src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2010-2021, the Friendica project
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\PageInfo;
  31 use Friendica\Content\Smilies;
  32 use Friendica\Core\Hook;
  33 use Friendica\Core\Logger;
  34 use Friendica\Core\Protocol;
  35 use Friendica\Core\Renderer;
  36 use Friendica\DI;
  37 use Friendica\Model\Contact;
  38 use Friendica\Model\Event;
  39 use Friendica\Model\Photo;
  40 use Friendica\Model\Tag;
  41 use Friendica\Object\Image;
  42 use Friendica\Protocol\Activity;
  43 use Friendica\Util\Images;
  44 use Friendica\Util\Map;
  45 use Friendica\Util\ParseUrl;
  46 use Friendica\Util\Proxy as ProxyUtils;
  47 use Friendica\Util\Strings;
  48 use Friendica\Util\XML;
  49
  50 class BBCode
  51 {
  52         // Update this value to the current date whenever changes are made to BBCode::convert
  53         const VERSION = '2021-05-01';
  54
  55         const INTERNAL = 0;
  56         const API = 2;
  57         const DIASPORA = 3;
  58         const CONNECTORS = 4;
  59         const OSTATUS = 7;
  60         const TWITTER = 8;
  61         const BACKLINK = 8;
  62         const ACTIVITYPUB = 9;
  63
  64         const TOP_ANCHOR = '<br class="top-anchor">';
  65         const BOTTOM_ANCHOR = '<br class="button-anchor">';
  66         /**
  67          * Fetches attachment data that were generated the old way
  68          *
  69          * @param string $body Message body
  70          * @return array
  71          *                     'type' -> Message type ('link', 'video', 'photo')
  72          *                     'text' -> Text before the shared message
  73          *                     'after' -> Text after the shared message
  74          *                     'image' -> Preview image of the message
  75          *                     'url' -> Url to the attached message
  76          *                     'title' -> Title of the attachment
  77          *                     'description' -> Description of the attachment
  78          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  79          */
  80         private static function getOldAttachmentData($body)
  81         {
  82                 $post = [];
  83
  84                 // Simplify image codes
  85                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  86
  87                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  88                         foreach ($attached as $data) {
  89                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  90                                         continue;
  91                                 }
  92
  93                                 $post['type'] = substr($data[1], 5);
  94
  95                                 $pos = strpos($body, $data[0]);
  96                                 if ($pos > 0) {
  97                                         $post['text'] = trim(substr($body, 0, $pos));
  98                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  99                                 } else {
 100                                         $post['text'] = trim(str_replace($data[0], '', $body));
 101                                         $post['after'] = '';
 102                                 }
 103
 104                                 $attacheddata = $data[2];
 105
 106                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 107
 108                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 109
 110                                         if ($picturedata) {
 111                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 112                                                         $post['image'] = $matches[1];
 113                                                 } else {
 114                                                         $post['preview'] = $matches[1];
 115                                                 }
 116                                         }
 117                                 }
 118
 119                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 120                                         $post['url'] = $matches[1];
 121                                         $post['title'] = $matches[2];
 122                                 }
 123                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 124                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 125                                         $post['url'] = $matches[1];
 126                                 }
 127
 128                                 // Search for description
 129                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 130                                         $post['description'] = $matches[1];
 131                                 }
 132                         }
 133                 }
 134                 return $post;
 135         }
 136
 137         /**
 138          * Fetches attachment data that were generated with the "attachment" element
 139          *
 140          * @param string $body Message body
 141          * @return array
 142          *                     'type' -> Message type ('link', 'video', 'photo')
 143          *                     'text' -> Text before the shared message
 144          *                     'after' -> Text after the shared message
 145          *                     'image' -> Preview image of the message
 146          *                     'url' -> Url to the attached message
 147          *                     'title' -> Title of the attachment
 148          *                     'description' -> Description of the attachment
 149          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 150          */
 151         public static function getAttachmentData($body)
 152         {
 153                 $data = [
 154                         'type'          => '',
 155                         'text'          => '',
 156                         'after'         => '',
 157                         'image'         => null,
 158                         'url'           => '',
 159                         'author_name'   => '',
 160                         'author_url'    => '',
 161                         'provider_name' => '',
 162                         'provider_url'  => '',
 163                         'title'         => '',
 164                         'description'   => '',
 165                 ];
 166
 167                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 168                         return self::getOldAttachmentData($body);
 169                 }
 170
 171                 $attributes = $match[2];
 172
 173                 $data['text'] = trim($match[1]);
 174
 175                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 176                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 177                         $value = $matches[2] ?? '';
 178
 179                         if ($value != '') {
 180                                 switch ($field) {
 181                                         case 'publisher_name':
 182                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 183                                                 break;
 184                                         case 'publisher_url':
 185                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 186                                                 break;
 187                                         case 'author_name':
 188                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 189                                                 if ($data['provider_name'] == $data['author_name']) {
 190                                                         $data['author_name'] = '';
 191                                                 }
 192                                                 break;
 193                                         case 'author_url':
 194                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 195                                                 if ($data['provider_url'] == $data['author_url']) {
 196                                                         $data['author_url'] = '';
 197                                                 }
 198                                                 break;
 199                                         case 'title':
 200                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
 201                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 202                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 203                                                 $data['title'] = $value;
 204                                         default:
 205                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 206                                                 break;
 207                                 }
 208                         }
 209                 }
 210
 211                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 212                         return [];
 213                 }
 214
 215                 $data['description'] = trim($match[3]);
 216
 217                 $data['after'] = trim($match[4]);
 218
 219                 $parts = parse_url($data['url']);
 220                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 221                         if (empty($data['provider_name'])) {
 222                                 $data['provider_name'] = $parts['host'];
 223                         }
 224                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 225                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 226
 227                                 if (!empty($parts['port'])) {
 228                                         $data['provider_url'] .= ':' . $parts['port'];
 229                                 }
 230                         }
 231                 }
 232
 233                 return $data;
 234         }
 235
 236         public static function getAttachedData($body, $item = [])
 237         {
 238                 /*
 239                 - text:
 240                 - type: link, video, photo
 241                 - title:
 242                 - url:
 243                 - image:
 244                 - description:
 245                 - (thumbnail)
 246                 */
 247
 248                 $has_title = !empty($item['title']);
 249                 $plink = $item['plink'] ?? '';
 250                 $post = self::getAttachmentData($body);
 251
 252                 // Get all linked images with alternative image description
 253                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 254                         foreach ($pictures as $picture) {
 255                                 if (Photo::isLocal($picture[1])) {
 256                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 257                                 } else {
 258                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
 259                                 }
 260                         }
 261                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 262                                 $post['image_description'] = $post['images'][0]['description'];
 263                         }
 264                 }
 265
 266                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 267                         foreach ($pictures as $picture) {
 268                                 if (Photo::isLocal($picture[1])) {
 269                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 270                                 } else {
 271                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
 272                                 }
 273                         }
 274                 }
 275
 276                 // if nothing is found, it maybe having an image.
 277                 if (!isset($post['type'])) {
 278                         // Simplify image codes
 279                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 280                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 281                         $post['text'] = $body;
 282
 283                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 284                                 if ((count($pictures) == 1) && !$has_title) {
 285                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 286                                                 // Replace the preview picture with the real picture
 287                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 288                                                 $data = ['url' => $url, 'type' => 'photo'];
 289                                         } else {
 290                                                 // Checking, if the link goes to a picture
 291                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 292                                         }
 293
 294                                         // Workaround:
 295                                         // Sometimes photo posts to the own album are not detected at the start.
 296                                         // So we seem to cannot use the cache for these cases. That's strange.
 297                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 298                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 299                                         }
 300
 301                                         if ($data['type'] == 'photo') {
 302                                                 $post['type'] = 'photo';
 303                                                 if (isset($data['images'][0])) {
 304                                                         $post['image'] = $data['images'][0]['src'];
 305                                                         $post['url'] = $data['url'];
 306                                                 } else {
 307                                                         $post['image'] = $data['url'];
 308                                                 }
 309
 310                                                 $post['preview'] = $pictures[0][2];
 311                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 312                                         } else {
 313                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 314                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 315                                                         $post['type'] = 'photo';
 316                                                         $post['image'] = $pictures[0][1];
 317                                                         $post['preview'] = $pictures[0][2];
 318                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 319                                                 }
 320                                         }
 321                                 } elseif (count($pictures) > 0) {
 322                                         $post['type'] = 'link';
 323                                         $post['url'] = $plink;
 324                                         $post['image'] = $pictures[0][2];
 325                                         $post['text'] = $body;
 326
 327                                         foreach ($pictures as $picture) {
 328                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 329                                         }
 330                                 }
 331                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 332                                 if ($has_title) {
 333                                         $post['type'] = 'link';
 334                                         $post['url'] = $plink;
 335                                 } else {
 336                                         $post['type'] = 'photo';
 337                                 }
 338
 339                                 $post['image'] = $pictures[0][1];
 340                                 $post['text'] = $body;
 341                                 foreach ($pictures as $picture) {
 342                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 343                                 }
 344                         }
 345
 346                         // Test for the external links
 347                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 348                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 349
 350                         $links = array_merge($links1, $links2);
 351
 352                         // If there is only a single one, then use it.
 353                         // This should cover link posts via API.
 354                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 355                                 $post['type'] = 'link';
 356                                 $post['url'] = $links[0][1];
 357                         }
 358
 359                         // Simplify "video" element
 360                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 361
 362                         // Now count the number of external media links
 363                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 364                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 365                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 366                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 367
 368                         // Add them to the other external links
 369                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 370
 371                         // Are there more than one?
 372                         if (count($links) > 1) {
 373                                 // The post will be the type "text", which means a blog post
 374                                 unset($post['type']);
 375                                 $post['url'] = $plink;
 376                         }
 377
 378                         if (!isset($post['type'])) {
 379                                 $post['type'] = "text";
 380                                 $post['text'] = trim($body);
 381                         }
 382
 383                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
 384                                 $post['images'] = $post['remote_images'];
 385                                 $post['image'] = $post['images'][0]['url'];
 386                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 387                                         $post['image_description'] = $post['images'][0]['description'];
 388                                 }
 389                         }
 390                         unset($post['remote_images']);
 391                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 392                         $data = ParseUrl::getSiteinfoCached($post['url']);
 393
 394                         if (isset($data['images'][0])) {
 395                                 $post['image'] = $data['images'][0]['src'];
 396                         }
 397                 }
 398
 399                 return $post;
 400         }
 401
 402         /**
 403          * Remove [attachment] BBCode and replaces it with a regular [url]
 404          *
 405          * @param string  $body
 406          * @param boolean $no_link_desc No link description
 407          *
 408          * @return string with replaced body
 409          */
 410         public static function removeAttachment($body, $no_link_desc = false)
 411         {
 412                 return preg_replace_callback("/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
 413                         function ($match) use ($no_link_desc) {
 414                                 $attach_data = self::getAttachmentData($match[0]);
 415                                 if (empty($attach_data['url'])) {
 416                                         return $match[0];
 417                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 418                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 419                                 } else {
 420                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 421                                 }
 422                 }, $body);
 423         }
 424
 425         /**
 426          * Converts a BBCode text into plaintext
 427          *
 428          * @param      $text
 429          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 430          *
 431          * @return string
 432          */
 433         public static function toPlaintext($text, $keep_urls = true)
 434         {
 435                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 436
 437                 return $naked_text;
 438         }
 439
 440         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 441         {
 442                 // Only send proxied pictures to API and for internal display
 443                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 444                         return ProxyUtils::proxifyUrl($image);
 445                 } else {
 446                         return $image;
 447                 }
 448         }
 449
 450         /**
 451          * This function changing the visual size (not the real size) of images.
 452          * The function does not work for pictures with an alternate text description.
 453          * This could only be changed by using some new "img" BBCode format.
 454          *
 455          * @param string $srctext The body with images
 456          * @return string The body with possibly scaled images
 457          */
 458         public static function scaleExternalImages(string $srctext)
 459         {
 460                 $s = $srctext;
 461
 462                 // Simplify image links
 463                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 464
 465                 $matches = null;
 466                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 467                 if ($c) {
 468                         foreach ($matches as $mtch) {
 469                                 Logger::info('scale_external_image', ['image' => $mtch[1]]);
 470
 471                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 472                                 if (stristr($mtch[1], $hostname)) {
 473                                         continue;
 474                                 }
 475
 476                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 477                                 if (!$curlResult->isSuccess()) {
 478                                         continue;
 479                                 }
 480
 481                                 $i = $curlResult->getBody();
 482                                 $type = $curlResult->getContentType();
 483                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 484
 485                                 if ($i) {
 486                                         $Image = new Image($i, $type);
 487                                         if ($Image->isValid()) {
 488                                                 $orig_width = $Image->getWidth();
 489                                                 $orig_height = $Image->getHeight();
 490
 491                                                 if ($orig_width > 640 || $orig_height > 640) {
 492                                                         $Image->scaleDown(640);
 493                                                         $new_width = $Image->getWidth();
 494                                                         $new_height = $Image->getHeight();
 495                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 496                                                         $s = str_replace(
 497                                                                 $mtch[0],
 498                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 499                                                                 . "\n",
 500                                                                 $s
 501                                                         );
 502                                                         Logger::info('New string', ['image' => $s]);
 503                                                 }
 504                                         }
 505                                 }
 506                         }
 507                 }
 508
 509                 return $s;
 510         }
 511
 512         /**
 513          * Truncates imported message body string length to max_import_size
 514          *
 515          * The purpose of this function is to apply system message length limits to
 516          * imported messages without including any embedded photos in the length
 517          *
 518          * @param string $body
 519          * @return string
 520          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 521          */
 522         public static function limitBodySize($body)
 523         {
 524                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 525
 526                 // If the length of the body, including the embedded images, is smaller
 527                 // than the maximum, then don't waste time looking for the images
 528                 if ($maxlen && (strlen($body) > $maxlen)) {
 529
 530                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 531
 532                         $orig_body = $body;
 533                         $new_body = '';
 534                         $textlen = 0;
 535
 536                         $img_start = strpos($orig_body, '[img');
 537                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 538                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 539                         while (($img_st_close !== false) && ($img_end !== false)) {
 540
 541                                 $img_st_close++; // make it point to AFTER the closing bracket
 542                                 $img_end += $img_start;
 543                                 $img_end += strlen('[/img]');
 544
 545                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 546                                         // This is an embedded image
 547
 548                                         if (($textlen + $img_start) > $maxlen) {
 549                                                 if ($textlen < $maxlen) {
 550                                                         Logger::info('the limit happens before an embedded image');
 551                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 552                                                         $textlen = $maxlen;
 553                                                 }
 554                                         } else {
 555                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 556                                                 $textlen += $img_start;
 557                                         }
 558
 559                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 560                                 } else {
 561
 562                                         if (($textlen + $img_end) > $maxlen) {
 563                                                 if ($textlen < $maxlen) {
 564                                                         Logger::info('the limit happens before the end of a non-embedded image');
 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_end);
 570                                                 $textlen += $img_end;
 571                                         }
 572                                 }
 573                                 $orig_body = substr($orig_body, $img_end);
 574
 575                                 if ($orig_body === false) {
 576                                         // in case the body ends on a closing image tag
 577                                         $orig_body = '';
 578                                 }
 579
 580                                 $img_start = strpos($orig_body, '[img');
 581                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 582                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 583                         }
 584
 585                         if (($textlen + strlen($orig_body)) > $maxlen) {
 586                                 if ($textlen < $maxlen) {
 587                                         Logger::info('the limit happens after the end of the last image');
 588                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 589                                 }
 590                         } else {
 591                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 592                                 $new_body = $new_body . $orig_body;
 593                         }
 594
 595                         return $new_body;
 596                 } else {
 597                         return $body;
 598                 }
 599         }
 600
 601         /**
 602          * Processes [attachment] tags
 603          *
 604          * Note: Can produce a [bookmark] tag in the returned string
 605          *
 606          * @param string  $text
 607          * @param integer $simplehtml
 608          * @param bool    $tryoembed
 609          * @param array   $data
 610          * @return string
 611          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 612          */
 613         public static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true, array $data = [])
 614         {
 615                 $data = $data ?: self::getAttachmentData($text);
 616                 if (empty($data) || empty($data['url'])) {
 617                         return $text;
 618                 }
 619
 620                 $stamp1 = microtime(true);
 621
 622                 if (isset($data['title'])) {
 623                         $data['title'] = strip_tags($data['title']);
 624                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 625                 } else {
 626                         $data['title'] = null;
 627                 }
 628
 629                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 630                         $data['preview'] = $data['image'];
 631                         $data['image'] = '';
 632                 }
 633
 634                 $return = '';
 635                 try {
 636                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 637                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 638                         } else {
 639                                 throw new Exception('OEmbed is disabled for this attachment.');
 640                         }
 641                 } catch (Exception $e) {
 642                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 643
 644                         if ($simplehtml != self::CONNECTORS) {
 645                                 $return = sprintf('<div class="type-%s">', $data['type']);
 646                         }
 647
 648                         if (!empty($data['title']) && !empty($data['url'])) {
 649                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 650                                         $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), $data['title']);
 651                                 } else {
 652                                         if (!empty($data['image'])) {
 653                                                 $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), $data['title']);
 654                                         } elseif (!empty($data['preview'])) {
 655                                                 $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']);
 656                                         }
 657                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 658                                 }
 659                         }
 660
 661                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 662                                 // Sanitize the HTML
 663                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 664                         }
 665
 666                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 667                                 if (!empty($data['author_name'])) {
 668                                         $return .= sprintf('<sup><a href="%s">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 669                                 } else {
 670                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 671                                 }
 672                         }
 673
 674                         if ($simplehtml != self::CONNECTORS) {
 675                                 $return .= '</div>';
 676                         }
 677                 }
 678
 679                 DI::profiler()->saveTimestamp($stamp1, 'rendering');
 680                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 681         }
 682
 683         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 684         {
 685                 $data = self::getAttachmentData($Text);
 686
 687                 if (!$data) {
 688                         return $Text;
 689                 } elseif ($nolink) {
 690                         return $data['text'] . ($data['after'] ?? '');
 691                 }
 692
 693                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 694                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 695                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 696                         $data['title'] = $data['url'];
 697                 } elseif (($text != '') && strstr($title, $text)) {
 698                         $data['text'] = $data['title'];
 699                         $data['title'] = $data['url'];
 700                 }
 701
 702                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 703                         return $data['title'] . $data['after'];
 704                 }
 705
 706                 // If the link already is included in the post, don't add it again
 707                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 708                         return $data['text'] . $data['after'];
 709                 }
 710
 711                 $text = $data['text'];
 712
 713                 if (!empty($data['url']) && !empty($data['title'])) {
 714                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 715                 } elseif (!empty($data['url'])) {
 716                         $text .= "\n[url]" . $data['url'] . '[/url]';
 717                 }
 718
 719                 return $text . "\n" . $data['after'];
 720         }
 721
 722         /**
 723          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 724          *
 725          * @param array $match Array with the matching values
 726          * @return string reformatted link including HTML codes
 727          */
 728         private static function convertUrlForActivityPubCallback($match)
 729         {
 730                 $url = $match[1];
 731
 732                 if (isset($match[2]) && ($match[1] != $match[2])) {
 733                         return $match[0];
 734                 }
 735
 736                 $parts = parse_url($url);
 737                 if (!isset($parts['scheme'])) {
 738                         return $match[0];
 739                 }
 740
 741                 return self::convertUrlForActivityPub($url);
 742         }
 743
 744         /**
 745          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 746          *
 747          * @param string $url URL that is about to be reformatted
 748          * @return string reformatted link including HTML codes
 749          */
 750         private static function convertUrlForActivityPub($url)
 751         {
 752                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 753                 return sprintf($html, $url, self::getStyledURL($url));
 754         }
 755
 756         /**
 757          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 758          *
 759          * @param string $url URL that is about to be reformatted
 760          * @return string reformatted link
 761          */
 762         private static function getStyledURL($url)
 763         {
 764                 $parts = parse_url($url);
 765                 $scheme = $parts['scheme'] . '://';
 766                 $styled_url = str_replace($scheme, '', $url);
 767
 768                 if (strlen($styled_url) > 30) {
 769                         $styled_url = substr($styled_url, 0, 30) . "…";
 770                 }
 771
 772                 return $styled_url;
 773         }
 774
 775         /*
 776          * [noparse][i]italic[/i][/noparse] turns into
 777          * [noparse][ i ]italic[ /i ][/noparse],
 778          * to hide them from parser.
 779          */
 780         private static function escapeNoparseCallback($match)
 781         {
 782                 $whole_match = $match[0];
 783                 $captured = $match[1];
 784                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 785                 $new_str = str_replace($captured, $spacefied, $whole_match);
 786                 return $new_str;
 787         }
 788
 789         /*
 790          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 791          * now turns back and the [noparse] tags are trimed
 792          * returning [i]italic[/i]
 793          */
 794         private static function unescapeNoparseCallback($match)
 795         {
 796                 $captured = $match[1];
 797                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 798                 return $unspacefied;
 799         }
 800
 801         /**
 802          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 803          * occurrences
 804          *
 805          * @param string $text        Text to search
 806          * @param string $name        Tag name
 807          * @param int    $occurrences Number of first occurrences to skip
 808          * @return boolean|array
 809          */
 810         public static function getTagPosition($text, $name, $occurrences = 0)
 811         {
 812                 if ($occurrences < 0) {
 813                         $occurrences = 0;
 814                 }
 815
 816                 $start_open = -1;
 817                 for ($i = 0; $i <= $occurrences; $i++) {
 818                         if ($start_open !== false) {
 819                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 820                         }
 821                 }
 822
 823                 if ($start_open === false) {
 824                         return false;
 825                 }
 826
 827                 $start_equal = strpos($text, '=', $start_open);
 828                 $start_close = strpos($text, ']', $start_open);
 829
 830                 if ($start_close === false) {
 831                         return false;
 832                 }
 833
 834                 $start_close++;
 835
 836                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 837
 838                 if ($end_open === false) {
 839                         return false;
 840                 }
 841
 842                 $res = [
 843                         'start' => [
 844                                 'open' => $start_open,
 845                                 'close' => $start_close
 846                         ],
 847                         'end' => [
 848                                 'open' => $end_open,
 849                                 'close' => $end_open + strlen('[/' . $name . ']')
 850                         ],
 851                 ];
 852
 853                 if ($start_equal !== false) {
 854                         $res['start']['equal'] = $start_equal + 1;
 855                 }
 856
 857                 return $res;
 858         }
 859
 860         /**
 861          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 862          *
 863          * @param string $pattern Preg pattern string
 864          * @param string $replace Preg replace string
 865          * @param string $name    BBCode tag name
 866          * @param string $text    Text to search
 867          * @return string
 868          */
 869         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 870         {
 871                 $occurrences = 0;
 872                 $pos = self::getTagPosition($text, $name, $occurrences);
 873                 while ($pos !== false && $occurrences++ < 1000) {
 874                         $start = substr($text, 0, $pos['start']['open']);
 875                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 876                         $end = substr($text, $pos['end']['close']);
 877                         if ($end === false) {
 878                                 $end = '';
 879                         }
 880
 881                         $subject = preg_replace($pattern, $replace, $subject);
 882                         $text = $start . $subject . $end;
 883
 884                         $pos = self::getTagPosition($text, $name, $occurrences);
 885                 }
 886
 887                 return $text;
 888         }
 889
 890         private static function extractImagesFromItemBody($body)
 891         {
 892                 $saved_image = [];
 893                 $orig_body = $body;
 894                 $new_body = '';
 895
 896                 $cnt = 0;
 897                 $img_start = strpos($orig_body, '[img');
 898                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 899                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 900                 while (($img_st_close !== false) && ($img_end !== false)) {
 901                         $img_st_close++; // make it point to AFTER the closing bracket
 902                         $img_end += $img_start;
 903
 904                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 905                                 // This is an embedded image
 906                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 907                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 908
 909                                 $cnt++;
 910                         } else {
 911                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 912                         }
 913
 914                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 915
 916                         if ($orig_body === false) {
 917                                 // in case the body ends on a closing image tag
 918                                 $orig_body = '';
 919                         }
 920
 921                         $img_start = strpos($orig_body, '[img');
 922                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 923                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 924                 }
 925
 926                 $new_body = $new_body . $orig_body;
 927
 928                 return ['body' => $new_body, 'images' => $saved_image];
 929         }
 930
 931         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 932         {
 933                 $newbody = $body;
 934
 935                 $cnt = 0;
 936                 foreach ($images as $image) {
 937                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 938                         // it loops over the array starting from the first element and going sequentially
 939                         // to the last element
 940                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 941                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 942                         $cnt++;
 943                 }
 944
 945                 return $newbody;
 946         }
 947
 948         /**
 949          *
 950          * @param  string   $text     A BBCode string
 951          * @return array share attributes
 952          */
 953         public static function fetchShareAttributes($text)
 954         {
 955                 $attributes = [];
 956                 if (!preg_match("/(.*?)\[share(.*?)\](.*)\[\/share\]/ism", $text, $matches)) {
 957                         return $attributes;
 958                 }
 959
 960                 $attribute_string = $matches[2];
 961                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 962                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 963                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 964                 }
 965                 return $attributes;
 966         }
 967
 968         /**
 969          * This function converts a [share] block to text according to a provided callback function whose signature is:
 970          *
 971          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 972          *
 973          * Where:
 974          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 975          * data lookup
 976          * - $author_contact is a contact record array
 977          * - $content is the inner content of the [share] block
 978          * - $is_quote_share indicates whether there's any content before the [share] block
 979          * - Return value is the string that should replace the [share] block in the provided text
 980          *
 981          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 982          *
 983          * @param  string   $text     A BBCode string
 984          * @param  callable $callback
 985          * @return string The BBCode string with all [share] blocks replaced
 986          */
 987         public static function convertShare($text, callable $callback)
 988         {
 989                 $return = preg_replace_callback(
 990                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 991                         function ($match) use ($callback) {
 992                                 $attribute_string = $match[2];
 993                                 $attributes = [];
 994                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 995                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 996                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 997                                 }
 998
 999                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
1000                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1001                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1002
1003                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1004                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1005                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1006
1007                                 if ($attributes['avatar']) {
1008                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1009                                 }
1010
1011                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1012
1013                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1014                         },
1015                         $text
1016                 );
1017
1018                 return $return;
1019         }
1020
1021         /**
1022          * Default [share] tag conversion callback
1023          *
1024          * Note: Can produce a [bookmark] tag in the output
1025          *
1026          * @see BBCode::convertShare()
1027          * @param array   $attributes     [share] block attribute values
1028          * @param array   $author_contact Contact row of the shared author
1029          * @param string  $content        Inner content of the [share] block
1030          * @param boolean $is_quote_share Whether there is content before the [share] block
1031          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1032          * @return string
1033          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1034          */
1035         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1036         {
1037                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1038
1039                 switch ($simplehtml) {
1040                         case self::API:
1041                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1042                                 break;
1043                         case self::DIASPORA:
1044                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1045                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1046                                 } else {
1047                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1048
1049                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1050                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1051                                         }
1052
1053                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1054
1055                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1056                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1057                                         }
1058                                 }
1059
1060                                 break;
1061                         case self::CONNECTORS:
1062                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1063                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1064                                 $headline .= ':</b></p>' . "\n";
1065
1066                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1067
1068                                 break;
1069                         case self::OSTATUS:
1070                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1071                                 break;
1072                         case self::ACTIVITYPUB:
1073                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1074                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1075                                 break;
1076                         default:
1077                                 $text = ($is_quote_share? "\n" : '');
1078
1079                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1080                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1081
1082                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1083                                 $text .= Renderer::replaceMacros($tpl, [
1084                                         '$profile'      => $attributes['profile'],
1085                                         '$avatar'       => $attributes['avatar'],
1086                                         '$author'       => $attributes['author'],
1087                                         '$link'         => $attributes['link'],
1088                                         '$link_title'   => DI::l10n()->t('link to source'),
1089                                         '$posted'       => $attributes['posted'],
1090                                         '$guid'         => $attributes['guid'],
1091                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1092                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1093                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
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                         $curlResult = DI::httpRequest()->head($match[1], ['timeout' => DI::config()->get('system', 'xrd_timeout')]);
1108                         if ($curlResult->isSuccess()) {
1109                                 $mimetype = $curlResult->getHeader('Content-Type');
1110                         } else {
1111                                 $mimetype = '';
1112                         }
1113
1114                         if (substr($mimetype, 0, 6) == 'image/') {
1115                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1116                         } else {
1117                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1118
1119                                 // if its not a picture then look if its a page that contains a picture link
1120                                 $body = DI::httpRequest()->fetch($match[1]);
1121                                 if (empty($body)) {
1122                                         DI::cache()->set($cache_key, $text);
1123                                         return $text;
1124                                 }
1125
1126                                 $doc = new DOMDocument();
1127                                 @$doc->loadHTML($body);
1128                                 $xpath = new DOMXPath($doc);
1129                                 $list = $xpath->query("//meta[@name]");
1130                                 foreach ($list as $node) {
1131                                         $attr = [];
1132
1133                                         if ($node->attributes->length) {
1134                                                 foreach ($node->attributes as $attribute) {
1135                                                         $attr[$attribute->name] = $attribute->value;
1136                                                 }
1137                                         }
1138
1139                                         if (strtolower($attr['name']) == 'twitter:image') {
1140                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1141                                         }
1142                                 }
1143                         }
1144                         DI::cache()->set($cache_key, $text);
1145                 }
1146
1147                 return $text;
1148         }
1149
1150         private static function expandLinksCallback($match)
1151         {
1152                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1153                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1154                 } else {
1155                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1156                 }
1157         }
1158
1159         private static function cleanPictureLinksCallback($match)
1160         {
1161                 // When the picture link is the own photo path then we can avoid fetching the link
1162                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1163                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1164                         if (!empty($match[3])) {
1165                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1166                         } else {
1167                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1168                         }
1169                         return $text;
1170                 }
1171
1172                 $cache_key = 'clean:' . $match[1];
1173                 $text = DI::cache()->get($cache_key);
1174                 if (!is_null($text)) {
1175                         return $text;
1176                 }
1177
1178                 $curlResult = DI::httpRequest()->head($match[1], ['timeout' => DI::config()->get('system', 'xrd_timeout')]);
1179                 if ($curlResult->isSuccess()) {
1180                         $mimetype = $curlResult->getHeader('Content-Type');
1181                 } else {
1182                         $mimetype = '';
1183                 }
1184
1185                 // if its a link to a picture then embed this picture
1186                 if (substr($mimetype, 0, 6) == 'image/') {
1187                         $text = '[img]' . $match[1] . '[/img]';
1188                 } else {
1189                         if (!empty($match[3])) {
1190                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1191                         } else {
1192                                 $text = '[img]' . $match[2] . '[/img]';
1193                         }
1194
1195                         // if its not a picture then look if its a page that contains a picture link
1196                         $body = DI::httpRequest()->fetch($match[1]);
1197                         if (empty($body)) {
1198                                 DI::cache()->set($cache_key, $text);
1199                                 return $text;
1200                         }
1201
1202                         $doc = new DOMDocument();
1203                         @$doc->loadHTML($body);
1204                         $xpath = new DOMXPath($doc);
1205                         $list = $xpath->query("//meta[@name]");
1206                         foreach ($list as $node) {
1207                                 $attr = [];
1208                                 if ($node->attributes->length) {
1209                                         foreach ($node->attributes as $attribute) {
1210                                                 $attr[$attribute->name] = $attribute->value;
1211                                         }
1212                                 }
1213
1214                                 if (strtolower($attr['name']) == "twitter:image") {
1215                                         if (!empty($match[3])) {
1216                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1217                                         } else {
1218                                                 $text = "[img]" . $attr['content'] . "[/img]";
1219                                         }
1220                                 }
1221                         }
1222                 }
1223                 DI::cache()->set($cache_key, $text);
1224
1225                 return $text;
1226         }
1227
1228         public static function cleanPictureLinks($text)
1229         {
1230                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1231                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1232                 return $return;
1233         }
1234
1235         public static function removeLinks(string $bbcode)
1236         {
1237                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1238                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1239
1240                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1241                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1242                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1243                 return $bbcode;
1244         }
1245
1246         /**
1247          * Converts a BBCode message to HTML message
1248          *
1249          * BBcode 2 HTML was written by WAY2WEB.net
1250          * extended to work with Mistpark/Friendica - Mike Macgirvin
1251          *
1252          * Simple HTML values meaning:
1253          * - 0: Friendica display
1254          * - 1: Unused
1255          * - 2: Used for Windows Phone push, Friendica API
1256          * - 3: Used before converting to Markdown in bb2diaspora.php
1257          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1258          * - 5: Unused
1259          * - 6: Unused
1260          * - 7: Used for dfrn, OStatus
1261          * - 8: Used for Twitter, WP backlink text setting
1262          * - 9: ActivityPub
1263          *
1264          * @param string $text
1265          * @param bool   $try_oembed
1266          * @param int    $simple_html
1267          * @param bool   $for_plaintext
1268          * @return string
1269          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1270          */
1271         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1272         {
1273                 // Accounting for null default column values
1274                 if (is_null($text) || $text === '') {
1275                         return '';
1276                 }
1277
1278                 Hook::callAll('bbcode', $text);
1279
1280                 $a = DI::app();
1281
1282                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1283                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1284                                 /*
1285                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1286                                  *
1287                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1288                                  * $match[1] = $url
1289                                  * $match[2] = $title or absent
1290                                  */
1291                                 $try_oembed_callback = function ($match)
1292                                 {
1293                                         $url = $match[1];
1294                                         $title = $match[2] ?? null;
1295
1296                                         try {
1297                                                 $return = OEmbed::getHTML($url, $title);
1298                                         } catch (Exception $ex) {
1299                                                 $return = $match[0];
1300                                         }
1301
1302                                         return $return;
1303                                 };
1304
1305                                 // Remove the abstract element. It is a non visible element.
1306                                 $text = self::stripAbstract($text);
1307
1308                                 // Line ending normalisation
1309                                 $text = str_replace("\r\n", "\n", $text);
1310
1311                                 // Move new lines outside of tags
1312                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1313                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1314
1315                                 // Extract the private images which use data urls since preg has issues with
1316                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1317                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1318
1319                                 $extracted = self::extractImagesFromItemBody($text);
1320                                 $text = $extracted['body'];
1321                                 $saved_image = $extracted['images'];
1322
1323                                 // If we find any event code, turn it into an event.
1324                                 // After we're finished processing the bbcode we'll
1325                                 // replace all of the event code with a reformatted version.
1326
1327                                 $ev = Event::fromBBCode($text);
1328
1329                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1330                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1331
1332                                 $text = str_replace("<", "&lt;", $text);
1333                                 $text = str_replace(">", "&gt;", $text);
1334
1335                                 // remove some newlines before the general conversion
1336                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1337                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1338
1339                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1340                                 if (!$try_oembed) {
1341                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1342                                 }
1343
1344                                 // Remove linefeeds inside of the table elements. See issue #6799
1345                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1346                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1347                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1348                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1349                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1350                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1351                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1352                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1353                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1354                                         "[hr]", "[hr]", "[hr]", "[hr]",
1355                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1356                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1357                                 do {
1358                                         $oldtext = $text;
1359                                         $text = str_replace($search, $replace, $text);
1360                                 } while ($oldtext != $text);
1361
1362                                 // Replace these here only once
1363                                 $search = ["\n[table]", "[/table]\n"];
1364                                 $replace = ["[table]", "[/table]"];
1365                                 $text = str_replace($search, $replace, $text);
1366
1367                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1368                                 $text = trim($text, "\n");
1369
1370                                 // removing multiplicated newlines
1371                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1372                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[*]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1373                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1374                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1375                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1376                                         do {
1377                                                 $oldtext = $text;
1378                                                 $text = str_replace($search, $replace, $text);
1379                                         } while ($oldtext != $text);
1380                                 }
1381
1382                                 // Add HTML new lines
1383                                 $text = str_replace("\n", '<br>', $text);
1384
1385                                 /// @todo Have a closer look at the different html modes
1386                                 // Handle attached links or videos
1387                                 if ($simple_html == self::ACTIVITYPUB) {
1388                                         $text = self::removeAttachment($text);
1389                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1390                                         $text = self::removeAttachment($text, true);
1391                                 } else {
1392                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1393                                 }
1394
1395                                 $nosmile = strpos($text, '[nosmile]') !== false;
1396                                 $text = str_replace('[nosmile]', '', $text);
1397
1398                                 // Replace non graphical smilies for external posts
1399                                 if (!$nosmile && !$for_plaintext) {
1400                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) {
1401                                                 return Smilies::replace($text);
1402                                         });
1403                                 }
1404
1405                                 // leave open the posibility of [map=something]
1406                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1407                                 if (strpos($text, '[/map]') !== false) {
1408                                         $text = preg_replace_callback(
1409                                                 "/\[map\](.*?)\[\/map\]/ism",
1410                                                 function ($match) use ($simple_html) {
1411                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1412                                                 },
1413                                                 $text
1414                                         );
1415                                 }
1416
1417                                 if (strpos($text, '[map=') !== false) {
1418                                         $text = preg_replace_callback(
1419                                                 "/\[map=(.*?)\]/ism",
1420                                                 function ($match) use ($simple_html) {
1421                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1422                                                 },
1423                                                 $text
1424                                         );
1425                                 }
1426
1427                                 if (strpos($text, '[map]') !== false) {
1428                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1429                                 }
1430
1431                                 // Check for headers
1432                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1433                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1434                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1435                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1436                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1437                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1438
1439                                 // Check for paragraph
1440                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1441
1442                                 // Check for bold text
1443                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1444
1445                                 // Check for Italics text
1446                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1447
1448                                 // Check for Underline text
1449                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1450
1451                                 // Check for strike-through text
1452                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1453
1454                                 // Check for over-line text
1455                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1456
1457                                 // Check for colored text
1458                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1459
1460                                 // Check for sized text
1461                                 // [size=50] --> font-size: 50px (with the unit).
1462                                 if ($simple_html != self::DIASPORA) {
1463                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1464                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1465                                 } else {
1466                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1467                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1468                                 }
1469
1470
1471                                 // Check for centered text
1472                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1473
1474                                 // Check for list text
1475                                 $text = str_replace("[*]", "<li>", $text);
1476
1477                                 // Check for style sheet commands
1478                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1479
1480                                 // Check for CSS classes
1481                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1482
1483                                 // handle nested lists
1484                                 $endlessloop = 0;
1485
1486                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1487                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1488                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1489                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1490                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1491                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1492                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1493                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1494                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1495                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1496                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1497                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1498                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1499                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1500                                 }
1501
1502                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1503                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1504                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1505                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1506
1507                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1508                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1509
1510                                 $text = str_replace('[hr]', '<hr />', $text);
1511
1512                                 if (!$for_plaintext) {
1513                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1514                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1515                                         });
1516                                 }
1517
1518                                 // Check for font change text
1519                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1520
1521                                 // Declare the format for [spoiler] layout
1522                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1523
1524                                 // Check for [spoiler] text
1525                                 // handle nested quotes
1526                                 $endlessloop = 0;
1527                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1528                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1529                                 }
1530
1531                                 // Check for [spoiler=Title] text
1532
1533                                 // handle nested quotes
1534                                 $endlessloop = 0;
1535                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1536                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1537                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1538                                                 $text);
1539                                 }
1540
1541                                 // Declare the format for [quote] layout
1542                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1543
1544                                 // Check for [quote] text
1545                                 // handle nested quotes
1546                                 $endlessloop = 0;
1547                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1548                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1549                                 }
1550
1551                                 // Check for [quote=Author] text
1552
1553                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1554
1555                                 // handle nested quotes
1556                                 $endlessloop = 0;
1557                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1558                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1559                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1560                                                 $text);
1561                                 }
1562
1563
1564                                 // [img=widthxheight]image source[/img]
1565                                 $text = preg_replace_callback(
1566                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1567                                         function ($matches) use ($simple_html) {
1568                                                 if (strpos($matches[3], "data:image/") === 0) {
1569                                                         return $matches[0];
1570                                                 }
1571
1572                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1573                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1574                                         },
1575                                         $text
1576                                 );
1577
1578                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1579                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1580
1581                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1582                                         function ($matches) use ($simple_html) {
1583                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1584                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1585                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1586                                         },
1587                                         $text);
1588
1589                                 // Images
1590                                 // [img]pathtoimage[/img]
1591                                 $text = preg_replace_callback(
1592                                         "/\[img\](.*?)\[\/img\]/ism",
1593                                         function ($matches) use ($simple_html) {
1594                                                 if (strpos($matches[1], "data:image/") === 0) {
1595                                                         return $matches[0];
1596                                                 }
1597
1598                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1599                                                 return "[img]" . $matches[1] . "[/img]";
1600                                         },
1601                                         $text
1602                                 );
1603
1604                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1605                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1606
1607                                 $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);
1608                                 $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>', $text);
1609                                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $Text);
1610
1611                                 // Simplify "video" element
1612                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1613
1614                                 if ($try_oembed) {
1615                                         // html5 video and audio
1616                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1617                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1618
1619                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1620                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1621
1622                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1623                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1624                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1625                                 } else {
1626                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1627                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1628                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1629                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1630                                 }
1631
1632                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1633                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1634
1635                                 // Youtube extensions
1636                                 if ($try_oembed) {
1637                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1638                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1639                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1640                                 }
1641
1642                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1643                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1644                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1645
1646                                 if ($try_oembed) {
1647                                         $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);
1648                                 } else {
1649                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1650                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1651                                 }
1652
1653                                 if ($try_oembed) {
1654                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1655                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1656                                 }
1657
1658                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1659                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1660
1661                                 if ($try_oembed) {
1662                                         $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);
1663                                 } else {
1664                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1665                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1666                                 }
1667
1668                                 // oembed tag
1669                                 $text = OEmbed::BBCode2HTML($text);
1670
1671                                 // Avoid triple linefeeds through oembed
1672                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1673
1674                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1675                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1676                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1677                                 // start which is always required). Allow desc with a missing summary for compatibility.
1678
1679                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1680                                         $sub = Event::getHTML($ev, $simple_html);
1681
1682                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1683                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1684                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1685                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1686                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1687                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1688                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1689                                 }
1690
1691                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1692                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1693                                         // Emojis are always 4 byte Unicode characters
1694                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1695                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1696                                         }
1697                                 }
1698
1699                                 if (!$for_plaintext) {
1700                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1701                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1702                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1703                                         }
1704                                 } else {
1705                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1706                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1707                                 }
1708
1709                                 // Remove all hashtag addresses
1710                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1711                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1712                                 } elseif ($simple_html == self::DIASPORA) {
1713                                         // The ! is converted to @ since Diaspora only understands the @
1714                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1715                                                 '@<a href="$2">$3</a>',
1716                                                 $text);
1717                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1718                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1719                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1720                                                 $text);
1721                                 } elseif (!$simple_html) {
1722                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1723                                                 '$1<a href="$2" class="userinfo mention" title="$3"><bdi>$3</bdi></a>',
1724                                                 $text);
1725                                 }
1726
1727                                 // Bookmarks in red - will be converted to bookmarks in friendica
1728                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1729                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1730                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1731                                                         "[bookmark=$1]$2[/bookmark]", $text);
1732
1733                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1734                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1735                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1736                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1737                                 }
1738
1739                                 // Perform URL Search
1740                                 if ($try_oembed) {
1741                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1742                                 }
1743
1744                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1745
1746                                 // Handle Diaspora posts
1747                                 $text = preg_replace_callback(
1748                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1749                                         function ($match) {
1750                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1751                                         }, $text
1752                                 );
1753
1754                                 $text = preg_replace_callback(
1755                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1756                                         function ($match) {
1757                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1758                                         }, $text
1759                                 );
1760
1761                                 // Server independent link to posts and comments
1762                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1763                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1764                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1765
1766                                 /* Tag conversion
1767                                  * Supports:
1768                                  * - #[url=<anything>]<term>[/url]
1769                                  * - [url=<anything>]#<term>[/url]
1770                                  */
1771                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1772                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1773                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1774                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1775                                                         . XML::escape($matches[1]) . '</a>';
1776                                         } else {
1777                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1778                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1779                                                         . XML::escape($matches[1]) . '</a>';
1780                                         }
1781                                 }, $text);
1782
1783                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1784                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1785                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1786                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1787                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1788
1789                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1790                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1791
1792                                 // Red compatibility, though the link can't be authenticated on Friendica
1793                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1794
1795
1796                                 // we may need to restrict this further if it picks up too many strays
1797                                 // link acct:user@host to a webfinger profile redirector
1798
1799                                 $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);
1800
1801                                 // Perform MAIL Search
1802                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1803                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1804
1805                                 /// @todo What is the meaning of these lines?
1806                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1807                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1808
1809                                 // Currently deactivated, it made problems with " inside of alt texts.
1810                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1811
1812                                 // fix any escaped ampersands that may have been converted into links
1813                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1814
1815                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1816                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1817
1818                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1819
1820                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1821                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1822
1823                                 // sanitize href attributes (only allowlisted protocols URLs)
1824                                 // default value for backward compatibility
1825                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1826
1827                                 // Always allowed protocol even if config isn't set or not including it
1828                                 $allowed_link_protocols[] = '//';
1829                                 $allowed_link_protocols[] = 'http://';
1830                                 $allowed_link_protocols[] = 'https://';
1831                                 $allowed_link_protocols[] = 'redir/';
1832
1833                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1834
1835                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1836                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1837
1838                                 // Shared content
1839                                 $text = self::convertShare(
1840                                         $text,
1841                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1842                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1843                                         }
1844                                 );
1845
1846                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1847
1848                                 return $text;
1849                         }); // Escaped noparse, nobb, pre
1850
1851                         // Remove escaping tags and replace new lines that remain
1852                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1853                                 return str_replace("\n", "<br>", $match[2]);
1854                         }, $text);
1855
1856                         // Additionally, [pre] tags preserve spaces
1857                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1858                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1859                         }, $text);
1860
1861                         return $text;
1862                 }); // Escaped code
1863
1864                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1865                         function ($matches) {
1866                                 if (strpos($matches[2], "\n") !== false) {
1867                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1868                                 } else {
1869                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1870                                 }
1871
1872                                 return $return;
1873                         },
1874                         $text
1875                 );
1876
1877                 // Default iframe allowed domains/path
1878                 $allowedIframeDomains = [
1879                         DI::baseUrl()->getHostname()
1880                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1881                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1882                         'www.youtube.com/embed/',
1883                         'player.vimeo.com/video/',
1884                 ];
1885
1886                 $allowedIframeDomains = array_merge(
1887                         $allowedIframeDomains,
1888                         DI::config()->get('system', 'allowed_oembed') ?
1889                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1890                                 : []
1891                 );
1892
1893                 $text = HTML::purify($text, $allowedIframeDomains);
1894
1895                 return $text;
1896         }
1897
1898         /**
1899          * Strips the "abstract" tag from the provided text
1900          *
1901          * @param string $text The text with BBCode
1902          * @return string The same text - but without "abstract" element
1903          */
1904         public static function stripAbstract($text)
1905         {
1906                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1907                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1908
1909                 return $text;
1910         }
1911
1912         /**
1913          * Returns the value of the "abstract" element
1914          *
1915          * @param string $text The text that maybe contains the element
1916          * @param string $addon The addon for which the abstract is meant for
1917          * @return string The abstract
1918          */
1919         public static function getAbstract($text, $addon = '')
1920         {
1921                 $abstract = '';
1922                 $abstracts = [];
1923                 $addon = strtolower($addon);
1924
1925                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1926                         foreach ($results AS $result) {
1927                                 $abstracts[strtolower($result[1])] = $result[2];
1928                         }
1929                 }
1930
1931                 if (isset($abstracts[$addon])) {
1932                         $abstract = $abstracts[$addon];
1933                 }
1934
1935                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1936                         $abstract = $result[1];
1937                 }
1938
1939                 return $abstract;
1940         }
1941
1942         /**
1943          * Callback function to replace a Friendica style mention in a mention for Diaspora
1944          *
1945          * @param array $match Matching values for the callback
1946          *                     [1] = Mention type (! or @)
1947          *                     [2] = Name
1948          *                     [3] = Address
1949          * @return string Replaced mention
1950          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1951          * @throws \ImagickException
1952          */
1953         private static function bbCodeMention2DiasporaCallback($match)
1954         {
1955                 $contact = Contact::getByURL($match[3], false, ['addr']);
1956                 if (empty($contact['addr'])) {
1957                         return $match[0];
1958                 }
1959
1960                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1961                 return $mention;
1962         }
1963
1964         /**
1965          * Converts a BBCode text into Markdown
1966          *
1967          * This function converts a BBCode item body to be sent to Markdown-enabled
1968          * systems like Diaspora and Libertree
1969          *
1970          * @param string $text
1971          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1972          * @return string
1973          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1974          */
1975         public static function toMarkdown($text, $for_diaspora = true)
1976         {
1977                 $original_text = $text;
1978
1979                 // Since Diaspora is creating a summary for links, this function removes them before posting
1980                 if ($for_diaspora) {
1981                         $text = self::removeShareInformation($text);
1982                 }
1983
1984                 /**
1985                  * Transform #tags, strip off the [url] and replace spaces with underscore
1986                  */
1987                 $url_search_string = "^\[\]";
1988                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1989                         function ($matches) {
1990                                 return '#' . str_replace(' ', '_', $matches[2]);
1991                         },
1992                         $text
1993                 );
1994
1995                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1996                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1997
1998                 // Convert it to HTML - don't try oembed
1999                 if ($for_diaspora) {
2000                         $text = self::convert($text, false, self::DIASPORA);
2001
2002                         // Add all tags that maybe were removed
2003                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2004                                 $tagline = '';
2005                                 foreach ($tags[2] as $tag) {
2006                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2007                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2008                                                 $tagline .= '#' . $tag . ' ';
2009                                         }
2010                                 }
2011                                 $text = $text . " " . $tagline;
2012                         }
2013                 } else {
2014                         $text = self::convert($text, false, self::CONNECTORS);
2015                 }
2016
2017                 // If a link is followed by a quote then there should be a newline before it
2018                 // Maybe we should make this newline at every time before a quote.
2019                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2020
2021                 $stamp1 = microtime(true);
2022
2023                 // Now convert HTML to Markdown
2024                 $text = HTML::toMarkdown($text);
2025
2026                 DI::profiler()->saveTimestamp($stamp1, "parser");
2027
2028                 // Libertree has a problem with escaped hashtags.
2029                 $text = str_replace(['\#'], ['#'], $text);
2030
2031                 // Remove any leading or trailing whitespace, as this will mess up
2032                 // the Diaspora signature verification and cause the item to disappear
2033                 $text = trim($text);
2034
2035                 if ($for_diaspora) {
2036                         $url_search_string = "^\[\]";
2037                         $text = preg_replace_callback(
2038                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2039                                 ['self', 'bbCodeMention2DiasporaCallback'],
2040                                 $text
2041                         );
2042                 }
2043
2044                 Hook::callAll('bb2diaspora', $text);
2045
2046                 return $text;
2047         }
2048
2049         /**
2050          * Pull out all #hashtags and @person tags from $string.
2051          *
2052          * We also get @person@domain.com - which would make
2053          * the regex quite complicated as tags can also
2054          * end a sentence. So we'll run through our results
2055          * and strip the period from any tags which end with one.
2056          * Returns array of tags found, or empty array.
2057          *
2058          * @param string $string Post content
2059          *
2060          * @return array List of tag and person names
2061          */
2062         public static function getTags($string)
2063         {
2064                 $ret = [];
2065
2066                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2067                         // Convert hashtag links to hashtags
2068                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2069
2070                         // Force line feeds at bbtags
2071                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2072
2073                         // ignore anything in a bbtag
2074                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2075
2076                         // Match full names against @tags including the space between first and last
2077                         // We will look these up afterward to see if they are full names or not recognisable.
2078
2079                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2080                                 foreach ($matches[1] as $match) {
2081                                         if (strstr($match, ']')) {
2082                                                 // we might be inside a bbcode color tag - leave it alone
2083                                                 continue;
2084                                         }
2085
2086                                         if (substr($match, -1, 1) === '.') {
2087                                                 $ret[] = substr($match, 0, -1);
2088                                         } else {
2089                                                 $ret[] = $match;
2090                                         }
2091                                 }
2092                         }
2093
2094                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2095                         // and #hash tags.
2096
2097                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2098                                 foreach ($matches[1] as $match) {
2099                                         if (strstr($match, ']')) {
2100                                                 // we might be inside a bbcode color tag - leave it alone
2101                                                 continue;
2102                                         }
2103
2104                                         // try not to catch url fragments
2105                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2106                                                 continue;
2107                                         }
2108
2109                                         $ret[] = $match;
2110                                 }
2111                         }
2112                 });
2113
2114                 return array_unique($ret);
2115         }
2116
2117         /**
2118          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2119          *
2120          * @param string   $text
2121          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2122          * @param callable $callback
2123          * @return string
2124          * @throws Exception
2125          *@see Strings::performWithEscapedBlocks
2126          *
2127          */
2128         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2129         {
2130                 $tagList = array_map('preg_quote', $tagList);
2131
2132                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2133         }
2134
2135         /**
2136          * Replaces mentions in the provided message body for the provided user and network if any
2137          *
2138          * @param $body
2139          * @param $profile_uid
2140          * @param $network
2141          * @return string
2142          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2143          * @throws \ImagickException
2144          */
2145         public static function setMentions($body, $profile_uid = 0, $network = '')
2146         {
2147                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2148                         $tags = BBCode::getTags($body);
2149
2150                         $tagged = [];
2151                         $inform = '';
2152
2153                         foreach ($tags as $tag) {
2154                                 $tag_type = substr($tag, 0, 1);
2155
2156                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2157                                         continue;
2158                                 }
2159
2160                                 /*
2161                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2162                                  * Robert Johnson should be first in the $tags array
2163                                  */
2164                                 foreach ($tagged as $nextTag) {
2165                                         if (stristr($nextTag, $tag . ' ')) {
2166                                                 continue 2;
2167                                         }
2168                                 }
2169
2170                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2171
2172                                 if ($success['replaced']) {
2173                                         $tagged[] = $tag;
2174                                 }
2175                         }
2176
2177                         return $body;
2178                 });
2179
2180                 return $body;
2181         }
2182
2183         /**
2184          * @param string      $author  Author display name
2185          * @param string      $profile Author profile URL
2186          * @param string      $avatar  Author profile picture URL
2187          * @param string      $link    Post source URL
2188          * @param string      $posted  Post created date
2189          * @param string|null $guid    Post guid (if any)
2190          * @return string
2191          * @TODO Rewrite to handle over whole record array
2192          */
2193         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2194         {
2195                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2196                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2197                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2198                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2199                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2200
2201                 if ($guid) {
2202                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2203                 }
2204
2205                 $header  .= "']";
2206
2207                 return $header;
2208         }
2209
2210         /**
2211          * Returns the BBCode relevant to embed the provided URL in a post body.
2212          * For media type, it will return [img], [video] and [audio] tags.
2213          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2214          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2215          *
2216          * @param string      $url
2217          * @param bool        $tryAttachment
2218          * @param string|null $title
2219          * @param string|null $description
2220          * @param string|null $tags
2221          * @return string
2222          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2223          *@see ParseUrl::getSiteinfoCached
2224          *
2225          */
2226         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2227         {
2228                 DI::logger()->info($url);
2229
2230                 // If there is already some content information submitted we don't
2231                 // need to parse the url for content.
2232                 if (!empty($title) && !empty($description)) {
2233                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2234
2235                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2236
2237                         $str_tags = '';
2238                         if (!empty($tags)) {
2239                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2240                                 if (count($arr_tags)) {
2241                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2242                                 }
2243                         }
2244
2245                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2246
2247                         DI::logger()->info('(unparsed): returns: ' . $result);
2248
2249                         return $result;
2250                 }
2251
2252                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2253
2254                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2255                         switch ($siteinfo['type']) {
2256                                 case 'video':
2257                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2258                                         break;
2259                                 case 'audio':
2260                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2261                                         break;
2262                                 default:
2263                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2264                                         break;
2265                         }
2266
2267                         return $bbcode;
2268                 }
2269
2270                 unset($siteinfo['keywords']);
2271
2272                 // Bypass attachment if parse url for a comment
2273                 if (!$tryAttachment) {
2274                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2275                 }
2276
2277                 // Format it as BBCode attachment
2278                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2279         }
2280 }