInformation($text);
1976                 }
1977
1978                 /**
1979                  * Transform #tags, strip off the [url] and replace spaces with underscore
1980                  */
1981                 $url_search_string = "^\[\]";
1982                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1983                         function ($matches) {
1984                                 return '#' . str_replace(' ', '_', $matches[2]);
1985                         },
1986                         $text
1987                 );
1988
1989                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1990                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1991
1992                 // Convert it to HTML - don't try oembed
1993                 if ($for_diaspora) {
1994                         $text = self::convert($text, false, self::DIASPORA);
1995
1996                         // Add all tags that maybe were removed
1997                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1998                                 $tagline = '';
1999                                 foreach ($tags[2] as $tag) {
2000                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2001                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2002                                                 $tagline .= '#' . $tag . ' ';
2003                                         }
2004                                 }
2005                                 $text = $text . " " . $tagline;
2006                         }
2007                 } else {
2008                         $text = self::convert($text, false, self::CONNECTORS);
2009                 }
2010
2011                 // If a link is followed by a quote then there should be a newline before it
2012                 // Maybe we should make this newline at every time before a quote.
2013                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2014
2015                 $stamp1 = microtime(true);
2016
2017                 // Now convert HTML to Markdown
2018                 $text = HTML::toMarkdown($text);
2019
2020                 DI::profiler()->saveTimestamp($stamp1, "parser");
2021
2022                 // Libertree has a problem with escaped hashtags.
2023                 $text = str_replace(['\#'], ['#'], $text);
2024
2025                 // Remove any leading or trailing whitespace, as this will mess up
2026                 // the Diaspora signature verification and cause the item to disappear
2027                 $text = trim($text);
2028
2029                 if ($for_diaspora) {
2030                         $url_search_string = "^\[\]";
2031                         $text = preg_replace_callback(
2032                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2033                                 ['self', 'bbCodeMention2DiasporaCallback'],
2034                                 $text
2035                         );
2036                 }
2037
2038                 Hook::callAll('bb2diaspora', $text);
2039
2040                 return $text;
2041         }
2042
2043         /**
2044          * Pull out all #hashtags and @person tags from $string.
2045          *
2046          * We also get @person@domain.com - which would make
2047          * the regex quite complicated as tags can also
2048          * end a sentence. So we'll run through our results
2049          * and strip the period from any tags which end with one.
2050          * Returns array of tags found, or empty array.
2051          *
2052          * @param string $string Post content
2053          *
2054          * @return array List of tag and person names
2055          */
2056         public static function getTags($string)
2057         {
2058                 $ret = [];
2059
2060                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2061                         // Convert hashtag links to hashtags
2062                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2063
2064                         // Force line feeds at bbtags
2065                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2066
2067                         // ignore anything in a bbtag
2068                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2069
2070                         // Match full names against @tags including the space between first and last
2071                         // We will look these up afterward to see if they are full names or not recognisable.
2072
2073                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2074                                 foreach ($matches[1] as $match) {
2075                                         if (strstr($match, ']')) {
2076                                                 // we might be inside a bbcode color tag - leave it alone
2077                                                 continue;
2078                                         }
2079
2080                                         if (substr($match, -1, 1) === '.') {
2081                                                 $ret[] = substr($match, 0, -1);
2082                                         } else {
2083                                                 $ret[] = $match;
2084                                         }
2085                                 }
2086                         }
2087
2088                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2089                         // and #hash tags.
2090
2091                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2092                                 foreach ($matches[1] as $match) {
2093                                         if (strstr($match, ']')) {
2094                                                 // we might be inside a bbcode color tag - leave it alone
2095                                                 continue;
2096                                         }
2097
2098                                         // try not to catch url fragments
2099                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2100                                                 continue;
2101                                         }
2102
2103                                         $ret[] = $match;
2104                                 }
2105                         }
2106                 });
2107
2108                 return array_unique($ret);
2109         }
2110
2111         /**
2112          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2113          *
2114          * @param string   $text
2115          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2116          * @param callable $callback
2117          * @return string
2118          * @throws Exception
2119          *@see Strings::performWithEscapedBlocks
2120          *
2121          */
2122         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2123         {
2124                 $tagList = array_map('preg_quote', $tagList);
2125
2126                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2127         }
2128
2129         /**
2130          * Replaces mentions in the provided message body for the provided user and network if any
2131          *
2132          * @param $body
2133          * @param $profile_uid
2134          * @param $network
2135          * @return string
2136          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2137          * @throws \ImagickException
2138          */
2139         public static function setMentions($body, $profile_uid = 0, $network = '')
2140         {
2141                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2142                         $tags = BBCode::getTags($body);
2143
2144                         $tagged = [];
2145                         $inform = '';
2146
2147                         foreach ($tags as $tag) {
2148                                 $tag_type = substr($tag, 0, 1);
2149
2150                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2151                                         continue;
2152                                 }
2153
2154                                 /*
2155                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2156                                  * Robert Johnson should be first in the $tags array
2157                                  */
2158                                 foreach ($tagged as $nextTag) {
2159                                         if (stristr($nextTag, $tag . ' ')) {
2160                                                 continue 2;
2161                                         }
2162                                 }
2163
2164                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2165
2166                                 if ($success['replaced']) {
2167                                         $tagged[] = $tag;
2168                                 }
2169                         }
2170
2171                         return $body;
2172                 });
2173
2174                 return $body;
2175         }
2176
2177         /**
2178          * @param string      $author  Author display name
2179          * @param string      $profile Author profile URL
2180          * @param string      $avatar  Author profile picture URL
2181          * @param string      $link    Post source URL
2182          * @param string      $posted  Post created date
2183          * @param string|null $guid    Post guid (if any)
2184          * @return string
2185          * @TODO Rewrite to handle over whole record array
2186          */
2187         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2188         {
2189                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2190                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2191                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2192                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2193                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2194
2195                 if ($guid) {
2196                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2197                 }
2198
2199                 $header  .= "']";
2200
2201                 return $header;
2202         }
2203
2204         /**
2205          * Returns the BBCode relevant to embed the provided URL in a post body.
2206          * For media type, it will return [img], [video] and [audio] tags.
2207          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2208          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2209          *
2210          * @param string      $url
2211          * @param bool        $tryAttachment
2212          * @param string|null $title
2213          * @param string|null $description
2214          * @param string|null $tags
2215          * @return string
2216          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2217          *@see ParseUrl::getSiteinfoCached
2218          *
2219          */
2220         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2221         {
2222                 DI::logger()->info($url);
2223
2224                 // If there is already some content information submitted we don't
2225                 // need to parse the url for content.
2226                 if (!empty($title) && !empty($description)) {
2227                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2228
2229                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2230
2231                         $str_tags = '';
2232                         if (!empty($tags)) {
2233                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2234                                 if (count($arr_tags)) {
2235                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2236                                 }
2237                         }
2238
2239                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2240
2241                         DI::logger()->info('(unparsed): returns: ' . $result);
2242
2243                         return $result;
2244                 }
2245
2246                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2247
2248                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2249                         switch ($siteinfo['type']) {
2250                                 case 'video':
2251                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2252                                         break;
2253                                 case 'audio':
2254                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2255                                         break;
2256                                 default:
2257                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2258                                         break;
2259                         }
2260
2261                         return $bbcode;
2262                 }
2263
2264                 unset($siteinfo['keywords']);
2265
2266                 // Bypass attachment if parse url for a comment
2267                 if (!$tryAttachment) {
2268                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2269                 }
2270
2271                 // Format it as BBCode attachment
2272                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2273         }
2274 }