src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2020, Friendica
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\Smilies;
  31 use Friendica\Core\Hook;
  32 use Friendica\Core\Logger;
  33 use Friendica\Core\Protocol;
  34 use Friendica\Core\Renderer;
  35 use Friendica\Core\System;
  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\Network\Probe;
  42 use Friendica\Object\Image;
  43 use Friendica\Protocol\Activity;
  44 use Friendica\Util\Images;
  45 use Friendica\Util\Map;
  46 use Friendica\Util\Network;
  47 use Friendica\Util\ParseUrl;
  48 use Friendica\Util\Proxy as ProxyUtils;
  49 use Friendica\Util\Strings;
  50 use Friendica\Util\XML;
  51
  52 class BBCode
  53 {
  54         const INTERNAL = 0;
  55         const API = 2;
  56         const DIASPORA = 3;
  57         const CONNECTORS = 4;
  58         const OSTATUS = 7;
  59         const TWITTER = 8;
  60         const BACKLINK = 8;
  61         const ACTIVITYPUB = 9;
  62
  63         /**
  64          * Fetches attachment data that were generated the old way
  65          *
  66          * @param string $body Message body
  67          * @return array
  68          *                     'type' -> Message type ('link', 'video', 'photo')
  69          *                     'text' -> Text before the shared message
  70          *                     'after' -> Text after the shared message
  71          *                     'image' -> Preview image of the message
  72          *                     'url' -> Url to the attached message
  73          *                     'title' -> Title of the attachment
  74          *                     'description' -> Description of the attachment
  75          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  76          */
  77         private static function getOldAttachmentData($body)
  78         {
  79                 $post = [];
  80
  81                 // Simplify image codes
  82                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  83
  84                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  85                         foreach ($attached as $data) {
  86                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  87                                         continue;
  88                                 }
  89
  90                                 $post['type'] = substr($data[1], 5);
  91
  92                                 $pos = strpos($body, $data[0]);
  93                                 if ($pos > 0) {
  94                                         $post['text'] = trim(substr($body, 0, $pos));
  95                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  96                                 } else {
  97                                         $post['text'] = trim(str_replace($data[0], '', $body));
  98                                         $post['after'] = '';
  99                                 }
 100
 101                                 $attacheddata = $data[2];
 102
 103                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 104
 105                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 106
 107                                         if ($picturedata) {
 108                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 109                                                         $post['image'] = $matches[1];
 110                                                 } else {
 111                                                         $post['preview'] = $matches[1];
 112                                                 }
 113                                         }
 114                                 }
 115
 116                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 117                                         $post['url'] = $matches[1];
 118                                         $post['title'] = $matches[2];
 119                                 }
 120                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 121                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 122                                         $post['url'] = $matches[1];
 123                                 }
 124
 125                                 // Search for description
 126                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 127                                         $post['description'] = $matches[1];
 128                                 }
 129                         }
 130                 }
 131                 return $post;
 132         }
 133
 134         /**
 135          * Fetches attachment data that were generated with the "attachment" element
 136          *
 137          * @param string $body Message body
 138          * @return array
 139          *                     'type' -> Message type ('link', 'video', 'photo')
 140          *                     'text' -> Text before the shared message
 141          *                     'after' -> Text after the shared message
 142          *                     'image' -> Preview image of the message
 143          *                     'url' -> Url to the attached message
 144          *                     'title' -> Title of the attachment
 145          *                     'description' -> Description of the attachment
 146          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 147          */
 148         public static function getAttachmentData($body)
 149         {
 150                 $data = [
 151                         'type'        => '',
 152                         'text'        => '',
 153                         'after'       => '',
 154                         'image'       => null,
 155                         'url'         => '',
 156                         'title'       => '',
 157                         'description' => '',
 158                 ];
 159
 160                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 161                         return self::getOldAttachmentData($body);
 162                 }
 163
 164                 $attributes = $match[2];
 165
 166                 $data['text'] = trim($match[1]);
 167
 168                 $type = '';
 169                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 170                 if (!empty($matches[1])) {
 171                         $type = strtolower($matches[1]);
 172                 }
 173
 174                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 175                 if (!empty($matches[1])) {
 176                         $type = strtolower($matches[1]);
 177                 }
 178
 179                 if ($type == '') {
 180                         return [];
 181                 }
 182
 183                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 184                         return [];
 185                 }
 186
 187                 if ($type != '') {
 188                         $data['type'] = $type;
 189                 }
 190
 191                 $url = '';
 192                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 193                 if (!empty($matches[1])) {
 194                         $url = $matches[1];
 195                 }
 196
 197                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 198                 if (!empty($matches[1])) {
 199                         $url = $matches[1];
 200                 }
 201
 202                 if ($url != '') {
 203                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 204                 }
 205
 206                 $title = '';
 207                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 208                 if (!empty($matches[1])) {
 209                         $title = $matches[1];
 210                 }
 211
 212                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 213                 if (!empty($matches[1])) {
 214                         $title = $matches[1];
 215                 }
 216
 217                 if ($title != '') {
 218                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 219                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 220                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 221                         $data['title'] = $title;
 222                 }
 223
 224                 $image = '';
 225                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 226                 if (!empty($matches[1])) {
 227                         $image = $matches[1];
 228                 }
 229
 230                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 231                 if (!empty($matches[1])) {
 232                         $image = $matches[1];
 233                 }
 234
 235                 if ($image != '') {
 236                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 237                 }
 238
 239                 $preview = '';
 240                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 241                 if (!empty($matches[1])) {
 242                         $preview = $matches[1];
 243                 }
 244
 245                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 246                 if (!empty($matches[1])) {
 247                         $preview = $matches[1];
 248                 }
 249
 250                 if ($preview != '') {
 251                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 252                 }
 253
 254                 $data['description'] = trim($match[3]);
 255
 256                 $data['after'] = trim($match[4]);
 257
 258                 return $data;
 259         }
 260
 261         public static function getAttachedData($body, $item = [])
 262         {
 263                 /*
 264                 - text:
 265                 - type: link, video, photo
 266                 - title:
 267                 - url:
 268                 - image:
 269                 - description:
 270                 - (thumbnail)
 271                 */
 272
 273                 $has_title = !empty($item['title']);
 274                 $plink = $item['plink'] ?? '';
 275                 $post = self::getAttachmentData($body);
 276
 277                 // Get all linked images with alternative image description
 278                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 279                         foreach ($pictures as $picture) {
 280                                 if (Photo::isLocal($picture[1])) {
 281                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 282                                 }
 283                         }
 284                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 285                                 $post['image_description'] = $post['images'][0]['description'];
 286                         }
 287                 }
 288
 289                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 290                         foreach ($pictures as $picture) {
 291                                 if (Photo::isLocal($picture[1])) {
 292                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 293                                 }
 294                         }
 295                 }
 296
 297                 // if nothing is found, it maybe having an image.
 298                 if (!isset($post['type'])) {
 299                         // Simplify image codes
 300                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 301                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 302                         $post['text'] = $body;
 303
 304                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 305                                 if ((count($pictures) == 1) && !$has_title) {
 306                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 307                                                 // Replace the preview picture with the real picture
 308                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 309                                                 $data = ['url' => $url, 'type' => 'photo'];
 310                                         } else {
 311                                                 // Checking, if the link goes to a picture
 312                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 313                                         }
 314
 315                                         // Workaround:
 316                                         // Sometimes photo posts to the own album are not detected at the start.
 317                                         // So we seem to cannot use the cache for these cases. That's strange.
 318                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 319                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 320                                         }
 321
 322                                         if ($data['type'] == 'photo') {
 323                                                 $post['type'] = 'photo';
 324                                                 if (isset($data['images'][0])) {
 325                                                         $post['image'] = $data['images'][0]['src'];
 326                                                         $post['url'] = $data['url'];
 327                                                 } else {
 328                                                         $post['image'] = $data['url'];
 329                                                 }
 330
 331                                                 $post['preview'] = $pictures[0][2];
 332                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 333                                         } else {
 334                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 335                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 336                                                         $post['type'] = 'photo';
 337                                                         $post['image'] = $pictures[0][1];
 338                                                         $post['preview'] = $pictures[0][2];
 339                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 340                                                 }
 341                                         }
 342                                 } elseif (count($pictures) > 0) {
 343                                         $post['type'] = 'link';
 344                                         $post['url'] = $plink;
 345                                         $post['image'] = $pictures[0][2];
 346                                         $post['text'] = $body;
 347
 348                                         foreach ($pictures as $picture) {
 349                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 350                                         }
 351                                 }
 352                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 353                                 if ((count($pictures) == 1) && !$has_title) {
 354                                         $post['type'] = 'photo';
 355                                         $post['image'] = $pictures[0][1];
 356                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 357                                 } elseif (count($pictures) > 0) {
 358                                         $post['type'] = 'link';
 359                                         $post['url'] = $plink;
 360                                         $post['image'] = $pictures[0][1];
 361                                         $post['text'] = $body;
 362
 363                                         foreach ($pictures as $picture) {
 364                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 365                                         }
 366                                 }
 367                         }
 368
 369                         // Test for the external links
 370                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 371                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 372
 373                         $links = array_merge($links1, $links2);
 374
 375                         // If there is only a single one, then use it.
 376                         // This should cover link posts via API.
 377                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 378                                 $post['type'] = 'link';
 379                                 $post['url'] = $links[0][1];
 380                         }
 381
 382                         // Simplify "video" element
 383                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 384
 385                         // Now count the number of external media links
 386                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 387                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 388                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 389                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 390
 391                         // Add them to the other external links
 392                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 393
 394                         // Are there more than one?
 395                         if (count($links) > 1) {
 396                                 // The post will be the type "text", which means a blog post
 397                                 unset($post['type']);
 398                                 $post['url'] = $plink;
 399                         }
 400
 401                         if (!isset($post['type'])) {
 402                                 $post['type'] = "text";
 403                                 $post['text'] = trim($body);
 404                         }
 405                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 406                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 407
 408                         if (isset($data['images'][0])) {
 409                                 $post['image'] = $data['images'][0]['src'];
 410                         }
 411                 }
 412
 413                 return $post;
 414         }
 415
 416         /**
 417          * Remove [attachment] BBCode and replaces it with a regular [url]
 418          *
 419          * @param string  $body
 420          * @param boolean $no_link_desc No link description
 421          *
 422          * @return string with replaced body
 423          */
 424         public static function removeAttachment($body, $no_link_desc = false)
 425         {
 426                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 427                         function ($match) use ($no_link_desc) {
 428                                 $attach_data = self::getAttachmentData($match[0]);
 429                                 if (empty($attach_data['url'])) {
 430                                         return $match[0];
 431                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 432                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 433                                 } else {
 434                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 435                                 }
 436                 }, $body);
 437         }
 438
 439         /**
 440          * Converts a BBCode text into plaintext
 441          *
 442          * @param      $text
 443          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 444          *
 445          * @return string
 446          */
 447         public static function toPlaintext($text, $keep_urls = true)
 448         {
 449                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 450
 451                 return $naked_text;
 452         }
 453
 454         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 455         {
 456                 // Only send proxied pictures to API and for internal display
 457                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 458                         return ProxyUtils::proxifyUrl($image);
 459                 } else {
 460                         return $image;
 461                 }
 462         }
 463
 464         /**
 465          * This function changing the visual size (not the real size) of images.
 466          * The function does not work for pictures with an alternate text description.
 467          * This could only be changed by using some new "img" BBCode format.
 468          *
 469          * @param string $srctext The body with images
 470          * @return string The body with possibly scaled images
 471          */
 472         public static function scaleExternalImages(string $srctext)
 473         {
 474                 $s = $srctext;
 475
 476                 // Simplify image links
 477                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 478
 479                 $matches = null;
 480                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 481                 if ($c) {
 482                         foreach ($matches as $mtch) {
 483                                 Logger::log('scale_external_image: ' . $mtch[1]);
 484
 485                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 486                                 if (stristr($mtch[1], $hostname)) {
 487                                         continue;
 488                                 }
 489
 490                                 $curlResult = Network::curl($mtch[1], true);
 491                                 if (!$curlResult->isSuccess()) {
 492                                         continue;
 493                                 }
 494
 495                                 $i = $curlResult->getBody();
 496                                 $type = $curlResult->getContentType();
 497                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 498
 499                                 if ($i) {
 500                                         $Image = new Image($i, $type);
 501                                         if ($Image->isValid()) {
 502                                                 $orig_width = $Image->getWidth();
 503                                                 $orig_height = $Image->getHeight();
 504
 505                                                 if ($orig_width > 640 || $orig_height > 640) {
 506                                                         $Image->scaleDown(640);
 507                                                         $new_width = $Image->getWidth();
 508                                                         $new_height = $Image->getHeight();
 509                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 510                                                         $s = str_replace(
 511                                                                 $mtch[0],
 512                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 513                                                                 . "\n",
 514                                                                 $s
 515                                                         );
 516                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 517                                                 }
 518                                         }
 519                                 }
 520                         }
 521                 }
 522
 523                 return $s;
 524         }
 525
 526         /**
 527          * Truncates imported message body string length to max_import_size
 528          *
 529          * The purpose of this function is to apply system message length limits to
 530          * imported messages without including any embedded photos in the length
 531          *
 532          * @param string $body
 533          * @return string
 534          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 535          */
 536         public static function limitBodySize($body)
 537         {
 538                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 539
 540                 // If the length of the body, including the embedded images, is smaller
 541                 // than the maximum, then don't waste time looking for the images
 542                 if ($maxlen && (strlen($body) > $maxlen)) {
 543
 544                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 545
 546                         $orig_body = $body;
 547                         $new_body = '';
 548                         $textlen = 0;
 549
 550                         $img_start = strpos($orig_body, '[img');
 551                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 552                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 553                         while (($img_st_close !== false) && ($img_end !== false)) {
 554
 555                                 $img_st_close++; // make it point to AFTER the closing bracket
 556                                 $img_end += $img_start;
 557                                 $img_end += strlen('[/img]');
 558
 559                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 560                                         // This is an embedded image
 561
 562                                         if (($textlen + $img_start) > $maxlen) {
 563                                                 if ($textlen < $maxlen) {
 564                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 565                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 566                                                         $textlen = $maxlen;
 567                                                 }
 568                                         } else {
 569                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 570                                                 $textlen += $img_start;
 571                                         }
 572
 573                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 574                                 } else {
 575
 576                                         if (($textlen + $img_end) > $maxlen) {
 577                                                 if ($textlen < $maxlen) {
 578                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 579                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 580                                                         $textlen = $maxlen;
 581                                                 }
 582                                         } else {
 583                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 584                                                 $textlen += $img_end;
 585                                         }
 586                                 }
 587                                 $orig_body = substr($orig_body, $img_end);
 588
 589                                 if ($orig_body === false) {
 590                                         // in case the body ends on a closing image tag
 591                                         $orig_body = '';
 592                                 }
 593
 594                                 $img_start = strpos($orig_body, '[img');
 595                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 596                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 597                         }
 598
 599                         if (($textlen + strlen($orig_body)) > $maxlen) {
 600                                 if ($textlen < $maxlen) {
 601                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 602                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 603                                 }
 604                         } else {
 605                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 606                                 $new_body = $new_body . $orig_body;
 607                         }
 608
 609                         return $new_body;
 610                 } else {
 611                         return $body;
 612                 }
 613         }
 614
 615         /**
 616          * Processes [attachment] tags
 617          *
 618          * Note: Can produce a [bookmark] tag in the returned string
 619          *
 620          * @param string  $text
 621          * @param integer $simplehtml
 622          * @param bool    $tryoembed
 623          * @return string
 624          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 625          */
 626         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 627         {
 628                 $data = self::getAttachmentData($text);
 629                 if (empty($data) || empty($data['url'])) {
 630                         return $text;
 631                 }
 632
 633                 if (isset($data['title'])) {
 634                         $data['title'] = strip_tags($data['title']);
 635                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 636                 } else {
 637                         $data['title'] = null;
 638                 }
 639
 640                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 641                         $data['preview'] = $data['image'];
 642                         $data['image'] = '';
 643                 }
 644
 645                 $return = '';
 646                 try {
 647                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 648                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 649                         } else {
 650                                 throw new Exception('OEmbed is disabled for this attachment.');
 651                         }
 652                 } catch (Exception $e) {
 653                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 654
 655                         if ($simplehtml != self::CONNECTORS) {
 656                                 $return = sprintf('<div class="type-%s">', $data['type']);
 657                         }
 658
 659                         if (!empty($data['title']) && !empty($data['url'])) {
 660                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 661                                         $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), $data['title']);
 662                                 } else {
 663                                         if (!empty($data['image'])) {
 664                                                 $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), $data['title']);
 665                                         } elseif (!empty($data['preview'])) {
 666                                                 $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']);
 667                                         }
 668                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 669                                 }
 670                         }
 671
 672                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 673                                 // Sanitize the HTML by converting it to BBCode
 674                                 $bbcode = HTML::toBBCode($data['description']);
 675                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 676                         }
 677
 678                         if (!empty($data['url'])) {
 679                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 680                         }
 681
 682                         if ($simplehtml != self::CONNECTORS) {
 683                                 $return .= '</div>';
 684                         }
 685                 }
 686
 687                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 688         }
 689
 690         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 691         {
 692                 $data = self::getAttachmentData($Text);
 693
 694                 if (!$data) {
 695                         return $Text;
 696                 } elseif ($nolink) {
 697                         return $data['text'] . ($data['after'] ?? '');
 698                 }
 699
 700                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 701                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 702                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 703                         $data['title'] = $data['url'];
 704                 } elseif (($text != '') && strstr($title, $text)) {
 705                         $data['text'] = $data['title'];
 706                         $data['title'] = $data['url'];
 707                 }
 708
 709                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 710                         return $data['title'] . $data['after'];
 711                 }
 712
 713                 // If the link already is included in the post, don't add it again
 714                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 715                         return $data['text'] . $data['after'];
 716                 }
 717
 718                 $text = $data['text'];
 719
 720                 if (!empty($data['url']) && !empty($data['title'])) {
 721                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 722                 } elseif (!empty($data['url'])) {
 723                         $text .= "\n[url]" . $data['url'] . '[/url]';
 724                 }
 725
 726                 return $text . "\n" . $data['after'];
 727         }
 728
 729         /**
 730          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 731          *
 732          * @param array $match Array with the matching values
 733          * @return string reformatted link including HTML codes
 734          */
 735         private static function convertUrlForActivityPubCallback($match)
 736         {
 737                 $url = $match[1];
 738
 739                 if (isset($match[2]) && ($match[1] != $match[2])) {
 740                         return $match[0];
 741                 }
 742
 743                 $parts = parse_url($url);
 744                 if (!isset($parts['scheme'])) {
 745                         return $match[0];
 746                 }
 747
 748                 return self::convertUrlForActivityPub($url);
 749         }
 750
 751         /**
 752          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 753          *
 754          * @param string $url URL that is about to be reformatted
 755          * @return string reformatted link including HTML codes
 756          */
 757         private static function convertUrlForActivityPub($url)
 758         {
 759                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 760                 return sprintf($html, $url, self::getStyledURL($url));
 761         }
 762
 763         /**
 764          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 765          *
 766          * @param string $url URL that is about to be reformatted
 767          * @return string reformatted link
 768          */
 769         private static function getStyledURL($url)
 770         {
 771                 $parts = parse_url($url);
 772                 $scheme = $parts['scheme'] . '://';
 773                 $styled_url = str_replace($scheme, '', $url);
 774
 775                 if (strlen($styled_url) > 30) {
 776                         $styled_url = substr($styled_url, 0, 30) . "…";
 777                 }
 778
 779                 return $styled_url;
 780         }
 781
 782         /*
 783          * [noparse][i]italic[/i][/noparse] turns into
 784          * [noparse][ i ]italic[ /i ][/noparse],
 785          * to hide them from parser.
 786          */
 787         private static function escapeNoparseCallback($match)
 788         {
 789                 $whole_match = $match[0];
 790                 $captured = $match[1];
 791                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 792                 $new_str = str_replace($captured, $spacefied, $whole_match);
 793                 return $new_str;
 794         }
 795
 796         /*
 797          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 798          * now turns back and the [noparse] tags are trimed
 799          * returning [i]italic[/i]
 800          */
 801         private static function unescapeNoparseCallback($match)
 802         {
 803                 $captured = $match[1];
 804                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 805                 return $unspacefied;
 806         }
 807
 808         /**
 809          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 810          * occurrences
 811          *
 812          * @param string $text        Text to search
 813          * @param string $name        Tag name
 814          * @param int    $occurrences Number of first occurrences to skip
 815          * @return boolean|array
 816          */
 817         public static function getTagPosition($text, $name, $occurrences = 0)
 818         {
 819                 if ($occurrences < 0) {
 820                         $occurrences = 0;
 821                 }
 822
 823                 $start_open = -1;
 824                 for ($i = 0; $i <= $occurrences; $i++) {
 825                         if ($start_open !== false) {
 826                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 827                         }
 828                 }
 829
 830                 if ($start_open === false) {
 831                         return false;
 832                 }
 833
 834                 $start_equal = strpos($text, '=', $start_open);
 835                 $start_close = strpos($text, ']', $start_open);
 836
 837                 if ($start_close === false) {
 838                         return false;
 839                 }
 840
 841                 $start_close++;
 842
 843                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 844
 845                 if ($end_open === false) {
 846                         return false;
 847                 }
 848
 849                 $res = [
 850                         'start' => [
 851                                 'open' => $start_open,
 852                                 'close' => $start_close
 853                         ],
 854                         'end' => [
 855                                 'open' => $end_open,
 856                                 'close' => $end_open + strlen('[/' . $name . ']')
 857                         ],
 858                 ];
 859
 860                 if ($start_equal !== false) {
 861                         $res['start']['equal'] = $start_equal + 1;
 862                 }
 863
 864                 return $res;
 865         }
 866
 867         /**
 868          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 869          *
 870          * @param string $pattern Preg pattern string
 871          * @param string $replace Preg replace string
 872          * @param string $name    BBCode tag name
 873          * @param string $text    Text to search
 874          * @return string
 875          */
 876         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 877         {
 878                 $occurrences = 0;
 879                 $pos = self::getTagPosition($text, $name, $occurrences);
 880                 while ($pos !== false && $occurrences++ < 1000) {
 881                         $start = substr($text, 0, $pos['start']['open']);
 882                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 883                         $end = substr($text, $pos['end']['close']);
 884                         if ($end === false) {
 885                                 $end = '';
 886                         }
 887
 888                         $subject = preg_replace($pattern, $replace, $subject);
 889                         $text = $start . $subject . $end;
 890
 891                         $pos = self::getTagPosition($text, $name, $occurrences);
 892                 }
 893
 894                 return $text;
 895         }
 896
 897         private static function extractImagesFromItemBody($body)
 898         {
 899                 $saved_image = [];
 900                 $orig_body = $body;
 901                 $new_body = '';
 902
 903                 $cnt = 0;
 904                 $img_start = strpos($orig_body, '[img');
 905                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 906                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 907                 while (($img_st_close !== false) && ($img_end !== false)) {
 908                         $img_st_close++; // make it point to AFTER the closing bracket
 909                         $img_end += $img_start;
 910
 911                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 912                                 // This is an embedded image
 913                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 914                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 915
 916                                 $cnt++;
 917                         } else {
 918                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 919                         }
 920
 921                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 922
 923                         if ($orig_body === false) {
 924                                 // in case the body ends on a closing image tag
 925                                 $orig_body = '';
 926                         }
 927
 928                         $img_start = strpos($orig_body, '[img');
 929                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 930                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 931                 }
 932
 933                 $new_body = $new_body . $orig_body;
 934
 935                 return ['body' => $new_body, 'images' => $saved_image];
 936         }
 937
 938         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 939         {
 940                 $newbody = $body;
 941
 942                 $cnt = 0;
 943                 foreach ($images as $image) {
 944                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 945                         // it loops over the array starting from the first element and going sequentially
 946                         // to the last element
 947                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 948                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 949                         $cnt++;
 950                 }
 951
 952                 return $newbody;
 953         }
 954
 955         /**
 956          * This function converts a [share] block to text according to a provided callback function whose signature is:
 957          *
 958          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 959          *
 960          * Where:
 961          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 962          * data lookup
 963          * - $author_contact is a contact record array
 964          * - $content is the inner content of the [share] block
 965          * - $is_quote_share indicates whether there's any content before the [share] block
 966          * - Return value is the string that should replace the [share] block in the provided text
 967          *
 968          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 969          *
 970          * @param  string   $text     A BBCode string
 971          * @param  callable $callback
 972          * @return string The BBCode string with all [share] blocks replaced
 973          */
 974         public static function convertShare($text, callable $callback)
 975         {
 976                 $return = preg_replace_callback(
 977                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 978                         function ($match) use ($callback) {
 979                                 $attribute_string = $match[2];
 980                                 $attributes = [];
 981                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 982                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 983                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 984                                 }
 985
 986                                 // We only call this so that a previously unknown contact can be added.
 987                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 988                                 // This function then can fetch an entry from the contact table.
 989                                 $default['url'] = $attributes['profile'];
 990
 991                                 if (!empty($attributes['author'])) {
 992                                         $default['name'] = $attributes['author'];
 993                                 }
 994
 995                                 if (!empty($attributes['avatar'])) {
 996                                         $default['photo'] = $attributes['avatar'];
 997                                 }
 998
 999                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
1000
1001                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
1002                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1003                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1004
1005                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1006                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1007                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1008
1009                                 if ($attributes['avatar']) {
1010                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1011                                 }
1012
1013                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
1014                         },
1015                         $text
1016                 );
1017
1018                 return $return;
1019         }
1020
1021         /**
1022          * Default [share] tag conversion callback
1023          *
1024          * Note: Can produce a [bookmark] tag in the output
1025          *
1026          * @see BBCode::convertShare()
1027          * @param array   $attributes     [share] block attribute values
1028          * @param array   $author_contact Contact row of the shared author
1029          * @param string  $content        Inner content of the [share] block
1030          * @param boolean $is_quote_share Whether there is content before the [share] block
1031          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1032          * @return string
1033          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1034          */
1035         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1036         {
1037                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1038
1039                 switch ($simplehtml) {
1040                         case self::API:
1041                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1042                                 break;
1043                         case self::DIASPORA:
1044                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1045                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1046                                 } else {
1047                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1048
1049                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1050                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1051                                         }
1052
1053                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1054
1055                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1056                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1057                                         }
1058                                 }
1059
1060                                 break;
1061                         case self::CONNECTORS:
1062                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1063                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1064                                 $headline .= ':</b></p>' . "\n";
1065
1066                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1067
1068                                 break;
1069                         case self::OSTATUS:
1070                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1071                                 break;
1072                         case self::ACTIVITYPUB:
1073                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1074                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1075                                 break;
1076                         default:
1077                                 $text = ($is_quote_share? "\n" : '');
1078
1079                                 $authorId = Contact::getIdForURL($attributes['profile']);
1080
1081                                 $contact = Contact::getById($authorId, ['network']);
1082
1083                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1084                                 $text .= Renderer::replaceMacros($tpl, [
1085                                         '$profile'      => $attributes['profile'],
1086                                         '$avatar'       => $attributes['avatar'],
1087                                         '$author'       => $attributes['author'],
1088                                         '$link'         => $attributes['link'],
1089                                         '$link_title'   => DI::l10n()->t('link to source'),
1090                                         '$posted'       => $attributes['posted'],
1091                                         '$network_name' => ContactSelector::networkToName($contact['network'], $attributes['profile']),
1092                                         '$network_icon' => ContactSelector::networkToIcon($contact['network'], $attributes['profile']),
1093                                         '$content'      => self::setMentions(trim($content), 0, $contact['network']),
1094                                 ]);
1095                                 break;
1096                 }
1097
1098                 return $text;
1099         }
1100
1101         private static function removePictureLinksCallback($match)
1102         {
1103                 $cache_key = 'remove:' . $match[1];
1104                 $text = DI::cache()->get($cache_key);
1105
1106                 if (is_null($text)) {
1107                         $a = DI::app();
1108
1109                         $stamp1 = microtime(true);
1110
1111                         $ch = @curl_init($match[1]);
1112                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1113                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1114                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1115                         @curl_exec($ch);
1116                         $curl_info = @curl_getinfo($ch);
1117
1118                         DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1119
1120                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1121                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1122                         } else {
1123                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1124
1125                                 // if its not a picture then look if its a page that contains a picture link
1126                                 $body = Network::fetchUrl($match[1]);
1127
1128                                 $doc = new DOMDocument();
1129                                 @$doc->loadHTML($body);
1130                                 $xpath = new DOMXPath($doc);
1131                                 $list = $xpath->query("//meta[@name]");
1132                                 foreach ($list as $node) {
1133                                         $attr = [];
1134
1135                                         if ($node->attributes->length) {
1136                                                 foreach ($node->attributes as $attribute) {
1137                                                         $attr[$attribute->name] = $attribute->value;
1138                                                 }
1139                                         }
1140
1141                                         if (strtolower($attr['name']) == 'twitter:image') {
1142                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1143                                         }
1144                                 }
1145                         }
1146                         DI::cache()->set($cache_key, $text);
1147                 }
1148
1149                 return $text;
1150         }
1151
1152         private static function expandLinksCallback($match)
1153         {
1154                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1155                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1156                 } else {
1157                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1158                 }
1159         }
1160
1161         private static function cleanPictureLinksCallback($match)
1162         {
1163                 $a = DI::app();
1164
1165                 // When the picture link is the own photo path then we can avoid fetching the link
1166                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1167                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1168                         if (!empty($match[3])) {
1169                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1170                         } else {
1171                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1172                         }
1173                         return $text;
1174                 }
1175
1176                 $cache_key = 'clean:' . $match[1];
1177                 $text = DI::cache()->get($cache_key);
1178                 if (!is_null($text)) {
1179                         return $text;
1180                 }
1181
1182                 // Only fetch the header, not the content
1183                 $stamp1 = microtime(true);
1184
1185                 $ch = @curl_init($match[1]);
1186                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1187                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1188                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1189                 @curl_exec($ch);
1190                 $curl_info = @curl_getinfo($ch);
1191
1192                 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1193
1194                 // if its a link to a picture then embed this picture
1195                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1196                         $text = '[img]' . $match[1] . '[/img]';
1197                 } else {
1198                         if (!empty($match[3])) {
1199                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1200                         } else {
1201                                 $text = '[img]' . $match[2] . '[/img]';
1202                         }
1203
1204                         // if its not a picture then look if its a page that contains a picture link
1205                         $body = Network::fetchUrl($match[1]);
1206
1207                         $doc = new DOMDocument();
1208                         @$doc->loadHTML($body);
1209                         $xpath = new DOMXPath($doc);
1210                         $list = $xpath->query("//meta[@name]");
1211                         foreach ($list as $node) {
1212                                 $attr = [];
1213                                 if ($node->attributes->length) {
1214                                         foreach ($node->attributes as $attribute) {
1215                                                 $attr[$attribute->name] = $attribute->value;
1216                                         }
1217                                 }
1218
1219                                 if (strtolower($attr['name']) == "twitter:image") {
1220                                         if (!empty($match[3])) {
1221                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1222                                         } else {
1223                                                 $text = "[img]" . $attr['content'] . "[/img]";
1224                                         }
1225                                 }
1226                         }
1227                 }
1228                 DI::cache()->set($cache_key, $text);
1229
1230                 return $text;
1231         }
1232
1233         public static function cleanPictureLinks($text)
1234         {
1235                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1236                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1237                 return $return;
1238         }
1239
1240         /**
1241          * Converts a BBCode message to HTML message
1242          *
1243          * BBcode 2 HTML was written by WAY2WEB.net
1244          * extended to work with Mistpark/Friendica - Mike Macgirvin
1245          *
1246          * Simple HTML values meaning:
1247          * - 0: Friendica display
1248          * - 1: Unused
1249          * - 2: Used for Windows Phone push, Friendica API
1250          * - 3: Used before converting to Markdown in bb2diaspora.php
1251          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1252          * - 5: Unused
1253          * - 6: Unused
1254          * - 7: Used for dfrn, OStatus
1255          * - 8: Used for Twitter, WP backlink text setting
1256          * - 9: ActivityPub
1257          *
1258          * @param string $text
1259          * @param bool   $try_oembed
1260          * @param int    $simple_html
1261          * @param bool   $for_plaintext
1262          * @return string
1263          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1264          */
1265         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1266         {
1267                 // Accounting for null default column values
1268                 if (is_null($text) || $text === '') {
1269                         return '';
1270                 }
1271
1272                 $a = DI::app();
1273
1274                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1275                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1276                                 /*
1277                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1278                                  *
1279                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1280                                  * $match[1] = $url
1281                                  * $match[2] = $title or absent
1282                                  */
1283                                 $try_oembed_callback = function ($match)
1284                                 {
1285                                         $url = $match[1];
1286                                         $title = $match[2] ?? null;
1287
1288                                         try {
1289                                                 $return = OEmbed::getHTML($url, $title);
1290                                         } catch (Exception $ex) {
1291                                                 $return = $match[0];
1292                                         }
1293
1294                                         return $return;
1295                                 };
1296
1297
1298
1299                                 // Remove the abstract element. It is a non visible element.
1300                                 $text = self::stripAbstract($text);
1301
1302                                 // Move all spaces out of the tags
1303                                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1304                                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1305
1306                                 // Extract the private images which use data urls since preg has issues with
1307                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1308                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1309
1310                                 $extracted = self::extractImagesFromItemBody($text);
1311                                 $text = $extracted['body'];
1312                                 $saved_image = $extracted['images'];
1313
1314                                 // If we find any event code, turn it into an event.
1315                                 // After we're finished processing the bbcode we'll
1316                                 // replace all of the event code with a reformatted version.
1317
1318                                 $ev = Event::fromBBCode($text);
1319
1320                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1321                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1322
1323                                 $text = str_replace("<", "&lt;", $text);
1324                                 $text = str_replace(">", "&gt;", $text);
1325
1326                                 // remove some newlines before the general conversion
1327                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1328                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1329
1330                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1331                                 if (!$try_oembed) {
1332                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1333                                 }
1334
1335                                 // Convert new line chars to html <br /> tags
1336
1337                                 // nlbr seems to be hopelessly messed up
1338                                 //      $Text = nl2br($Text);
1339
1340                                 // We'll emulate it.
1341
1342                                 $text = trim($text);
1343                                 $text = str_replace("\r\n", "\n", $text);
1344
1345                                 // Remove linefeeds inside of the table elements. See issue #6799
1346                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1347                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1348                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1349                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1350                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1351                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1352                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1353                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1354                                 do {
1355                                         $oldtext = $text;
1356                                         $text = str_replace($search, $replace, $text);
1357                                 } while ($oldtext != $text);
1358
1359                                 // Replace these here only once
1360                                 $search = ["\n[table]", "[/table]\n"];
1361                                 $replace = ["[table]", "[/table]"];
1362                                 $text = str_replace($search, $replace, $text);
1363
1364                                 // removing multiplicated newlines
1365                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1366                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1367                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1368                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1369                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1370                                         do {
1371                                                 $oldtext = $text;
1372                                                 $text = str_replace($search, $replace, $text);
1373                                         } while ($oldtext != $text);
1374                                 }
1375
1376                                 /// @todo Have a closer look at the different html modes
1377                                 // Handle attached links or videos
1378                                 if ($simple_html == self::ACTIVITYPUB) {
1379                                         $text = self::removeAttachment($text);
1380                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1381                                         $text = self::removeAttachment($text, true);
1382                                 } else {
1383                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1384                                 }
1385
1386                                 // leave open the posibility of [map=something]
1387                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1388                                 if (strpos($text, '[/map]') !== false) {
1389                                         $text = preg_replace_callback(
1390                                                 "/\[map\](.*?)\[\/map\]/ism",
1391                                                 function ($match) use ($simple_html) {
1392                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1393                                                 },
1394                                                 $text
1395                                         );
1396                                 }
1397
1398                                 if (strpos($text, '[map=') !== false) {
1399                                         $text = preg_replace_callback(
1400                                                 "/\[map=(.*?)\]/ism",
1401                                                 function ($match) use ($simple_html) {
1402                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1403                                                 },
1404                                                 $text
1405                                         );
1406                                 }
1407
1408                                 if (strpos($text, '[map]') !== false) {
1409                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1410                                 }
1411
1412                                 // Check for headers
1413                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1414                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1415                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1416                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1417                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1418                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1419
1420                                 // Check for paragraph
1421                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1422
1423                                 // Check for bold text
1424                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1425
1426                                 // Check for Italics text
1427                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1428
1429                                 // Check for Underline text
1430                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1431
1432                                 // Check for strike-through text
1433                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1434
1435                                 // Check for over-line text
1436                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1437
1438                                 // Check for colored text
1439                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1440
1441                                 // Check for sized text
1442                                 // [size=50] --> font-size: 50px (with the unit).
1443                                 if ($simple_html != self::DIASPORA) {
1444                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1445                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1446                                 } else {
1447                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1448                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1449                                 }
1450
1451
1452                                 // Check for centered text
1453                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1454
1455                                 // Check for list text
1456                                 $text = str_replace("[*]", "<li>", $text);
1457
1458                                 // Check for style sheet commands
1459                                 $text = preg_replace_callback(
1460                                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1461                                         function ($match) {
1462                                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1463                                         },
1464                                         $text
1465                                 );
1466
1467                                 // Check for CSS classes
1468                                 $text = preg_replace_callback(
1469                                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1470                                         function ($match) {
1471                                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1472                                         },
1473                                         $text
1474                                 );
1475
1476                                 // handle nested lists
1477                                 $endlessloop = 0;
1478
1479                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1480                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1481                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1482                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1483                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1484                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1485                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1486                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1487                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1488                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1489                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1490                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1491                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1492                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1493                                 }
1494
1495                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1496                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1497                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1498                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1499
1500                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1501                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1502
1503                                 $text = str_replace('[hr]', '<hr />', $text);
1504
1505                                 if (!$for_plaintext) {
1506                                         $escaped = [];
1507
1508                                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1509                                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1510                                                 function ($matches) use (&$escaped) {
1511                                                         $return = '{escaped-' . count($escaped) . '}';
1512                                                         $escaped[] = $matches[0];
1513
1514                                                         return $return;
1515                                                 },
1516                                                 $text
1517                                         );
1518
1519                                         // Autolinker for isolated URLs
1520                                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1521
1522                                         // Restoring escaped blocks
1523                                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1524                                                 function ($matches) use ($escaped) {
1525                                                         return $escaped[intval($matches[1])] ?? $matches[0];
1526                                                 },
1527                                                 $text
1528                                         );
1529                                 }
1530
1531                                 // This is actually executed in Item::prepareBody()
1532
1533                                 $nosmile = strpos($text, '[nosmile]') !== false;
1534                                 $text = str_replace('[nosmile]', '', $text);
1535
1536                                 // Check for font change text
1537                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1538
1539                                 // Declare the format for [spoiler] layout
1540                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1541
1542                                 // Check for [spoiler] text
1543                                 // handle nested quotes
1544                                 $endlessloop = 0;
1545                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1546                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1547                                 }
1548
1549                                 // Check for [spoiler=Title] text
1550
1551                                 // handle nested quotes
1552                                 $endlessloop = 0;
1553                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1554                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1555                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1556                                                 $text);
1557                                 }
1558
1559                                 // Declare the format for [quote] layout
1560                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1561
1562                                 // Check for [quote] text
1563                                 // handle nested quotes
1564                                 $endlessloop = 0;
1565                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1566                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1567                                 }
1568
1569                                 // Check for [quote=Author] text
1570
1571                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1572
1573                                 // handle nested quotes
1574                                 $endlessloop = 0;
1575                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1576                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1577                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1578                                                 $text);
1579                                 }
1580
1581
1582                                 // [img=widthxheight]image source[/img]
1583                                 $text = preg_replace_callback(
1584                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1585                                         function ($matches) use ($simple_html) {
1586                                                 if (strpos($matches[3], "data:image/") === 0) {
1587                                                         return $matches[0];
1588                                                 }
1589
1590                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1591                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1592                                         },
1593                                         $text
1594                                 );
1595
1596                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1597                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1598
1599                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1600                                         function ($matches) use ($simple_html) {
1601                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1602                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1603                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1604                                         },
1605                                         $text);
1606
1607                                 // Images
1608                                 // [img]pathtoimage[/img]
1609                                 $text = preg_replace_callback(
1610                                         "/\[img\](.*?)\[\/img\]/ism",
1611                                         function ($matches) use ($simple_html) {
1612                                                 if (strpos($matches[1], "data:image/") === 0) {
1613                                                         return $matches[0];
1614                                                 }
1615
1616                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1617                                                 return "[img]" . $matches[1] . "[/img]";
1618                                         },
1619                                         $text
1620                                 );
1621
1622                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1623                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1624
1625                                 $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);
1626                                 $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 />', $text);
1627                                 //$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);
1628
1629                                 // Simplify "video" element
1630                                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1631
1632                                 // Try to Oembed
1633                                 if ($try_oembed) {
1634                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism", '<video src="$1" controls="controls" width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1635                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1636
1637                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1638                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1639                                 } else {
1640                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1641                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1642                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1643                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1644                                 }
1645
1646                                 // html5 video and audio
1647
1648
1649                                 if ($try_oembed) {
1650                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1651                                 } else {
1652                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1653                                 }
1654
1655                                 // Youtube extensions
1656                                 if ($try_oembed) {
1657                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1658                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1659                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1660                                 }
1661
1662                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1663                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1664                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1665
1666                                 if ($try_oembed) {
1667                                         $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);
1668                                 } else {
1669                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1670                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1671                                 }
1672
1673                                 if ($try_oembed) {
1674                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1675                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1676                                 }
1677
1678                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1679                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1680
1681                                 if ($try_oembed) {
1682                                         $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);
1683                                 } else {
1684                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1685                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1686                                 }
1687
1688                                 // oembed tag
1689                                 $text = OEmbed::BBCode2HTML($text);
1690
1691                                 // Avoid triple linefeeds through oembed
1692                                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1693
1694                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1695                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1696                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1697                                 // start which is always required). Allow desc with a missing summary for compatibility.
1698
1699                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1700                                         $sub = Event::getHTML($ev, $simple_html);
1701
1702                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1703                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1704                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1705                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1706                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1707                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1708                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1709                                 }
1710
1711                                 // Replace non graphical smilies for external posts
1712                                 if (!$nosmile && !$for_plaintext) {
1713                                         $text = Smilies::replace($text);
1714                                 }
1715
1716                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1717                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1718                                         // Emojis are always 4 byte Unicode characters
1719                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1720                                                 $text = '<span style="font-size: xx-large; line-height: initial;">' . $text . '</span>';
1721                                         }
1722                                 }
1723
1724                                 if (!$for_plaintext) {
1725                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1726                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1727                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1728                                         }
1729                                 } else {
1730                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1731                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1732                                 }
1733
1734                                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1735
1736                                 // Remove all hashtag addresses
1737                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1738                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1739                                 } elseif ($simple_html == self::DIASPORA) {
1740                                         // The ! is converted to @ since Diaspora only understands the @
1741                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1742                                                 '@<a href="$2">$3</a>',
1743                                                 $text);
1744                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1745                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1746                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1747                                                 $text);
1748                                 } elseif (!$simple_html) {
1749                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1750                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1751                                                 $text);
1752                                 }
1753
1754                                 // Bookmarks in red - will be converted to bookmarks in friendica
1755                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1756                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1757                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1758                                                         "[bookmark=$1]$2[/bookmark]", $text);
1759
1760                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1761                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1762                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1763                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1764                                 }
1765
1766                                 // Perform URL Search
1767                                 if ($try_oembed) {
1768                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1769                                 }
1770
1771                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1772
1773                                 // Handle Diaspora posts
1774                                 $text = preg_replace_callback(
1775                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1776                                         function ($match) {
1777                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1778                                         }, $text
1779                                 );
1780
1781                                 $text = preg_replace_callback(
1782                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1783                                         function ($match) {
1784                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1785                                         }, $text
1786                                 );
1787
1788                                 // Server independent link to posts and comments
1789                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1790                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1791                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1792
1793                                 /* Tag conversion
1794                                  * Supports:
1795                                  * - #[url=<anything>]<term>[/url]
1796                                  * - [url=<anything>]#<term>[/url]
1797                                  */
1798                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1799                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1800                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1801                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1802                                                         . XML::escape($matches[1]) . '</a>';
1803                                         } else {
1804                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1805                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1806                                                         . XML::escape($matches[1]) . '</a>';
1807                                         }
1808                                 }, $text);
1809
1810                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1811                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1812                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1813                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1814                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1815
1816                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1817                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1818
1819                                 // Red compatibility, though the link can't be authenticated on Friendica
1820                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1821
1822
1823                                 // we may need to restrict this further if it picks up too many strays
1824                                 // link acct:user@host to a webfinger profile redirector
1825
1826                                 $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);
1827
1828                                 // Perform MAIL Search
1829                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1830                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1831
1832                                 /// @todo What is the meaning of these lines?
1833                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1834                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1835
1836                                 // Currently deactivated, it made problems with " inside of alt texts.
1837                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1838
1839                                 // fix any escaped ampersands that may have been converted into links
1840                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1841
1842                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1843                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1844
1845                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1846
1847                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1848                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1849
1850                                 // sanitize href attributes (only allowlisted protocols URLs)
1851                                 // default value for backward compatibility
1852                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1853
1854                                 // Always allowed protocol even if config isn't set or not including it
1855                                 $allowed_link_protocols[] = '//';
1856                                 $allowed_link_protocols[] = 'http://';
1857                                 $allowed_link_protocols[] = 'https://';
1858                                 $allowed_link_protocols[] = 'redir/';
1859
1860                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1861
1862                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1863                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1864
1865                                 // Shared content
1866                                 $text = self::convertShare(
1867                                         $text,
1868                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1869                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1870                                         }
1871                                 );
1872
1873                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1874
1875                                 return $text;
1876                         }); // Escaped noparse, nobb, pre
1877
1878                         // Remove escaping tags
1879                         $text = preg_replace("/\[noparse\](.*?)\[\/noparse\]/ism", '\1', $text);
1880                         $text = preg_replace("/\[nobb\](.*?)\[\/nobb\]/ism", '\1', $text);
1881                         $text = preg_replace("/\[pre\](.*?)\[\/pre\]/ism", '\1', $text);
1882
1883                         return $text;
1884                 }); // Escaped code
1885
1886                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1887                         function ($matches) {
1888                                 if (strpos($matches[2], "\n") !== false) {
1889                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1890                                 } else {
1891                                         $return = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1892                                 }
1893
1894                                 return $return;
1895                         },
1896                         $text
1897                 );
1898
1899                 // Clean up the HTML by loading and saving the HTML with the DOM.
1900                 // Bad structured html can break a whole page.
1901                 // For performance reasons do it only with activated item cache or at export.
1902                 if (!$try_oembed || (get_itemcachepath() != '')) {
1903                         $doc = new DOMDocument();
1904                         $doc->preserveWhiteSpace = false;
1905
1906                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1907
1908                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1909                         $encoding = '<?xml encoding="UTF-8">';
1910                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1911                         $doc->encoding = 'UTF-8';
1912                         $text = $doc->saveHTML();
1913                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1914
1915                         $text = str_replace('<br></li>', '</li>', $text);
1916
1917                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1918                 }
1919
1920                 // Clean up some useless linebreaks in lists
1921                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1922                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1923                 //$Text = str_replace('</li><br />', '</li>', $Text);
1924                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1925                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1926
1927                 Hook::callAll('bbcode', $text);
1928
1929                 return trim($text);
1930         }
1931
1932         /**
1933          * Strips the "abstract" tag from the provided text
1934          *
1935          * @param string $text The text with BBCode
1936          * @return string The same text - but without "abstract" element
1937          */
1938         public static function stripAbstract($text)
1939         {
1940                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1941                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1942
1943                 return $text;
1944         }
1945
1946         /**
1947          * Returns the value of the "abstract" element
1948          *
1949          * @param string $text The text that maybe contains the element
1950          * @param string $addon The addon for which the abstract is meant for
1951          * @return string The abstract
1952          */
1953         public static function getAbstract($text, $addon = '')
1954         {
1955                 $abstract = '';
1956                 $abstracts = [];
1957                 $addon = strtolower($addon);
1958
1959                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1960                         foreach ($results AS $result) {
1961                                 $abstracts[strtolower($result[1])] = $result[2];
1962                         }
1963                 }
1964
1965                 if (isset($abstracts[$addon])) {
1966                         $abstract = $abstracts[$addon];
1967                 }
1968
1969                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1970                         $abstract = $result[1];
1971                 }
1972
1973                 return $abstract;
1974         }
1975
1976         /**
1977          * Callback function to replace a Friendica style mention in a mention for Diaspora
1978          *
1979          * @param array $match Matching values for the callback
1980          *                     [1] = Mention type (! or @)
1981          *                     [2] = Name
1982          *                     [3] = Address
1983          * @return string Replaced mention
1984          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1985          * @throws \ImagickException
1986          */
1987         private static function bbCodeMention2DiasporaCallback($match)
1988         {
1989                 $contact = Contact::getDetailsByURL($match[3]);
1990
1991                 if (empty($contact['addr'])) {
1992                         $contact = Probe::uri($match[3]);
1993                 }
1994
1995                 if (empty($contact['addr'])) {
1996                         return $match[0];
1997                 }
1998
1999                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2000                 return $mention;
2001         }
2002
2003         /**
2004          * Converts a BBCode text into Markdown
2005          *
2006          * This function converts a BBCode item body to be sent to Markdown-enabled
2007          * systems like Diaspora and Libertree
2008          *
2009          * @param string $text
2010          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2011          * @return string
2012          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2013          */
2014         public static function toMarkdown($text, $for_diaspora = true)
2015         {
2016                 $original_text = $text;
2017
2018                 // Since Diaspora is creating a summary for links, this function removes them before posting
2019                 if ($for_diaspora) {
2020                         $text = self::removeShareInformation($text);
2021                 }
2022
2023                 /**
2024                  * Transform #tags, strip off the [url] and replace spaces with underscore
2025                  */
2026                 $url_search_string = "^\[\]";
2027                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2028                         function ($matches) {
2029                                 return '#' . str_replace(' ', '_', $matches[2]);
2030                         },
2031                         $text
2032                 );
2033
2034                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2035                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2036
2037                 // Convert it to HTML - don't try oembed
2038                 if ($for_diaspora) {
2039                         $text = self::convert($text, false, self::DIASPORA);
2040
2041                         // Add all tags that maybe were removed
2042                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2043                                 $tagline = '';
2044                                 foreach ($tags[2] as $tag) {
2045                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2046                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2047                                                 $tagline .= '#' . $tag . ' ';
2048                                         }
2049                                 }
2050                                 $text = $text . " " . $tagline;
2051                         }
2052                 } else {
2053                         $text = self::convert($text, false, self::CONNECTORS);
2054                 }
2055
2056                 // If a link is followed by a quote then there should be a newline before it
2057                 // Maybe we should make this newline at every time before a quote.
2058                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2059
2060                 $stamp1 = microtime(true);
2061
2062                 // Now convert HTML to Markdown
2063                 $text = HTML::toMarkdown($text);
2064
2065                 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2066
2067                 // Libertree has a problem with escaped hashtags.
2068                 $text = str_replace(['\#'], ['#'], $text);
2069
2070                 // Remove any leading or trailing whitespace, as this will mess up
2071                 // the Diaspora signature verification and cause the item to disappear
2072                 $text = trim($text);
2073
2074                 if ($for_diaspora) {
2075                         $url_search_string = "^\[\]";
2076                         $text = preg_replace_callback(
2077                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2078                                 ['self', 'bbCodeMention2DiasporaCallback'],
2079                                 $text
2080                         );
2081                 }
2082
2083                 Hook::callAll('bb2diaspora', $text);
2084
2085                 return $text;
2086         }
2087
2088         /**
2089          * Pull out all #hashtags and @person tags from $string.
2090          *
2091          * We also get @person@domain.com - which would make
2092          * the regex quite complicated as tags can also
2093          * end a sentence. So we'll run through our results
2094          * and strip the period from any tags which end with one.
2095          * Returns array of tags found, or empty array.
2096          *
2097          * @param string $string Post content
2098          *
2099          * @return array List of tag and person names
2100          */
2101         public static function getTags($string)
2102         {
2103                 $ret = [];
2104
2105                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code'], function ($string) use (&$ret) {
2106                         // Convert hashtag links to hashtags
2107                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2108
2109                         // Force line feeds at bbtags
2110                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2111
2112                         // ignore anything in a bbtag
2113                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2114
2115                         // Match full names against @tags including the space between first and last
2116                         // We will look these up afterward to see if they are full names or not recognisable.
2117
2118                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2119                                 foreach ($matches[1] as $match) {
2120                                         if (strstr($match, ']')) {
2121                                                 // we might be inside a bbcode color tag - leave it alone
2122                                                 continue;
2123                                         }
2124
2125                                         if (substr($match, -1, 1) === '.') {
2126                                                 $ret[] = substr($match, 0, -1);
2127                                         } else {
2128                                                 $ret[] = $match;
2129                                         }
2130                                 }
2131                         }
2132
2133                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2134                         // and #hash tags.
2135
2136                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2137                                 foreach ($matches[1] as $match) {
2138                                         if (strstr($match, ']')) {
2139                                                 // we might be inside a bbcode color tag - leave it alone
2140                                                 continue;
2141                                         }
2142
2143                                         // ignore strictly numeric tags like #1
2144                                         if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2145                                                 continue;
2146                                         }
2147
2148                                         // try not to catch url fragments
2149                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2150                                                 continue;
2151                                         }
2152
2153                                         $ret[] = $match;
2154                                 }
2155                         }
2156                 });
2157
2158                 return array_unique($ret);
2159         }
2160
2161         /**
2162          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2163          *
2164          * @param string   $text
2165          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2166          * @param callable $callback
2167          * @return string
2168          * @throws Exception
2169          *@see Strings::performWithEscapedBlocks
2170          *
2171          */
2172         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2173         {
2174                 $tagList = array_map('preg_quote', $tagList);
2175
2176                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2177         }
2178
2179         /**
2180          * Replaces mentions in the provided message body for the provided user and network if any
2181          *
2182          * @param $body
2183          * @param $profile_uid
2184          * @param $network
2185          * @return string
2186          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2187          * @throws \ImagickException
2188          */
2189         public static function setMentions($body, $profile_uid = 0, $network = '')
2190         {
2191                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2192                         $tags = BBCode::getTags($body);
2193
2194                         $tagged = [];
2195                         $inform = '';
2196
2197                         foreach ($tags as $tag) {
2198                                 $tag_type = substr($tag, 0, 1);
2199
2200                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2201                                         continue;
2202                                 }
2203
2204                                 /*
2205                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2206                                  * Robert Johnson should be first in the $tags array
2207                                  */
2208                                 foreach ($tagged as $nextTag) {
2209                                         if (stristr($nextTag, $tag . ' ')) {
2210                                                 continue 2;
2211                                         }
2212                                 }
2213
2214                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2215
2216                                 if ($success['replaced']) {
2217                                         $tagged[] = $tag;
2218                                 }
2219                         }
2220
2221                         return $body;
2222                 });
2223
2224                 return $body;
2225         }
2226 }