src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2021, the Friendica project
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\PageInfo;
  31 use Friendica\Content\Smilies;
  32 use Friendica\Core\Hook;
  33 use Friendica\Core\Logger;
  34 use Friendica\Core\Protocol;
  35 use Friendica\Core\Renderer;
  36 use Friendica\DI;
  37 use Friendica\Model\Contact;
  38 use Friendica\Model\Event;
  39 use Friendica\Model\Photo;
  40 use Friendica\Model\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 = '2021-04-07';
  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                         'author_name'   => '',
 158                         'provider_name' => '',
 159                         'provider_url'  => '',
 160                         'title'         => '',
 161                         'description'   => '',
 162                 ];
 163
 164                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 165                         return self::getOldAttachmentData($body);
 166                 }
 167
 168                 $attributes = $match[2];
 169
 170                 $data['text'] = trim($match[1]);
 171
 172                 $type = '';
 173                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 174                 if (!empty($matches[1])) {
 175                         $type = strtolower($matches[1]);
 176                 }
 177
 178                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 179                 if (!empty($matches[1])) {
 180                         $type = strtolower($matches[1]);
 181                 }
 182
 183                 if ($type == '') {
 184                         return [];
 185                 }
 186
 187                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 188                         return [];
 189                 }
 190
 191                 if ($type != '') {
 192                         $data['type'] = $type;
 193                 }
 194
 195                 $url = '';
 196                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 197                 if (!empty($matches[1])) {
 198                         $url = $matches[1];
 199                 }
 200
 201                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 202                 if (!empty($matches[1])) {
 203                         $url = $matches[1];
 204                 }
 205
 206                 if ($url != '') {
 207                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 208                 }
 209
 210                 $title = '';
 211                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 212                 if (!empty($matches[1])) {
 213                         $title = $matches[1];
 214                 }
 215
 216                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 217                 if (!empty($matches[1])) {
 218                         $title = $matches[1];
 219                 }
 220
 221                 if ($title != '') {
 222                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 223                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 224                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 225                         $data['title'] = $title;
 226                 }
 227
 228                 $image = '';
 229                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 230                 if (!empty($matches[1])) {
 231                         $image = $matches[1];
 232                 }
 233
 234                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 235                 if (!empty($matches[1])) {
 236                         $image = $matches[1];
 237                 }
 238
 239                 if ($image != '') {
 240                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 241                 }
 242
 243                 $preview = '';
 244                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 245                 if (!empty($matches[1])) {
 246                         $preview = $matches[1];
 247                 }
 248
 249                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 250                 if (!empty($matches[1])) {
 251                         $preview = $matches[1];
 252                 }
 253
 254                 if ($preview != '') {
 255                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 256                 }
 257
 258                 $provider_name = '';
 259                 preg_match("/publisher_name='(.*?)'/ism", $attributes, $matches);
 260                 if (!empty($matches[1])) {
 261                         $provider_name = $matches[1];
 262                 }
 263
 264                 preg_match('/publisher_name="(.*?)"/ism', $attributes, $matches);
 265                 if (!empty($matches[1])) {
 266                         $provider_name = $matches[1];
 267                 }
 268
 269                 if ($provider_name != '') {
 270                         $data['provider_name'] = html_entity_decode($provider_name, ENT_QUOTES, 'UTF-8');
 271                 }
 272
 273                 $provider_url = '';
 274                 preg_match("/publisher_url='(.*?)'/ism", $attributes, $matches);
 275                 if (!empty($matches[1])) {
 276                         $provider_url = $matches[1];
 277                 }
 278
 279                 preg_match('/publisher_url="(.*?)"/ism', $attributes, $matches);
 280                 if (!empty($matches[1])) {
 281                         $provider_url = $matches[1];
 282                 }
 283
 284                 if ($provider_url != '') {
 285                         $data['provider_url'] = html_entity_decode($provider_url, ENT_QUOTES, 'UTF-8');
 286                 }
 287
 288                 $author_name = '';
 289                 preg_match("/author_name='(.*?)'/ism", $attributes, $matches);
 290                 if (!empty($matches[1])) {
 291                         $author_name = $matches[1];
 292                 }
 293
 294                 preg_match('/author_name="(.*?)"/ism', $attributes, $matches);
 295                 if (!empty($matches[1])) {
 296                         $author_name = $matches[1];
 297                 }
 298
 299                 if (($author_name != '') && ($author_name != $provider_name)) {
 300                         $data['author_name'] = html_entity_decode($author_name, ENT_QUOTES, 'UTF-8');
 301                 }
 302
 303                 $data['description'] = trim($match[3]);
 304
 305                 $data['after'] = trim($match[4]);
 306
 307                 $parts = parse_url($data['url']);
 308                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 309                         if (empty($data['provider_name'])) {
 310                                 $data['provider_name'] = $parts['host'];
 311                         }
 312                         if (empty($data['provider_url'])) {
 313                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 314
 315                                 if (!empty($parts['port'])) {
 316                                         $data['provider_url'] .= ':' . $parts['port'];
 317                                 }
 318                         }
 319                 }
 320
 321                 return $data;
 322         }
 323
 324         public static function getAttachedData($body, $item = [])
 325         {
 326                 /*
 327                 - text:
 328                 - type: link, video, photo
 329                 - title:
 330                 - url:
 331                 - image:
 332                 - description:
 333                 - (thumbnail)
 334                 */
 335
 336                 $has_title = !empty($item['title']);
 337                 $plink = $item['plink'] ?? '';
 338                 $post = self::getAttachmentData($body);
 339
 340                 // Get all linked images with alternative image description
 341                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 342                         foreach ($pictures as $picture) {
 343                                 if (Photo::isLocal($picture[1])) {
 344                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 345                                 }
 346                         }
 347                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 348                                 $post['image_description'] = $post['images'][0]['description'];
 349                         }
 350                 }
 351
 352                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 353                         foreach ($pictures as $picture) {
 354                                 if (Photo::isLocal($picture[1])) {
 355                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 356                                 }
 357                         }
 358                 }
 359
 360                 // if nothing is found, it maybe having an image.
 361                 if (!isset($post['type'])) {
 362                         // Simplify image codes
 363                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 364                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 365                         $post['text'] = $body;
 366
 367                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 368                                 if ((count($pictures) == 1) && !$has_title) {
 369                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 370                                                 // Replace the preview picture with the real picture
 371                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 372                                                 $data = ['url' => $url, 'type' => 'photo'];
 373                                         } else {
 374                                                 // Checking, if the link goes to a picture
 375                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 376                                         }
 377
 378                                         // Workaround:
 379                                         // Sometimes photo posts to the own album are not detected at the start.
 380                                         // So we seem to cannot use the cache for these cases. That's strange.
 381                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 382                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 383                                         }
 384
 385                                         if ($data['type'] == 'photo') {
 386                                                 $post['type'] = 'photo';
 387                                                 if (isset($data['images'][0])) {
 388                                                         $post['image'] = $data['images'][0]['src'];
 389                                                         $post['url'] = $data['url'];
 390                                                 } else {
 391                                                         $post['image'] = $data['url'];
 392                                                 }
 393
 394                                                 $post['preview'] = $pictures[0][2];
 395                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 396                                         } else {
 397                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 398                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 399                                                         $post['type'] = 'photo';
 400                                                         $post['image'] = $pictures[0][1];
 401                                                         $post['preview'] = $pictures[0][2];
 402                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 403                                                 }
 404                                         }
 405                                 } elseif (count($pictures) > 0) {
 406                                         $post['type'] = 'link';
 407                                         $post['url'] = $plink;
 408                                         $post['image'] = $pictures[0][2];
 409                                         $post['text'] = $body;
 410
 411                                         foreach ($pictures as $picture) {
 412                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 413                                         }
 414                                 }
 415                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 416                                 if ((count($pictures) == 1) && !$has_title) {
 417                                         $post['type'] = 'photo';
 418                                         $post['image'] = $pictures[0][1];
 419                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 420                                 } elseif (count($pictures) > 0) {
 421                                         $post['type'] = 'link';
 422                                         $post['url'] = $plink;
 423                                         $post['image'] = $pictures[0][1];
 424                                         $post['text'] = $body;
 425
 426                                         foreach ($pictures as $picture) {
 427                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 428                                         }
 429                                 }
 430                         }
 431
 432                         // Test for the external links
 433                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 434                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 435
 436                         $links = array_merge($links1, $links2);
 437
 438                         // If there is only a single one, then use it.
 439                         // This should cover link posts via API.
 440                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 441                                 $post['type'] = 'link';
 442                                 $post['url'] = $links[0][1];
 443                         }
 444
 445                         // Simplify "video" element
 446                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 447
 448                         // Now count the number of external media links
 449                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 450                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 451                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 452                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 453
 454                         // Add them to the other external links
 455                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 456
 457                         // Are there more than one?
 458                         if (count($links) > 1) {
 459                                 // The post will be the type "text", which means a blog post
 460                                 unset($post['type']);
 461                                 $post['url'] = $plink;
 462                         }
 463
 464                         if (!isset($post['type'])) {
 465                                 $post['type'] = "text";
 466                                 $post['text'] = trim($body);
 467                         }
 468                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 469                         $data = ParseUrl::getSiteinfoCached($post['url']);
 470
 471                         if (isset($data['images'][0])) {
 472                                 $post['image'] = $data['images'][0]['src'];
 473                         }
 474                 }
 475
 476                 return $post;
 477         }
 478
 479         /**
 480          * Remove [attachment] BBCode and replaces it with a regular [url]
 481          *
 482          * @param string  $body
 483          * @param boolean $no_link_desc No link description
 484          *
 485          * @return string with replaced body
 486          */
 487         public static function removeAttachment($body, $no_link_desc = false)
 488         {
 489                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 490                         function ($match) use ($no_link_desc) {
 491                                 $attach_data = self::getAttachmentData($match[0]);
 492                                 if (empty($attach_data['url'])) {
 493                                         return $match[0];
 494                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 495                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 496                                 } else {
 497                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 498                                 }
 499                 }, $body);
 500         }
 501
 502         /**
 503          * Converts a BBCode text into plaintext
 504          *
 505          * @param      $text
 506          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 507          *
 508          * @return string
 509          */
 510         public static function toPlaintext($text, $keep_urls = true)
 511         {
 512                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 513
 514                 return $naked_text;
 515         }
 516
 517         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 518         {
 519                 // Only send proxied pictures to API and for internal display
 520                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 521                         return ProxyUtils::proxifyUrl($image);
 522                 } else {
 523                         return $image;
 524                 }
 525         }
 526
 527         /**
 528          * This function changing the visual size (not the real size) of images.
 529          * The function does not work for pictures with an alternate text description.
 530          * This could only be changed by using some new "img" BBCode format.
 531          *
 532          * @param string $srctext The body with images
 533          * @return string The body with possibly scaled images
 534          */
 535         public static function scaleExternalImages(string $srctext)
 536         {
 537                 $s = $srctext;
 538
 539                 // Simplify image links
 540                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 541
 542                 $matches = null;
 543                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 544                 if ($c) {
 545                         foreach ($matches as $mtch) {
 546                                 Logger::log('scale_external_image: ' . $mtch[1]);
 547
 548                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 549                                 if (stristr($mtch[1], $hostname)) {
 550                                         continue;
 551                                 }
 552
 553                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 554                                 if (!$curlResult->isSuccess()) {
 555                                         continue;
 556                                 }
 557
 558                                 $i = $curlResult->getBody();
 559                                 $type = $curlResult->getContentType();
 560                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 561
 562                                 if ($i) {
 563                                         $Image = new Image($i, $type);
 564                                         if ($Image->isValid()) {
 565                                                 $orig_width = $Image->getWidth();
 566                                                 $orig_height = $Image->getHeight();
 567
 568                                                 if ($orig_width > 640 || $orig_height > 640) {
 569                                                         $Image->scaleDown(640);
 570                                                         $new_width = $Image->getWidth();
 571                                                         $new_height = $Image->getHeight();
 572                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 573                                                         $s = str_replace(
 574                                                                 $mtch[0],
 575                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 576                                                                 . "\n",
 577                                                                 $s
 578                                                         );
 579                                                         Logger::info('New string', ['image' => $s]);
 580                                                 }
 581                                         }
 582                                 }
 583                         }
 584                 }
 585
 586                 return $s;
 587         }
 588
 589         /**
 590          * Truncates imported message body string length to max_import_size
 591          *
 592          * The purpose of this function is to apply system message length limits to
 593          * imported messages without including any embedded photos in the length
 594          *
 595          * @param string $body
 596          * @return string
 597          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 598          */
 599         public static function limitBodySize($body)
 600         {
 601                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 602
 603                 // If the length of the body, including the embedded images, is smaller
 604                 // than the maximum, then don't waste time looking for the images
 605                 if ($maxlen && (strlen($body) > $maxlen)) {
 606
 607                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 608
 609                         $orig_body = $body;
 610                         $new_body = '';
 611                         $textlen = 0;
 612
 613                         $img_start = strpos($orig_body, '[img');
 614                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 615                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 616                         while (($img_st_close !== false) && ($img_end !== false)) {
 617
 618                                 $img_st_close++; // make it point to AFTER the closing bracket
 619                                 $img_end += $img_start;
 620                                 $img_end += strlen('[/img]');
 621
 622                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 623                                         // This is an embedded image
 624
 625                                         if (($textlen + $img_start) > $maxlen) {
 626                                                 if ($textlen < $maxlen) {
 627                                                         Logger::info('the limit happens before an embedded image');
 628                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 629                                                         $textlen = $maxlen;
 630                                                 }
 631                                         } else {
 632                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 633                                                 $textlen += $img_start;
 634                                         }
 635
 636                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 637                                 } else {
 638
 639                                         if (($textlen + $img_end) > $maxlen) {
 640                                                 if ($textlen < $maxlen) {
 641                                                         Logger::info('the limit happens before the end of a non-embedded image');
 642                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 643                                                         $textlen = $maxlen;
 644                                                 }
 645                                         } else {
 646                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 647                                                 $textlen += $img_end;
 648                                         }
 649                                 }
 650                                 $orig_body = substr($orig_body, $img_end);
 651
 652                                 if ($orig_body === false) {
 653                                         // in case the body ends on a closing image tag
 654                                         $orig_body = '';
 655                                 }
 656
 657                                 $img_start = strpos($orig_body, '[img');
 658                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 659                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 660                         }
 661
 662                         if (($textlen + strlen($orig_body)) > $maxlen) {
 663                                 if ($textlen < $maxlen) {
 664                                         Logger::info('the limit happens after the end of the last image');
 665                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 666                                 }
 667                         } else {
 668                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 669                                 $new_body = $new_body . $orig_body;
 670                         }
 671
 672                         return $new_body;
 673                 } else {
 674                         return $body;
 675                 }
 676         }
 677
 678         /**
 679          * Processes [attachment] tags
 680          *
 681          * Note: Can produce a [bookmark] tag in the returned string
 682          *
 683          * @param string  $text
 684          * @param integer $simplehtml
 685          * @param bool    $tryoembed
 686          * @return string
 687          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 688          */
 689         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 690         {
 691                 $data = self::getAttachmentData($text);
 692                 if (empty($data) || empty($data['url'])) {
 693                         return $text;
 694                 }
 695
 696                 if (isset($data['title'])) {
 697                         $data['title'] = strip_tags($data['title']);
 698                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 699                 } else {
 700                         $data['title'] = null;
 701                 }
 702
 703                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 704                         $data['preview'] = $data['image'];
 705                         $data['image'] = '';
 706                 }
 707
 708                 $return = '';
 709                 try {
 710                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 711                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 712                         } else {
 713                                 throw new Exception('OEmbed is disabled for this attachment.');
 714                         }
 715                 } catch (Exception $e) {
 716                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 717
 718                         if ($simplehtml != self::CONNECTORS) {
 719                                 $return = sprintf('<div class="type-%s">', $data['type']);
 720                         }
 721
 722                         if (!empty($data['title']) && !empty($data['url'])) {
 723                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 724                                         $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), $data['title']);
 725                                 } else {
 726                                         if (!empty($data['image'])) {
 727                                                 $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), $data['title']);
 728                                         } elseif (!empty($data['preview'])) {
 729                                                 $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']);
 730                                         }
 731                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 732                                 }
 733                         }
 734
 735                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 736                                 // Sanitize the HTML by converting it to BBCode
 737                                 $bbcode = HTML::toBBCode($data['description']);
 738                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 739                         }
 740
 741                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 742                                 if (!empty($data['author_name'])) {
 743                                         $return .= sprintf('<sup><a href="%s">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 744                                 } else {
 745                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 746                                 }
 747                         }
 748
 749                         if ($simplehtml != self::CONNECTORS) {
 750                                 $return .= '</div>';
 751                         }
 752                 }
 753
 754                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 755         }
 756
 757         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 758         {
 759                 $data = self::getAttachmentData($Text);
 760
 761                 if (!$data) {
 762                         return $Text;
 763                 } elseif ($nolink) {
 764                         return $data['text'] . ($data['after'] ?? '');
 765                 }
 766
 767                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 768                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 769                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 770                         $data['title'] = $data['url'];
 771                 } elseif (($text != '') && strstr($title, $text)) {
 772                         $data['text'] = $data['title'];
 773                         $data['title'] = $data['url'];
 774                 }
 775
 776                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 777                         return $data['title'] . $data['after'];
 778                 }
 779
 780                 // If the link already is included in the post, don't add it again
 781                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 782                         return $data['text'] . $data['after'];
 783                 }
 784
 785                 $text = $data['text'];
 786
 787                 if (!empty($data['url']) && !empty($data['title'])) {
 788                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 789                 } elseif (!empty($data['url'])) {
 790                         $text .= "\n[url]" . $data['url'] . '[/url]';
 791                 }
 792
 793                 return $text . "\n" . $data['after'];
 794         }
 795
 796         /**
 797          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 798          *
 799          * @param array $match Array with the matching values
 800          * @return string reformatted link including HTML codes
 801          */
 802         private static function convertUrlForActivityPubCallback($match)
 803         {
 804                 $url = $match[1];
 805
 806                 if (isset($match[2]) && ($match[1] != $match[2])) {
 807                         return $match[0];
 808                 }
 809
 810                 $parts = parse_url($url);
 811                 if (!isset($parts['scheme'])) {
 812                         return $match[0];
 813                 }
 814
 815                 return self::convertUrlForActivityPub($url);
 816         }
 817
 818         /**
 819          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 820          *
 821          * @param string $url URL that is about to be reformatted
 822          * @return string reformatted link including HTML codes
 823          */
 824         private static function convertUrlForActivityPub($url)
 825         {
 826                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 827                 return sprintf($html, $url, self::getStyledURL($url));
 828         }
 829
 830         /**
 831          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 832          *
 833          * @param string $url URL that is about to be reformatted
 834          * @return string reformatted link
 835          */
 836         private static function getStyledURL($url)
 837         {
 838                 $parts = parse_url($url);
 839                 $scheme = $parts['scheme'] . '://';
 840                 $styled_url = str_replace($scheme, '', $url);
 841
 842                 if (strlen($styled_url) > 30) {
 843                         $styled_url = substr($styled_url, 0, 30) . "…";
 844                 }
 845
 846                 return $styled_url;
 847         }
 848
 849         /*
 850          * [noparse][i]italic[/i][/noparse] turns into
 851          * [noparse][ i ]italic[ /i ][/noparse],
 852          * to hide them from parser.
 853          */
 854         private static function escapeNoparseCallback($match)
 855         {
 856                 $whole_match = $match[0];
 857                 $captured = $match[1];
 858                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 859                 $new_str = str_replace($captured, $spacefied, $whole_match);
 860                 return $new_str;
 861         }
 862
 863         /*
 864          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 865          * now turns back and the [noparse] tags are trimed
 866          * returning [i]italic[/i]
 867          */
 868         private static function unescapeNoparseCallback($match)
 869         {
 870                 $captured = $match[1];
 871                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 872                 return $unspacefied;
 873         }
 874
 875         /**
 876          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 877          * occurrences
 878          *
 879          * @param string $text        Text to search
 880          * @param string $name        Tag name
 881          * @param int    $occurrences Number of first occurrences to skip
 882          * @return boolean|array
 883          */
 884         public static function getTagPosition($text, $name, $occurrences = 0)
 885         {
 886                 if ($occurrences < 0) {
 887                         $occurrences = 0;
 888                 }
 889
 890                 $start_open = -1;
 891                 for ($i = 0; $i <= $occurrences; $i++) {
 892                         if ($start_open !== false) {
 893                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 894                         }
 895                 }
 896
 897                 if ($start_open === false) {
 898                         return false;
 899                 }
 900
 901                 $start_equal = strpos($text, '=', $start_open);
 902                 $start_close = strpos($text, ']', $start_open);
 903
 904                 if ($start_close === false) {
 905                         return false;
 906                 }
 907
 908                 $start_close++;
 909
 910                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 911
 912                 if ($end_open === false) {
 913                         return false;
 914                 }
 915
 916                 $res = [
 917                         'start' => [
 918                                 'open' => $start_open,
 919                                 'close' => $start_close
 920                         ],
 921                         'end' => [
 922                                 'open' => $end_open,
 923                                 'close' => $end_open + strlen('[/' . $name . ']')
 924                         ],
 925                 ];
 926
 927                 if ($start_equal !== false) {
 928                         $res['start']['equal'] = $start_equal + 1;
 929                 }
 930
 931                 return $res;
 932         }
 933
 934         /**
 935          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 936          *
 937          * @param string $pattern Preg pattern string
 938          * @param string $replace Preg replace string
 939          * @param string $name    BBCode tag name
 940          * @param string $text    Text to search
 941          * @return string
 942          */
 943         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 944         {
 945                 $occurrences = 0;
 946                 $pos = self::getTagPosition($text, $name, $occurrences);
 947                 while ($pos !== false && $occurrences++ < 1000) {
 948                         $start = substr($text, 0, $pos['start']['open']);
 949                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 950                         $end = substr($text, $pos['end']['close']);
 951                         if ($end === false) {
 952                                 $end = '';
 953                         }
 954
 955                         $subject = preg_replace($pattern, $replace, $subject);
 956                         $text = $start . $subject . $end;
 957
 958                         $pos = self::getTagPosition($text, $name, $occurrences);
 959                 }
 960
 961                 return $text;
 962         }
 963
 964         private static function extractImagesFromItemBody($body)
 965         {
 966                 $saved_image = [];
 967                 $orig_body = $body;
 968                 $new_body = '';
 969
 970                 $cnt = 0;
 971                 $img_start = strpos($orig_body, '[img');
 972                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 973                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 974                 while (($img_st_close !== false) && ($img_end !== false)) {
 975                         $img_st_close++; // make it point to AFTER the closing bracket
 976                         $img_end += $img_start;
 977
 978                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 979                                 // This is an embedded image
 980                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 981                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 982
 983                                 $cnt++;
 984                         } else {
 985                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 986                         }
 987
 988                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 989
 990                         if ($orig_body === false) {
 991                                 // in case the body ends on a closing image tag
 992                                 $orig_body = '';
 993                         }
 994
 995                         $img_start = strpos($orig_body, '[img');
 996                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 997                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 998                 }
 999
1000                 $new_body = $new_body . $orig_body;
1001
1002                 return ['body' => $new_body, 'images' => $saved_image];
1003         }
1004
1005         private static function interpolateSavedImagesIntoItemBody($body, array $images)
1006         {
1007                 $newbody = $body;
1008
1009                 $cnt = 0;
1010                 foreach ($images as $image) {
1011                         // We're depending on the property of 'foreach' (specified on the PHP website) that
1012                         // it loops over the array starting from the first element and going sequentially
1013                         // to the last element
1014                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1015                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
1016                         $cnt++;
1017                 }
1018
1019                 return $newbody;
1020         }
1021
1022         /**
1023          * This function converts a [share] block to text according to a provided callback function whose signature is:
1024          *
1025          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
1026          *
1027          * Where:
1028          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
1029          * data lookup
1030          * - $author_contact is a contact record array
1031          * - $content is the inner content of the [share] block
1032          * - $is_quote_share indicates whether there's any content before the [share] block
1033          * - Return value is the string that should replace the [share] block in the provided text
1034          *
1035          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
1036          *
1037          * @param  string   $text     A BBCode string
1038          * @param  callable $callback
1039          * @return string The BBCode string with all [share] blocks replaced
1040          */
1041         public static function convertShare($text, callable $callback)
1042         {
1043                 $return = preg_replace_callback(
1044                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
1045                         function ($match) use ($callback) {
1046                                 $attribute_string = $match[2];
1047                                 $attributes = [];
1048                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
1049                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1050                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1051                                 }
1052
1053                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
1054                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1055                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1056
1057                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1058                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1059                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1060
1061                                 if ($attributes['avatar']) {
1062                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1063                                 }
1064
1065                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1066
1067                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1068                         },
1069                         $text
1070                 );
1071
1072                 return $return;
1073         }
1074
1075         /**
1076          * Default [share] tag conversion callback
1077          *
1078          * Note: Can produce a [bookmark] tag in the output
1079          *
1080          * @see BBCode::convertShare()
1081          * @param array   $attributes     [share] block attribute values
1082          * @param array   $author_contact Contact row of the shared author
1083          * @param string  $content        Inner content of the [share] block
1084          * @param boolean $is_quote_share Whether there is content before the [share] block
1085          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1086          * @return string
1087          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1088          */
1089         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1090         {
1091                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1092
1093                 switch ($simplehtml) {
1094                         case self::API:
1095                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1096                                 break;
1097                         case self::DIASPORA:
1098                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1099                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1100                                 } else {
1101                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1102
1103                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1104                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1105                                         }
1106
1107                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1108
1109                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1110                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1111                                         }
1112                                 }
1113
1114                                 break;
1115                         case self::CONNECTORS:
1116                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1117                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1118                                 $headline .= ':</b></p>' . "\n";
1119
1120                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1121
1122                                 break;
1123                         case self::OSTATUS:
1124                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1125                                 break;
1126                         case self::ACTIVITYPUB:
1127                                 $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>';
1128                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1129                                 break;
1130                         default:
1131                                 $text = ($is_quote_share? "\n" : '');
1132
1133                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1134                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1135
1136                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1137                                 $text .= Renderer::replaceMacros($tpl, [
1138                                         '$profile'      => $attributes['profile'],
1139                                         '$avatar'       => $attributes['avatar'],
1140                                         '$author'       => $attributes['author'],
1141                                         '$link'         => $attributes['link'],
1142                                         '$link_title'   => DI::l10n()->t('link to source'),
1143                                         '$posted'       => $attributes['posted'],
1144                                         '$guid'         => $attributes['guid'],
1145                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1146                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1147                                         '$content'      => self::setMentions(trim($content), 0, $network),
1148                                 ]);
1149                                 break;
1150                 }
1151
1152                 return $text;
1153         }
1154
1155         private static function removePictureLinksCallback($match)
1156         {
1157                 $cache_key = 'remove:' . $match[1];
1158                 $text = DI::cache()->get($cache_key);
1159
1160                 if (is_null($text)) {
1161                         $a = DI::app();
1162
1163                         $stamp1 = microtime(true);
1164
1165                         $ch = @curl_init($match[1]);
1166                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1167                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1168                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1169                         @curl_exec($ch);
1170                         $curl_info = @curl_getinfo($ch);
1171
1172                         DI::profiler()->saveTimestamp($stamp1, "network");
1173
1174                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1175                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1176                         } else {
1177                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1178
1179                                 // if its not a picture then look if its a page that contains a picture link
1180                                 $body = DI::httpRequest()->fetch($match[1]);
1181
1182                                 $doc = new DOMDocument();
1183                                 @$doc->loadHTML($body);
1184                                 $xpath = new DOMXPath($doc);
1185                                 $list = $xpath->query("//meta[@name]");
1186                                 foreach ($list as $node) {
1187                                         $attr = [];
1188
1189                                         if ($node->attributes->length) {
1190                                                 foreach ($node->attributes as $attribute) {
1191                                                         $attr[$attribute->name] = $attribute->value;
1192                                                 }
1193                                         }
1194
1195                                         if (strtolower($attr['name']) == 'twitter:image') {
1196                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1197                                         }
1198                                 }
1199                         }
1200                         DI::cache()->set($cache_key, $text);
1201                 }
1202
1203                 return $text;
1204         }
1205
1206         private static function expandLinksCallback($match)
1207         {
1208                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1209                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1210                 } else {
1211                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1212                 }
1213         }
1214
1215         private static function cleanPictureLinksCallback($match)
1216         {
1217                 $a = DI::app();
1218
1219                 // When the picture link is the own photo path then we can avoid fetching the link
1220                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1221                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1222                         if (!empty($match[3])) {
1223                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1224                         } else {
1225                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1226                         }
1227                         return $text;
1228                 }
1229
1230                 $cache_key = 'clean:' . $match[1];
1231                 $text = DI::cache()->get($cache_key);
1232                 if (!is_null($text)) {
1233                         return $text;
1234                 }
1235
1236                 // Only fetch the header, not the content
1237                 $stamp1 = microtime(true);
1238
1239                 $ch = @curl_init($match[1]);
1240                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1241                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1242                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1243                 @curl_exec($ch);
1244                 $curl_info = @curl_getinfo($ch);
1245
1246                 DI::profiler()->saveTimestamp($stamp1, "network");
1247
1248                 // if its a link to a picture then embed this picture
1249                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1250                         $text = '[img]' . $match[1] . '[/img]';
1251                 } else {
1252                         if (!empty($match[3])) {
1253                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1254                         } else {
1255                                 $text = '[img]' . $match[2] . '[/img]';
1256                         }
1257
1258                         // if its not a picture then look if its a page that contains a picture link
1259                         $body = DI::httpRequest()->fetch($match[1]);
1260
1261                         $doc = new DOMDocument();
1262                         @$doc->loadHTML($body);
1263                         $xpath = new DOMXPath($doc);
1264                         $list = $xpath->query("//meta[@name]");
1265                         foreach ($list as $node) {
1266                                 $attr = [];
1267                                 if ($node->attributes->length) {
1268                                         foreach ($node->attributes as $attribute) {
1269                                                 $attr[$attribute->name] = $attribute->value;
1270                                         }
1271                                 }
1272
1273                                 if (strtolower($attr['name']) == "twitter:image") {
1274                                         if (!empty($match[3])) {
1275                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1276                                         } else {
1277                                                 $text = "[img]" . $attr['content'] . "[/img]";
1278                                         }
1279                                 }
1280                         }
1281                 }
1282                 DI::cache()->set($cache_key, $text);
1283
1284                 return $text;
1285         }
1286
1287         public static function cleanPictureLinks($text)
1288         {
1289                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1290                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1291                 return $return;
1292         }
1293
1294         public static function removeLinks(string $bbcode)
1295         {
1296                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1297                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1298
1299                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1300                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1301                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1302                 return $bbcode;
1303         }
1304
1305         /**
1306          * Converts a BBCode message to HTML message
1307          *
1308          * BBcode 2 HTML was written by WAY2WEB.net
1309          * extended to work with Mistpark/Friendica - Mike Macgirvin
1310          *
1311          * Simple HTML values meaning:
1312          * - 0: Friendica display
1313          * - 1: Unused
1314          * - 2: Used for Windows Phone push, Friendica API
1315          * - 3: Used before converting to Markdown in bb2diaspora.php
1316          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1317          * - 5: Unused
1318          * - 6: Unused
1319          * - 7: Used for dfrn, OStatus
1320          * - 8: Used for Twitter, WP backlink text setting
1321          * - 9: ActivityPub
1322          *
1323          * @param string $text
1324          * @param bool   $try_oembed
1325          * @param int    $simple_html
1326          * @param bool   $for_plaintext
1327          * @return string
1328          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1329          */
1330         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1331         {
1332                 // Accounting for null default column values
1333                 if (is_null($text) || $text === '') {
1334                         return '';
1335                 }
1336
1337                 Hook::callAll('bbcode', $text);
1338
1339                 $a = DI::app();
1340
1341                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1342                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1343                                 /*
1344                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1345                                  *
1346                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1347                                  * $match[1] = $url
1348                                  * $match[2] = $title or absent
1349                                  */
1350                                 $try_oembed_callback = function ($match)
1351                                 {
1352                                         $url = $match[1];
1353                                         $title = $match[2] ?? null;
1354
1355                                         try {
1356                                                 $return = OEmbed::getHTML($url, $title);
1357                                         } catch (Exception $ex) {
1358                                                 $return = $match[0];
1359                                         }
1360
1361                                         return $return;
1362                                 };
1363
1364                                 // Remove the abstract element. It is a non visible element.
1365                                 $text = self::stripAbstract($text);
1366
1367                                 // Line ending normalisation
1368                                 $text = str_replace("\r\n", "\n", $text);
1369
1370                                 // Move new lines outside of tags
1371                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1372                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1373
1374                                 // Extract the private images which use data urls since preg has issues with
1375                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1376                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1377
1378                                 $extracted = self::extractImagesFromItemBody($text);
1379                                 $text = $extracted['body'];
1380                                 $saved_image = $extracted['images'];
1381
1382                                 // If we find any event code, turn it into an event.
1383                                 // After we're finished processing the bbcode we'll
1384                                 // replace all of the event code with a reformatted version.
1385
1386                                 $ev = Event::fromBBCode($text);
1387
1388                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1389                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1390
1391                                 $text = str_replace("<", "&lt;", $text);
1392                                 $text = str_replace(">", "&gt;", $text);
1393
1394                                 // remove some newlines before the general conversion
1395                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1396                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1397
1398                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1399                                 if (!$try_oembed) {
1400                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1401                                 }
1402
1403                                 // Remove linefeeds inside of the table elements. See issue #6799
1404                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1405                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1406                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1407                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1408                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1409                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1410                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1411                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1412                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1413                                         "[hr]", "[hr]", "[hr]", "[hr]",
1414                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1415                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1416                                 do {
1417                                         $oldtext = $text;
1418                                         $text = str_replace($search, $replace, $text);
1419                                 } while ($oldtext != $text);
1420
1421                                 // Replace these here only once
1422                                 $search = ["\n[table]", "[/table]\n"];
1423                                 $replace = ["[table]", "[/table]"];
1424                                 $text = str_replace($search, $replace, $text);
1425
1426                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1427                                 $text = trim($text, "\n");
1428
1429                                 // removing multiplicated newlines
1430                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1431                                         $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",
1432                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1433                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1434                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1435                                         do {
1436                                                 $oldtext = $text;
1437                                                 $text = str_replace($search, $replace, $text);
1438                                         } while ($oldtext != $text);
1439                                 }
1440
1441                                 // Add HTML new lines
1442                                 $text = str_replace("\n", '<br>', $text);
1443
1444                                 /// @todo Have a closer look at the different html modes
1445                                 // Handle attached links or videos
1446                                 if ($simple_html == self::ACTIVITYPUB) {
1447                                         $text = self::removeAttachment($text);
1448                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1449                                         $text = self::removeAttachment($text, true);
1450                                 } else {
1451                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1452                                 }
1453
1454                                 $nosmile = strpos($text, '[nosmile]') !== false;
1455                                 $text = str_replace('[nosmile]', '', $text);
1456
1457                                 // Replace non graphical smilies for external posts
1458                                 if (!$nosmile && !$for_plaintext) {
1459                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) {
1460                                                 return Smilies::replace($text);
1461                                         });
1462                                 }
1463
1464                                 // leave open the posibility of [map=something]
1465                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1466                                 if (strpos($text, '[/map]') !== false) {
1467                                         $text = preg_replace_callback(
1468                                                 "/\[map\](.*?)\[\/map\]/ism",
1469                                                 function ($match) use ($simple_html) {
1470                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1471                                                 },
1472                                                 $text
1473                                         );
1474                                 }
1475
1476                                 if (strpos($text, '[map=') !== false) {
1477                                         $text = preg_replace_callback(
1478                                                 "/\[map=(.*?)\]/ism",
1479                                                 function ($match) use ($simple_html) {
1480                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1481                                                 },
1482                                                 $text
1483                                         );
1484                                 }
1485
1486                                 if (strpos($text, '[map]') !== false) {
1487                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1488                                 }
1489
1490                                 // Check for headers
1491                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1492                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1493                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1494                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1495                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1496                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1497
1498                                 // Check for paragraph
1499                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1500
1501                                 // Check for bold text
1502                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1503
1504                                 // Check for Italics text
1505                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1506
1507                                 // Check for Underline text
1508                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1509
1510                                 // Check for strike-through text
1511                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1512
1513                                 // Check for over-line text
1514                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1515
1516                                 // Check for colored text
1517                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1518
1519                                 // Check for sized text
1520                                 // [size=50] --> font-size: 50px (with the unit).
1521                                 if ($simple_html != self::DIASPORA) {
1522                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1523                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1524                                 } else {
1525                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1526                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1527                                 }
1528
1529
1530                                 // Check for centered text
1531                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1532
1533                                 // Check for list text
1534                                 $text = str_replace("[*]", "<li>", $text);
1535
1536                                 // Check for style sheet commands
1537                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1538
1539                                 // Check for CSS classes
1540                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1541
1542                                 // handle nested lists
1543                                 $endlessloop = 0;
1544
1545                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1546                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1547                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1548                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1549                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1550                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1551                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1552                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1553                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1554                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1555                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1556                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1557                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1558                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1559                                 }
1560
1561                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1562                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1563                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1564                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1565
1566                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1567                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1568
1569                                 $text = str_replace('[hr]', '<hr />', $text);
1570
1571                                 if (!$for_plaintext) {
1572                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1573                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1574                                         });
1575                                 }
1576
1577                                 // Check for font change text
1578                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1579
1580                                 // Declare the format for [spoiler] layout
1581                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1582
1583                                 // Check for [spoiler] text
1584                                 // handle nested quotes
1585                                 $endlessloop = 0;
1586                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1587                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1588                                 }
1589
1590                                 // Check for [spoiler=Title] text
1591
1592                                 // handle nested quotes
1593                                 $endlessloop = 0;
1594                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1595                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1596                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1597                                                 $text);
1598                                 }
1599
1600                                 // Declare the format for [quote] layout
1601                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1602
1603                                 // Check for [quote] text
1604                                 // handle nested quotes
1605                                 $endlessloop = 0;
1606                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1607                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1608                                 }
1609
1610                                 // Check for [quote=Author] text
1611
1612                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1613
1614                                 // handle nested quotes
1615                                 $endlessloop = 0;
1616                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1617                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1618                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1619                                                 $text);
1620                                 }
1621
1622
1623                                 // [img=widthxheight]image source[/img]
1624                                 $text = preg_replace_callback(
1625                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1626                                         function ($matches) use ($simple_html) {
1627                                                 if (strpos($matches[3], "data:image/") === 0) {
1628                                                         return $matches[0];
1629                                                 }
1630
1631                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1632                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1633                                         },
1634                                         $text
1635                                 );
1636
1637                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1638                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1639
1640                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1641                                         function ($matches) use ($simple_html) {
1642                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1643                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1644                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1645                                         },
1646                                         $text);
1647
1648                                 // Images
1649                                 // [img]pathtoimage[/img]
1650                                 $text = preg_replace_callback(
1651                                         "/\[img\](.*?)\[\/img\]/ism",
1652                                         function ($matches) use ($simple_html) {
1653                                                 if (strpos($matches[1], "data:image/") === 0) {
1654                                                         return $matches[0];
1655                                                 }
1656
1657                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1658                                                 return "[img]" . $matches[1] . "[/img]";
1659                                         },
1660                                         $text
1661                                 );
1662
1663                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1664                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1665
1666                                 $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);
1667                                 $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>', $text);
1668                                 //$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);
1669
1670                                 // Simplify "video" element
1671                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1672
1673                                 if ($try_oembed) {
1674                                         // html5 video and audio
1675                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1676                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1677
1678                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1679                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1680
1681                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1682                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1683                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1684                                 } else {
1685                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1686                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1687                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1688                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1689                                 }
1690
1691                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1692                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1693
1694                                 // Youtube extensions
1695                                 if ($try_oembed) {
1696                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1697                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1698                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1699                                 }
1700
1701                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1702                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1703                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1704
1705                                 if ($try_oembed) {
1706                                         $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);
1707                                 } else {
1708                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1709                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1710                                 }
1711
1712                                 if ($try_oembed) {
1713                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1714                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1715                                 }
1716
1717                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1718                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1719
1720                                 if ($try_oembed) {
1721                                         $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);
1722                                 } else {
1723                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1724                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1725                                 }
1726
1727                                 // oembed tag
1728                                 $text = OEmbed::BBCode2HTML($text);
1729
1730                                 // Avoid triple linefeeds through oembed
1731                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1732
1733                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1734                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1735                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1736                                 // start which is always required). Allow desc with a missing summary for compatibility.
1737
1738                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1739                                         $sub = Event::getHTML($ev, $simple_html);
1740
1741                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1742                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1743                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1744                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1745                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1746                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1747                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1748                                 }
1749
1750                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1751                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1752                                         // Emojis are always 4 byte Unicode characters
1753                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1754                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1755                                         }
1756                                 }
1757
1758                                 if (!$for_plaintext) {
1759                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1760                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1761                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1762                                         }
1763                                 } else {
1764                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1765                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1766                                 }
1767
1768                                 // Remove all hashtag addresses
1769                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1770                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1771                                 } elseif ($simple_html == self::DIASPORA) {
1772                                         // The ! is converted to @ since Diaspora only understands the @
1773                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1774                                                 '@<a href="$2">$3</a>',
1775                                                 $text);
1776                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1777                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1778                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1779                                                 $text);
1780                                 } elseif (!$simple_html) {
1781                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1782                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1783                                                 $text);
1784                                 }
1785
1786                                 // Bookmarks in red - will be converted to bookmarks in friendica
1787                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1788                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1789                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1790                                                         "[bookmark=$1]$2[/bookmark]", $text);
1791
1792                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1793                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1794                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1795                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1796                                 }
1797
1798                                 // Perform URL Search
1799                                 if ($try_oembed) {
1800                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1801                                 }
1802
1803                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1804
1805                                 // Handle Diaspora posts
1806                                 $text = preg_replace_callback(
1807                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1808                                         function ($match) {
1809                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1810                                         }, $text
1811                                 );
1812
1813                                 $text = preg_replace_callback(
1814                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1815                                         function ($match) {
1816                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1817                                         }, $text
1818                                 );
1819
1820                                 // Server independent link to posts and comments
1821                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1822                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1823                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1824
1825                                 /* Tag conversion
1826                                  * Supports:
1827                                  * - #[url=<anything>]<term>[/url]
1828                                  * - [url=<anything>]#<term>[/url]
1829                                  */
1830                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1831                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1832                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1833                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1834                                                         . XML::escape($matches[1]) . '</a>';
1835                                         } else {
1836                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1837                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1838                                                         . XML::escape($matches[1]) . '</a>';
1839                                         }
1840                                 }, $text);
1841
1842                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1843                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1844                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1845                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1846                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1847
1848                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1849                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1850
1851                                 // Red compatibility, though the link can't be authenticated on Friendica
1852                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1853
1854
1855                                 // we may need to restrict this further if it picks up too many strays
1856                                 // link acct:user@host to a webfinger profile redirector
1857
1858                                 $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);
1859
1860                                 // Perform MAIL Search
1861                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1862                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1863
1864                                 /// @todo What is the meaning of these lines?
1865                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1866                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1867
1868                                 // Currently deactivated, it made problems with " inside of alt texts.
1869                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1870
1871                                 // fix any escaped ampersands that may have been converted into links
1872                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1873
1874                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1875                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1876
1877                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1878
1879                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1880                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1881
1882                                 // sanitize href attributes (only allowlisted protocols URLs)
1883                                 // default value for backward compatibility
1884                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1885
1886                                 // Always allowed protocol even if config isn't set or not including it
1887                                 $allowed_link_protocols[] = '//';
1888                                 $allowed_link_protocols[] = 'http://';
1889                                 $allowed_link_protocols[] = 'https://';
1890                                 $allowed_link_protocols[] = 'redir/';
1891
1892                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1893
1894                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1895                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1896
1897                                 // Shared content
1898                                 $text = self::convertShare(
1899                                         $text,
1900                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1901                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1902                                         }
1903                                 );
1904
1905                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1906
1907                                 return $text;
1908                         }); // Escaped noparse, nobb, pre
1909
1910                         // Remove escaping tags and replace new lines that remain
1911                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1912                                 return str_replace("\n", "<br>", $match[2]);
1913                         }, $text);
1914
1915                         // Additionally, [pre] tags preserve spaces
1916                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1917                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1918                         }, $text);
1919
1920                         return $text;
1921                 }); // Escaped code
1922
1923                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1924                         function ($matches) {
1925                                 if (strpos($matches[2], "\n") !== false) {
1926                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1927                                 } else {
1928                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1929                                 }
1930
1931                                 return $return;
1932                         },
1933                         $text
1934                 );
1935
1936                 // Default iframe allowed domains/path
1937                 $allowedIframeDomains = [
1938                         DI::baseUrl()->getHostname()
1939                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1940                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1941                         'www.youtube.com/embed/',
1942                         'player.vimeo.com/video/',
1943                 ];
1944
1945                 $allowedIframeDomains = array_merge(
1946                         $allowedIframeDomains,
1947                         DI::config()->get('system', 'allowed_oembed') ?
1948                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1949                                 : []
1950                 );
1951
1952                 $text = HTML::purify($text, $allowedIframeDomains);
1953
1954                 return $text;
1955         }
1956
1957         /**
1958          * Strips the "abstract" tag from the provided text
1959          *
1960          * @param string $text The text with BBCode
1961          * @return string The same text - but without "abstract" element
1962          */
1963         public static function stripAbstract($text)
1964         {
1965                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1966                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1967
1968                 return $text;
1969         }
1970
1971         /**
1972          * Returns the value of the "abstract" element
1973          *
1974          * @param string $text The text that maybe contains the element
1975          * @param string $addon The addon for which the abstract is meant for
1976          * @return string The abstract
1977          */
1978         public static function getAbstract($text, $addon = '')
1979         {
1980                 $abstract = '';
1981                 $abstracts = [];
1982                 $addon = strtolower($addon);
1983
1984                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1985                         foreach ($results AS $result) {
1986                                 $abstracts[strtolower($result[1])] = $result[2];
1987                         }
1988                 }
1989
1990                 if (isset($abstracts[$addon])) {
1991                         $abstract = $abstracts[$addon];
1992                 }
1993
1994                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1995                         $abstract = $result[1];
1996                 }
1997
1998                 return $abstract;
1999         }
2000
2001         /**
2002          * Callback function to replace a Friendica style mention in a mention for Diaspora
2003          *
2004          * @param array $match Matching values for the callback
2005          *                     [1] = Mention type (! or @)
2006          *                     [2] = Name
2007          *                     [3] = Address
2008          * @return string Replaced mention
2009          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2010          * @throws \ImagickException
2011          */
2012         private static function bbCodeMention2DiasporaCallback($match)
2013         {
2014                 $contact = Contact::getByURL($match[3], false, ['addr']);
2015                 if (empty($contact['addr'])) {
2016                         return $match[0];
2017                 }
2018
2019                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2020                 return $mention;
2021         }
2022
2023         /**
2024          * Converts a BBCode text into Markdown
2025          *
2026          * This function converts a BBCode item body to be sent to Markdown-enabled
2027          * systems like Diaspora and Libertree
2028          *
2029          * @param string $text
2030          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2031          * @return string
2032          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2033          */
2034         public static function toMarkdown($text, $for_diaspora = true)
2035         {
2036                 $original_text = $text;
2037
2038                 // Since Diaspora is creating a summary for links, this function removes them before posting
2039                 if ($for_diaspora) {
2040                         $text = self::removeShareInformation($text);
2041                 }
2042
2043                 /**
2044                  * Transform #tags, strip off the [url] and replace spaces with underscore
2045                  */
2046                 $url_search_string = "^\[\]";
2047                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2048                         function ($matches) {
2049                                 return '#' . str_replace(' ', '_', $matches[2]);
2050                         },
2051                         $text
2052                 );
2053
2054                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2055                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2056
2057                 // Convert it to HTML - don't try oembed
2058                 if ($for_diaspora) {
2059                         $text = self::convert($text, false, self::DIASPORA);
2060
2061                         // Add all tags that maybe were removed
2062                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2063                                 $tagline = '';
2064                                 foreach ($tags[2] as $tag) {
2065                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2066                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2067                                                 $tagline .= '#' . $tag . ' ';
2068                                         }
2069                                 }
2070                                 $text = $text . " " . $tagline;
2071                         }
2072                 } else {
2073                         $text = self::convert($text, false, self::CONNECTORS);
2074                 }
2075
2076                 // If a link is followed by a quote then there should be a newline before it
2077                 // Maybe we should make this newline at every time before a quote.
2078                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2079
2080                 $stamp1 = microtime(true);
2081
2082                 // Now convert HTML to Markdown
2083                 $text = HTML::toMarkdown($text);
2084
2085                 DI::profiler()->saveTimestamp($stamp1, "parser");
2086
2087                 // Libertree has a problem with escaped hashtags.
2088                 $text = str_replace(['\#'], ['#'], $text);
2089
2090                 // Remove any leading or trailing whitespace, as this will mess up
2091                 // the Diaspora signature verification and cause the item to disappear
2092                 $text = trim($text);
2093
2094                 if ($for_diaspora) {
2095                         $url_search_string = "^\[\]";
2096                         $text = preg_replace_callback(
2097                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2098                                 ['self', 'bbCodeMention2DiasporaCallback'],
2099                                 $text
2100                         );
2101                 }
2102
2103                 Hook::callAll('bb2diaspora', $text);
2104
2105                 return $text;
2106         }
2107
2108         /**
2109          * Pull out all #hashtags and @person tags from $string.
2110          *
2111          * We also get @person@domain.com - which would make
2112          * the regex quite complicated as tags can also
2113          * end a sentence. So we'll run through our results
2114          * and strip the period from any tags which end with one.
2115          * Returns array of tags found, or empty array.
2116          *
2117          * @param string $string Post content
2118          *
2119          * @return array List of tag and person names
2120          */
2121         public static function getTags($string)
2122         {
2123                 $ret = [];
2124
2125                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2126                         // Convert hashtag links to hashtags
2127                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2128
2129                         // Force line feeds at bbtags
2130                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2131
2132                         // ignore anything in a bbtag
2133                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2134
2135                         // Match full names against @tags including the space between first and last
2136                         // We will look these up afterward to see if they are full names or not recognisable.
2137
2138                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2139                                 foreach ($matches[1] as $match) {
2140                                         if (strstr($match, ']')) {
2141                                                 // we might be inside a bbcode color tag - leave it alone
2142                                                 continue;
2143                                         }
2144
2145                                         if (substr($match, -1, 1) === '.') {
2146                                                 $ret[] = substr($match, 0, -1);
2147                                         } else {
2148                                                 $ret[] = $match;
2149                                         }
2150                                 }
2151                         }
2152
2153                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2154                         // and #hash tags.
2155
2156                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2157                                 foreach ($matches[1] as $match) {
2158                                         if (strstr($match, ']')) {
2159                                                 // we might be inside a bbcode color tag - leave it alone
2160                                                 continue;
2161                                         }
2162
2163                                         // try not to catch url fragments
2164                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2165                                                 continue;
2166                                         }
2167
2168                                         $ret[] = $match;
2169                                 }
2170                         }
2171                 });
2172
2173                 return array_unique($ret);
2174         }
2175
2176         /**
2177          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2178          *
2179          * @param string   $text
2180          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2181          * @param callable $callback
2182          * @return string
2183          * @throws Exception
2184          *@see Strings::performWithEscapedBlocks
2185          *
2186          */
2187         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2188         {
2189                 $tagList = array_map('preg_quote', $tagList);
2190
2191                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2192         }
2193
2194         /**
2195          * Replaces mentions in the provided message body for the provided user and network if any
2196          *
2197          * @param $body
2198          * @param $profile_uid
2199          * @param $network
2200          * @return string
2201          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2202          * @throws \ImagickException
2203          */
2204         public static function setMentions($body, $profile_uid = 0, $network = '')
2205         {
2206                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2207                         $tags = BBCode::getTags($body);
2208
2209                         $tagged = [];
2210                         $inform = '';
2211
2212                         foreach ($tags as $tag) {
2213                                 $tag_type = substr($tag, 0, 1);
2214
2215                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2216                                         continue;
2217                                 }
2218
2219                                 /*
2220                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2221                                  * Robert Johnson should be first in the $tags array
2222                                  */
2223                                 foreach ($tagged as $nextTag) {
2224                                         if (stristr($nextTag, $tag . ' ')) {
2225                                                 continue 2;
2226                                         }
2227                                 }
2228
2229                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2230
2231                                 if ($success['replaced']) {
2232                                         $tagged[] = $tag;
2233                                 }
2234                         }
2235
2236                         return $body;
2237                 });
2238
2239                 return $body;
2240         }
2241
2242         /**
2243          * @param string      $author  Author display name
2244          * @param string      $profile Author profile URL
2245          * @param string      $avatar  Author profile picture URL
2246          * @param string      $link    Post source URL
2247          * @param string      $posted  Post created date
2248          * @param string|null $guid    Post guid (if any)
2249          * @return string
2250          * @TODO Rewrite to handle over whole record array
2251          */
2252         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2253         {
2254                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2255                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2256                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2257                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2258                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2259
2260                 if ($guid) {
2261                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2262                 }
2263
2264                 $header  .= "']";
2265
2266                 return $header;
2267         }
2268
2269         /**
2270          * Returns the BBCode relevant to embed the provided URL in a post body.
2271          * For media type, it will return [img], [video] and [audio] tags.
2272          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2273          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2274          *
2275          * @param string      $url
2276          * @param bool        $tryAttachment
2277          * @param string|null $title
2278          * @param string|null $description
2279          * @param string|null $tags
2280          * @return string
2281          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2282          *@see ParseUrl::getSiteinfoCached
2283          *
2284          */
2285         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2286         {
2287                 DI::logger()->info($url);
2288
2289                 // If there is already some content information submitted we don't
2290                 // need to parse the url for content.
2291                 if (!empty($title) && !empty($description)) {
2292                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2293
2294                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2295
2296                         $str_tags = '';
2297                         if (!empty($tags)) {
2298                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2299                                 if (count($arr_tags)) {
2300                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2301                                 }
2302                         }
2303
2304                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2305
2306                         DI::logger()->info('(unparsed): returns: ' . $result);
2307
2308                         return $result;
2309                 }
2310
2311                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2312
2313                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2314                         switch ($siteinfo['type']) {
2315                                 case 'video':
2316                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2317                                         break;
2318                                 case 'audio':
2319                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2320                                         break;
2321                                 default:
2322                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2323                                         break;
2324                         }
2325
2326                         return $bbcode;
2327                 }
2328
2329                 unset($siteinfo['keywords']);
2330
2331                 // Bypass attachment if parse url for a comment
2332                 if (!$tryAttachment) {
2333                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2334                 }
2335
2336                 // Format it as BBCode attachment
2337                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2338         }
2339 }