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 = '2021-05-21';
  54
  55         const INTERNAL = 0;
  56         const EXTERNAL = 1;
  57         const API = 2;
  58         const DIASPORA = 3;
  59         const CONNECTORS = 4;
  60         const OSTATUS = 7;
  61         const TWITTER = 8;
  62         const BACKLINK = 8;
  63         const ACTIVITYPUB = 9;
  64
  65         const TOP_ANCHOR = '<br class="top-anchor">';
  66         const BOTTOM_ANCHOR = '<br class="button-anchor">';
  67         /**
  68          * Fetches attachment data that were generated the old way
  69          *
  70          * @param string $body Message body
  71          * @return array
  72          *                     'type' -> Message type ('link', 'video', 'photo')
  73          *                     'text' -> Text before the shared message
  74          *                     'after' -> Text after the shared message
  75          *                     'image' -> Preview image of the message
  76          *                     'url' -> Url to the attached message
  77          *                     'title' -> Title of the attachment
  78          *                     'description' -> Description of the attachment
  79          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  80          */
  81         private static function getOldAttachmentData($body)
  82         {
  83                 $post = [];
  84
  85                 // Simplify image codes
  86                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  87
  88                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  89                         foreach ($attached as $data) {
  90                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  91                                         continue;
  92                                 }
  93
  94                                 $post['type'] = substr($data[1], 5);
  95
  96                                 $pos = strpos($body, $data[0]);
  97                                 if ($pos > 0) {
  98                                         $post['text'] = trim(substr($body, 0, $pos));
  99                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
 100                                 } else {
 101                                         $post['text'] = trim(str_replace($data[0], '', $body));
 102                                         $post['after'] = '';
 103                                 }
 104
 105                                 $attacheddata = $data[2];
 106
 107                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 108
 109                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 110
 111                                         if ($picturedata) {
 112                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 113                                                         $post['image'] = $matches[1];
 114                                                 } else {
 115                                                         $post['preview'] = $matches[1];
 116                                                 }
 117                                         }
 118                                 }
 119
 120                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 121                                         $post['url'] = $matches[1];
 122                                         $post['title'] = $matches[2];
 123                                 }
 124                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 125                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 126                                         $post['url'] = $matches[1];
 127                                 }
 128
 129                                 // Search for description
 130                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 131                                         $post['description'] = $matches[1];
 132                                 }
 133                         }
 134                 }
 135                 return $post;
 136         }
 137
 138         /**
 139          * Fetches attachment data that were generated with the "attachment" element
 140          *
 141          * @param string $body Message body
 142          * @return array
 143          *                     'type' -> Message type ('link', 'video', 'photo')
 144          *                     'text' -> Text before the shared message
 145          *                     'after' -> Text after the shared message
 146          *                     'image' -> Preview image of the message
 147          *                     'url' -> Url to the attached message
 148          *                     'title' -> Title of the attachment
 149          *                     'description' -> Description of the attachment
 150          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 151          */
 152         public static function getAttachmentData($body)
 153         {
 154                 $data = [
 155                         'type'          => '',
 156                         'text'          => '',
 157                         'after'         => '',
 158                         'image'         => null,
 159                         'url'           => '',
 160                         'author_name'   => '',
 161                         'author_url'    => '',
 162                         'provider_name' => '',
 163                         'provider_url'  => '',
 164                         'title'         => '',
 165                         'description'   => '',
 166                 ];
 167
 168                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 169                         return self::getOldAttachmentData($body);
 170                 }
 171
 172                 $attributes = $match[2];
 173
 174                 $data['text'] = trim($match[1]);
 175
 176                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 177                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 178                         $value = $matches[2] ?? '';
 179
 180                         if ($value != '') {
 181                                 switch ($field) {
 182                                         case 'publisher_name':
 183                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 184                                                 break;
 185                                         case 'publisher_url':
 186                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 187                                                 break;
 188                                         case 'author_name':
 189                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 190                                                 if ($data['provider_name'] == $data['author_name']) {
 191                                                         $data['author_name'] = '';
 192                                                 }
 193                                                 break;
 194                                         case 'author_url':
 195                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 196                                                 if ($data['provider_url'] == $data['author_url']) {
 197                                                         $data['author_url'] = '';
 198                                                 }
 199                                                 break;
 200                                         case 'title':
 201                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
 202                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 203                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 204                                                 $data['title'] = $value;
 205                                         default:
 206                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 207                                                 break;
 208                                 }
 209                         }
 210                 }
 211
 212                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 213                         return [];
 214                 }
 215
 216                 $data['description'] = trim($match[3]);
 217
 218                 $data['after'] = trim($match[4]);
 219
 220                 $parts = parse_url($data['url']);
 221                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 222                         if (empty($data['provider_name'])) {
 223                                 $data['provider_name'] = $parts['host'];
 224                         }
 225                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 226                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 227
 228                                 if (!empty($parts['port'])) {
 229                                         $data['provider_url'] .= ':' . $parts['port'];
 230                                 }
 231                         }
 232                 }
 233
 234                 return $data;
 235         }
 236
 237         public static function getAttachedData($body, $item = [])
 238         {
 239                 /*
 240                 - text:
 241                 - type: link, video, photo
 242                 - title:
 243                 - url:
 244                 - image:
 245                 - description:
 246                 - (thumbnail)
 247                 */
 248
 249                 $has_title = !empty($item['title']);
 250                 $plink = $item['plink'] ?? '';
 251                 $post = self::getAttachmentData($body);
 252
 253                 // Get all linked images with alternative image description
 254                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 255                         foreach ($pictures as $picture) {
 256                                 if (Photo::isLocal($picture[1])) {
 257                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 258                                 } else {
 259                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
 260                                 }
 261                         }
 262                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 263                                 $post['image_description'] = $post['images'][0]['description'];
 264                         }
 265                 }
 266
 267                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 268                         foreach ($pictures as $picture) {
 269                                 if (Photo::isLocal($picture[1])) {
 270                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 271                                 } else {
 272                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
 273                                 }
 274                         }
 275                 }
 276
 277                 // if nothing is found, it maybe having an image.
 278                 if (!isset($post['type'])) {
 279                         // Simplify image codes
 280                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 281                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 282                         $post['text'] = $body;
 283
 284                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 285                                 if ((count($pictures) == 1) && !$has_title) {
 286                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 287                                                 // Replace the preview picture with the real picture
 288                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 289                                                 $data = ['url' => $url, 'type' => 'photo'];
 290                                         } else {
 291                                                 // Checking, if the link goes to a picture
 292                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 293                                         }
 294
 295                                         // Workaround:
 296                                         // Sometimes photo posts to the own album are not detected at the start.
 297                                         // So we seem to cannot use the cache for these cases. That's strange.
 298                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 299                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 300                                         }
 301
 302                                         if ($data['type'] == 'photo') {
 303                                                 $post['type'] = 'photo';
 304                                                 if (isset($data['images'][0])) {
 305                                                         $post['image'] = $data['images'][0]['src'];
 306                                                         $post['url'] = $data['url'];
 307                                                 } else {
 308                                                         $post['image'] = $data['url'];
 309                                                 }
 310
 311                                                 $post['preview'] = $pictures[0][2];
 312                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 313                                         } else {
 314                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 315                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 316                                                         $post['type'] = 'photo';
 317                                                         $post['image'] = $pictures[0][1];
 318                                                         $post['preview'] = $pictures[0][2];
 319                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 320                                                 }
 321                                         }
 322                                 } elseif (count($pictures) > 0) {
 323                                         $post['type'] = 'link';
 324                                         $post['url'] = $plink;
 325                                         $post['image'] = $pictures[0][2];
 326                                         $post['text'] = $body;
 327
 328                                         foreach ($pictures as $picture) {
 329                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 330                                         }
 331                                 }
 332                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 333                                 if ($has_title) {
 334                                         $post['type'] = 'link';
 335                                         $post['url'] = $plink;
 336                                 } else {
 337                                         $post['type'] = 'photo';
 338                                 }
 339
 340                                 $post['image'] = $pictures[0][1];
 341                                 $post['text'] = $body;
 342                                 foreach ($pictures as $picture) {
 343                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 344                                 }
 345                         }
 346
 347                         // Test for the external links
 348                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 349                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 350
 351                         $links = array_merge($links1, $links2);
 352
 353                         // If there is only a single one, then use it.
 354                         // This should cover link posts via API.
 355                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 356                                 $post['type'] = 'link';
 357                                 $post['url'] = $links[0][1];
 358                         }
 359
 360                         // Simplify "video" element
 361                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 362
 363                         // Now count the number of external media links
 364                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 365                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 366                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 367                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 368
 369                         // Add them to the other external links
 370                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 371
 372                         // Are there more than one?
 373                         if (count($links) > 1) {
 374                                 // The post will be the type "text", which means a blog post
 375                                 unset($post['type']);
 376                                 $post['url'] = $plink;
 377                         }
 378
 379                         if (!isset($post['type'])) {
 380                                 $post['type'] = "text";
 381                                 $post['text'] = trim($body);
 382                         }
 383
 384                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
 385                                 $post['images'] = $post['remote_images'];
 386                                 $post['image'] = $post['images'][0]['url'];
 387                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 388                                         $post['image_description'] = $post['images'][0]['description'];
 389                                 }
 390                         }
 391                         unset($post['remote_images']);
 392                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 393                         $data = ParseUrl::getSiteinfoCached($post['url']);
 394
 395                         if (isset($data['images'][0])) {
 396                                 $post['image'] = $data['images'][0]['src'];
 397                         }
 398                 }
 399
 400                 return $post;
 401         }
 402
 403         /**
 404          * Remove [attachment] BBCode and replaces it with a regular [url]
 405          *
 406          * @param string  $body
 407          * @param boolean $no_link_desc No link description
 408          *
 409          * @return string with replaced body
 410          */
 411         public static function removeAttachment($body, $no_link_desc = false)
 412         {
 413                 return preg_replace_callback("/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
 414                         function ($match) use ($no_link_desc) {
 415                                 $attach_data = self::getAttachmentData($match[0]);
 416                                 if (empty($attach_data['url'])) {
 417                                         return $match[0];
 418                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 419                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 420                                 } else {
 421                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 422                                 }
 423                 }, $body);
 424         }
 425
 426         /**
 427          * Converts a BBCode text into plaintext
 428          *
 429          * @param      $text
 430          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 431          *
 432          * @return string
 433          */
 434         public static function toPlaintext($text, $keep_urls = true)
 435         {
 436                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 437
 438                 return $naked_text;
 439         }
 440
 441         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 442         {
 443                 // Only send proxied pictures to API and for internal display
 444                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 445                         return ProxyUtils::proxifyUrl($image);
 446                 } else {
 447                         return $image;
 448                 }
 449         }
 450
 451         /**
 452          * This function changing the visual size (not the real size) of images.
 453          * The function does not work for pictures with an alternate text description.
 454          * This could only be changed by using some new "img" BBCode format.
 455          *
 456          * @param string $srctext The body with images
 457          * @return string The body with possibly scaled images
 458          */
 459         public static function scaleExternalImages(string $srctext)
 460         {
 461                 $s = $srctext;
 462
 463                 // Simplify image links
 464                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 465
 466                 $matches = null;
 467                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 468                 if ($c) {
 469                         foreach ($matches as $mtch) {
 470                                 Logger::info('scale_external_image', ['image' => $mtch[1]]);
 471
 472                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 473                                 if (stristr($mtch[1], $hostname)) {
 474                                         continue;
 475                                 }
 476
 477                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 478                                 if (!$curlResult->isSuccess()) {
 479                                         continue;
 480                                 }
 481
 482                                 $i = $curlResult->getBody();
 483                                 $type = $curlResult->getContentType();
 484                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 485
 486                                 if ($i) {
 487                                         $Image = new Image($i, $type);
 488                                         if ($Image->isValid()) {
 489                                                 $orig_width = $Image->getWidth();
 490                                                 $orig_height = $Image->getHeight();
 491
 492                                                 if ($orig_width > 640 || $orig_height > 640) {
 493                                                         $Image->scaleDown(640);
 494                                                         $new_width = $Image->getWidth();
 495                                                         $new_height = $Image->getHeight();
 496                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 497                                                         $s = str_replace(
 498                                                                 $mtch[0],
 499                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 500                                                                 . "\n",
 501                                                                 $s
 502                                                         );
 503                                                         Logger::info('New string', ['image' => $s]);
 504                                                 }
 505                                         }
 506                                 }
 507                         }
 508                 }
 509
 510                 return $s;
 511         }
 512
 513         /**
 514          * Truncates imported message body string length to max_import_size
 515          *
 516          * The purpose of this function is to apply system message length limits to
 517          * imported messages without including any embedded photos in the length
 518          *
 519          * @param string $body
 520          * @return string
 521          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 522          */
 523         public static function limitBodySize($body)
 524         {
 525                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 526
 527                 // If the length of the body, including the embedded images, is smaller
 528                 // than the maximum, then don't waste time looking for the images
 529                 if ($maxlen && (strlen($body) > $maxlen)) {
 530
 531                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 532
 533                         $orig_body = $body;
 534                         $new_body = '';
 535                         $textlen = 0;
 536
 537                         $img_start = strpos($orig_body, '[img');
 538                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 539                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 540                         while (($img_st_close !== false) && ($img_end !== false)) {
 541
 542                                 $img_st_close++; // make it point to AFTER the closing bracket
 543                                 $img_end += $img_start;
 544                                 $img_end += strlen('[/img]');
 545
 546                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 547                                         // This is an embedded image
 548
 549                                         if (($textlen + $img_start) > $maxlen) {
 550                                                 if ($textlen < $maxlen) {
 551                                                         Logger::info('the limit happens before an embedded image');
 552                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 553                                                         $textlen = $maxlen;
 554                                                 }
 555                                         } else {
 556                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 557                                                 $textlen += $img_start;
 558                                         }
 559
 560                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 561                                 } else {
 562
 563                                         if (($textlen + $img_end) > $maxlen) {
 564                                                 if ($textlen < $maxlen) {
 565                                                         Logger::info('the limit happens before the end of a non-embedded image');
 566                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 567                                                         $textlen = $maxlen;
 568                                                 }
 569                                         } else {
 570                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 571                                                 $textlen += $img_end;
 572                                         }
 573                                 }
 574                                 $orig_body = substr($orig_body, $img_end);
 575
 576                                 if ($orig_body === false) {
 577                                         // in case the body ends on a closing image tag
 578                                         $orig_body = '';
 579                                 }
 580
 581                                 $img_start = strpos($orig_body, '[img');
 582                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 583                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 584                         }
 585
 586                         if (($textlen + strlen($orig_body)) > $maxlen) {
 587                                 if ($textlen < $maxlen) {
 588                                         Logger::info('the limit happens after the end of the last image');
 589                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 590                                 }
 591                         } else {
 592                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 593                                 $new_body = $new_body . $orig_body;
 594                         }
 595
 596                         return $new_body;
 597                 } else {
 598                         return $body;
 599                 }
 600         }
 601
 602         /**
 603          * Processes [attachment] tags
 604          *
 605          * Note: Can produce a [bookmark] tag in the returned string
 606          *
 607          * @param string  $text
 608          * @param integer $simplehtml
 609          * @param bool    $tryoembed
 610          * @param array   $data
 611          * @return string
 612          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 613          */
 614         public static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true, array $data = [])
 615         {
 616                 $data = $data ?: self::getAttachmentData($text);
 617                 if (empty($data) || empty($data['url'])) {
 618                         return $text;
 619                 }
 620
 621                 $stamp1 = microtime(true);
 622
 623                 if (isset($data['title'])) {
 624                         $data['title'] = strip_tags($data['title']);
 625                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 626                 } else {
 627                         $data['title'] = null;
 628                 }
 629
 630                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 631                         $data['preview'] = $data['image'];
 632                         $data['image'] = '';
 633                 }
 634
 635                 $return = '';
 636                 try {
 637                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 638                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 639                         } else {
 640                                 throw new Exception('OEmbed is disabled for this attachment.');
 641                         }
 642                 } catch (Exception $e) {
 643                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 644
 645                         if ($simplehtml != self::CONNECTORS) {
 646                                 $return = sprintf('<div class="type-%s">', $data['type']);
 647                         }
 648
 649                         if (!empty($data['title']) && !empty($data['url'])) {
 650                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 651                                         $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), $data['title']);
 652                                 } else {
 653                                         if (!empty($data['image'])) {
 654                                                 $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), $data['title']);
 655                                         } elseif (!empty($data['preview'])) {
 656                                                 $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']);
 657                                         }
 658                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 659                                 }
 660                         }
 661
 662                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 663                                 // Sanitize the HTML
 664                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 665                         }
 666
 667                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 668                                 if (!empty($data['author_name'])) {
 669                                         $return .= sprintf('<sup><a href="%s">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 670                                 } else {
 671                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 672                                 }
 673                         }
 674
 675                         if ($simplehtml != self::CONNECTORS) {
 676                                 $return .= '</div>';
 677                         }
 678                 }
 679
 680                 DI::profiler()->saveTimestamp($stamp1, 'rendering');
 681                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 682         }
 683
 684         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 685         {
 686                 $data = self::getAttachmentData($Text);
 687
 688                 if (!$data) {
 689                         return $Text;
 690                 } elseif ($nolink) {
 691                         return $data['text'] . ($data['after'] ?? '');
 692                 }
 693
 694                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 695                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 696                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 697                         $data['title'] = $data['url'];
 698                 } elseif (($text != '') && strstr($title, $text)) {
 699                         $data['text'] = $data['title'];
 700                         $data['title'] = $data['url'];
 701                 }
 702
 703                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 704                         return $data['title'] . $data['after'];
 705                 }
 706
 707                 // If the link already is included in the post, don't add it again
 708                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 709                         return $data['text'] . $data['after'];
 710                 }
 711
 712                 $text = $data['text'];
 713
 714                 if (!empty($data['url']) && !empty($data['title'])) {
 715                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 716                 } elseif (!empty($data['url'])) {
 717                         $text .= "\n[url]" . $data['url'] . '[/url]';
 718                 }
 719
 720                 return $text . "\n" . $data['after'];
 721         }
 722
 723         /**
 724          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 725          *
 726          * @param array $match Array with the matching values
 727          * @return string reformatted link including HTML codes
 728          */
 729         private static function convertUrlForActivityPubCallback($match)
 730         {
 731                 $url = $match[1];
 732
 733                 if (isset($match[2]) && ($match[1] != $match[2])) {
 734                         return $match[0];
 735                 }
 736
 737                 $parts = parse_url($url);
 738                 if (!isset($parts['scheme'])) {
 739                         return $match[0];
 740                 }
 741
 742                 return self::convertUrlForActivityPub($url);
 743         }
 744
 745         /**
 746          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 747          *
 748          * @param string $url URL that is about to be reformatted
 749          * @return string reformatted link including HTML codes
 750          */
 751         private static function convertUrlForActivityPub($url)
 752         {
 753                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 754                 return sprintf($html, $url, self::getStyledURL($url));
 755         }
 756
 757         /**
 758          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 759          *
 760          * @param string $url URL that is about to be reformatted
 761          * @return string reformatted link
 762          */
 763         private static function getStyledURL($url)
 764         {
 765                 $parts = parse_url($url);
 766                 $scheme = $parts['scheme'] . '://';
 767                 $styled_url = str_replace($scheme, '', $url);
 768
 769                 if (strlen($styled_url) > 30) {
 770                         $styled_url = substr($styled_url, 0, 30) . "…";
 771                 }
 772
 773                 return $styled_url;
 774         }
 775
 776         /*
 777          * [noparse][i]italic[/i][/noparse] turns into
 778          * [noparse][ i ]italic[ /i ][/noparse],
 779          * to hide them from parser.
 780          */
 781         private static function escapeNoparseCallback($match)
 782         {
 783                 $whole_match = $match[0];
 784                 $captured = $match[1];
 785                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 786                 $new_str = str_replace($captured, $spacefied, $whole_match);
 787                 return $new_str;
 788         }
 789
 790         /*
 791          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 792          * now turns back and the [noparse] tags are trimed
 793          * returning [i]italic[/i]
 794          */
 795         private static function unescapeNoparseCallback($match)
 796         {
 797                 $captured = $match[1];
 798                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 799                 return $unspacefied;
 800         }
 801
 802         /**
 803          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 804          * occurrences
 805          *
 806          * @param string $text        Text to search
 807          * @param string $name        Tag name
 808          * @param int    $occurrences Number of first occurrences to skip
 809          * @return boolean|array
 810          */
 811         public static function getTagPosition($text, $name, $occurrences = 0)
 812         {
 813                 if ($occurrences < 0) {
 814                         $occurrences = 0;
 815                 }
 816
 817                 $start_open = -1;
 818                 for ($i = 0; $i <= $occurrences; $i++) {
 819                         if ($start_open !== false) {
 820                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 821                         }
 822                 }
 823
 824                 if ($start_open === false) {
 825                         return false;
 826                 }
 827
 828                 $start_equal = strpos($text, '=', $start_open);
 829                 $start_close = strpos($text, ']', $start_open);
 830
 831                 if ($start_close === false) {
 832                         return false;
 833                 }
 834
 835                 $start_close++;
 836
 837                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 838
 839                 if ($end_open === false) {
 840                         return false;
 841                 }
 842
 843                 $res = [
 844                         'start' => [
 845                                 'open' => $start_open,
 846                                 'close' => $start_close
 847                         ],
 848                         'end' => [
 849                                 'open' => $end_open,
 850                                 'close' => $end_open + strlen('[/' . $name . ']')
 851                         ],
 852                 ];
 853
 854                 if ($start_equal !== false) {
 855                         $res['start']['equal'] = $start_equal + 1;
 856                 }
 857
 858                 return $res;
 859         }
 860
 861         /**
 862          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 863          *
 864          * @param string $pattern Preg pattern string
 865          * @param string $replace Preg replace string
 866          * @param string $name    BBCode tag name
 867          * @param string $text    Text to search
 868          * @return string
 869          */
 870         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 871         {
 872                 $occurrences = 0;
 873                 $pos = self::getTagPosition($text, $name, $occurrences);
 874                 while ($pos !== false && $occurrences++ < 1000) {
 875                         $start = substr($text, 0, $pos['start']['open']);
 876                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 877                         $end = substr($text, $pos['end']['close']);
 878                         if ($end === false) {
 879                                 $end = '';
 880                         }
 881
 882                         $subject = preg_replace($pattern, $replace, $subject);
 883                         $text = $start . $subject . $end;
 884
 885                         $pos = self::getTagPosition($text, $name, $occurrences);
 886                 }
 887
 888                 return $text;
 889         }
 890
 891         private static function extractImagesFromItemBody($body)
 892         {
 893                 $saved_image = [];
 894                 $orig_body = $body;
 895                 $new_body = '';
 896
 897                 $cnt = 0;
 898                 $img_start = strpos($orig_body, '[img');
 899                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 900                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 901                 while (($img_st_close !== false) && ($img_end !== false)) {
 902                         $img_st_close++; // make it point to AFTER the closing bracket
 903                         $img_end += $img_start;
 904
 905                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 906                                 // This is an embedded image
 907                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 908                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 909
 910                                 $cnt++;
 911                         } else {
 912                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 913                         }
 914
 915                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 916
 917                         if ($orig_body === false) {
 918                                 // in case the body ends on a closing image tag
 919                                 $orig_body = '';
 920                         }
 921
 922                         $img_start = strpos($orig_body, '[img');
 923                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 924                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 925                 }
 926
 927                 $new_body = $new_body . $orig_body;
 928
 929                 return ['body' => $new_body, 'images' => $saved_image];
 930         }
 931
 932         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 933         {
 934                 $newbody = $body;
 935
 936                 $cnt = 0;
 937                 foreach ($images as $image) {
 938                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 939                         // it loops over the array starting from the first element and going sequentially
 940                         // to the last element
 941                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 942                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 943                         $cnt++;
 944                 }
 945
 946                 return $newbody;
 947         }
 948
 949         /**
 950          *
 951          * @param  string   $text     A BBCode string
 952          * @return array share attributes
 953          */
 954         public static function fetchShareAttributes($text)
 955         {
 956                 $attributes = [];
 957                 if (!preg_match("/(.*?)\[share(.*?)\](.*)\[\/share\]/ism", $text, $matches)) {
 958                         return $attributes;
 959                 }
 960
 961                 $attribute_string = $matches[2];
 962                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 963                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 964                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 965                 }
 966                 return $attributes;
 967         }
 968
 969         /**
 970          * This function converts a [share] block to text according to a provided callback function whose signature is:
 971          *
 972          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 973          *
 974          * Where:
 975          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 976          * data lookup
 977          * - $author_contact is a contact record array
 978          * - $content is the inner content of the [share] block
 979          * - $is_quote_share indicates whether there's any content before the [share] block
 980          * - Return value is the string that should replace the [share] block in the provided text
 981          *
 982          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 983          *
 984          * @param  string   $text     A BBCode string
 985          * @param  callable $callback
 986          * @return string The BBCode string with all [share] blocks replaced
 987          */
 988         public static function convertShare($text, callable $callback)
 989         {
 990                 $return = preg_replace_callback(
 991                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 992                         function ($match) use ($callback) {
 993                                 $attribute_string = $match[2];
 994                                 $attributes = [];
 995                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 996                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 997                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 998                                 }
 999
1000                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
1001                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1002                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1003
1004                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1005                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1006                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1007
1008                                 if (!empty($author_contact['id'])) {
1009                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], ProxyUtils::SIZE_THUMB);
1010                                 } elseif ($attributes['avatar']) {
1011                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1012                                 }
1013
1014                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1015
1016                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1017                         },
1018                         $text
1019                 );
1020
1021                 return $return;
1022         }
1023
1024         /**
1025          * Default [share] tag conversion callback
1026          *
1027          * Note: Can produce a [bookmark] tag in the output
1028          *
1029          * @see BBCode::convertShare()
1030          * @param array   $attributes     [share] block attribute values
1031          * @param array   $author_contact Contact row of the shared author
1032          * @param string  $content        Inner content of the [share] block
1033          * @param boolean $is_quote_share Whether there is content before the [share] block
1034          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1035          * @return string
1036          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1037          */
1038         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1039         {
1040                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1041
1042                 switch ($simplehtml) {
1043                         case self::API:
1044                                 $text = ($is_quote_share? '<br>' : '') .
1045                                 '<p><b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b> </p>\n" .
1046                                 '<blockquote class="shared_content">' . $content . '</blockquote>';
1047                                 break;
1048                         case self::DIASPORA:
1049                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1050                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1051                                 } else {
1052                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1053
1054                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1055                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1056                                         }
1057
1058                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1059
1060                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1061                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1062                                         }
1063                                 }
1064
1065                                 break;
1066                         case self::CONNECTORS:
1067                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1068                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1069                                 $headline .= ':</b></p>' . "\n";
1070
1071                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1072
1073                                 break;
1074                         case self::OSTATUS:
1075                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1076                                 break;
1077                         case self::ACTIVITYPUB:
1078                                 $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>';
1079                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1080                                 break;
1081                         default:
1082                                 $text = ($is_quote_share? "\n" : '');
1083
1084                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1085                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1086
1087                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1088                                 $text .= Renderer::replaceMacros($tpl, [
1089                                         '$profile'      => $attributes['profile'],
1090                                         '$avatar'       => $attributes['avatar'],
1091                                         '$author'       => $attributes['author'],
1092                                         '$link'         => $attributes['link'],
1093                                         '$link_title'   => DI::l10n()->t('Link to source'),
1094                                         '$posted'       => $attributes['posted'],
1095                                         '$guid'         => $attributes['guid'],
1096                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1097                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1098                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
1099                                 ]);
1100                                 break;
1101                 }
1102
1103                 return $text;
1104         }
1105
1106         private static function removePictureLinksCallback($match)
1107         {
1108                 $cache_key = 'remove:' . $match[1];
1109                 $text = DI::cache()->get($cache_key);
1110
1111                 if (is_null($text)) {
1112                         $curlResult = DI::httpRequest()->head($match[1], ['timeout' => DI::config()->get('system', 'xrd_timeout')]);
1113                         if ($curlResult->isSuccess()) {
1114                                 $mimetype = $curlResult->getHeader('Content-Type');
1115                         } else {
1116                                 $mimetype = '';
1117                         }
1118
1119                         if (substr($mimetype, 0, 6) == 'image/') {
1120                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1121                         } else {
1122                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1123
1124                                 // if its not a picture then look if its a page that contains a picture link
1125                                 $body = DI::httpRequest()->fetch($match[1]);
1126                                 if (empty($body)) {
1127                                         DI::cache()->set($cache_key, $text);
1128                                         return $text;
1129                                 }
1130
1131                                 $doc = new DOMDocument();
1132                                 @$doc->loadHTML($body);
1133                                 $xpath = new DOMXPath($doc);
1134                                 $list = $xpath->query("//meta[@name]");
1135                                 foreach ($list as $node) {
1136                                         $attr = [];
1137
1138                                         if ($node->attributes->length) {
1139                                                 foreach ($node->attributes as $attribute) {
1140                                                         $attr[$attribute->name] = $attribute->value;
1141                                                 }
1142                                         }
1143
1144                                         if (strtolower($attr['name']) == 'twitter:image') {
1145                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1146                                         }
1147                                 }
1148                         }
1149                         DI::cache()->set($cache_key, $text);
1150                 }
1151
1152                 return $text;
1153         }
1154
1155         private static function expandLinksCallback($match)
1156         {
1157                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1158                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1159                 } else {
1160                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1161                 }
1162         }
1163
1164         private static function cleanPictureLinksCallback($match)
1165         {
1166                 // When the picture link is the own photo path then we can avoid fetching the link
1167                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1168                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1169                         if (!empty($match[3])) {
1170                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1171                         } else {
1172                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1173                         }
1174                         return $text;
1175                 }
1176
1177                 $cache_key = 'clean:' . $match[1];
1178                 $text = DI::cache()->get($cache_key);
1179                 if (!is_null($text)) {
1180                         return $text;
1181                 }
1182
1183                 $curlResult = DI::httpRequest()->head($match[1], ['timeout' => DI::config()->get('system', 'xrd_timeout')]);
1184                 if ($curlResult->isSuccess()) {
1185                         $mimetype = $curlResult->getHeader('Content-Type');
1186                 } else {
1187                         $mimetype = '';
1188                 }
1189
1190                 // if its a link to a picture then embed this picture
1191                 if (substr($mimetype, 0, 6) == 'image/') {
1192                         $text = '[img]' . $match[1] . '[/img]';
1193                 } else {
1194                         if (!empty($match[3])) {
1195                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1196                         } else {
1197                                 $text = '[img]' . $match[2] . '[/img]';
1198                         }
1199
1200                         // if its not a picture then look if its a page that contains a picture link
1201                         $body = DI::httpRequest()->fetch($match[1]);
1202                         if (empty($body)) {
1203                                 DI::cache()->set($cache_key, $text);
1204                                 return $text;
1205                         }
1206
1207                         $doc = new DOMDocument();
1208                         @$doc->loadHTML($body);
1209                         $xpath = new DOMXPath($doc);
1210                         $list = $xpath->query("//meta[@name]");
1211                         foreach ($list as $node) {
1212                                 $attr = [];
1213                                 if ($node->attributes->length) {
1214                                         foreach ($node->attributes as $attribute) {
1215                                                 $attr[$attribute->name] = $attribute->value;
1216                                         }
1217                                 }
1218
1219                                 if (strtolower($attr['name']) == "twitter:image") {
1220                                         if (!empty($match[3])) {
1221                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1222                                         } else {
1223                                                 $text = "[img]" . $attr['content'] . "[/img]";
1224                                         }
1225                                 }
1226                         }
1227                 }
1228                 DI::cache()->set($cache_key, $text);
1229
1230                 return $text;
1231         }
1232
1233         public static function cleanPictureLinks($text)
1234         {
1235                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1236                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1237                 return $return;
1238         }
1239
1240         public static function removeLinks(string $bbcode)
1241         {
1242                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1243                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1244
1245                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1246                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1247                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1248                 return $bbcode;
1249         }
1250
1251         /**
1252          * Converts a BBCode message to HTML message
1253          *
1254          * BBcode 2 HTML was written by WAY2WEB.net
1255          * extended to work with Mistpark/Friendica - Mike Macgirvin
1256          *
1257          * Simple HTML values meaning:
1258          * - 0: Friendica display
1259          * - 1: Unused
1260          * - 2: Used for Windows Phone push, Friendica API
1261          * - 3: Used before converting to Markdown in bb2diaspora.php
1262          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1263          * - 5: Unused
1264          * - 6: Unused
1265          * - 7: Used for dfrn, OStatus
1266          * - 8: Used for Twitter, WP backlink text setting
1267          * - 9: ActivityPub
1268          *
1269          * @param string $text
1270          * @param bool   $try_oembed
1271          * @param int    $simple_html
1272          * @param bool   $for_plaintext
1273          * @return string
1274          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1275          */
1276         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1277         {
1278                 // Accounting for null default column values
1279                 if (is_null($text) || $text === '') {
1280                         return '';
1281                 }
1282
1283                 Hook::callAll('bbcode', $text);
1284
1285                 $a = DI::app();
1286
1287                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1288                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1289                                 /*
1290                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1291                                  *
1292                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1293                                  * $match[1] = $url
1294                                  * $match[2] = $title or absent
1295                                  */
1296                                 $try_oembed_callback = function ($match)
1297                                 {
1298                                         $url = $match[1];
1299                                         $title = $match[2] ?? null;
1300
1301                                         try {
1302                                                 $return = OEmbed::getHTML($url, $title);
1303                                         } catch (Exception $ex) {
1304                                                 $return = $match[0];
1305                                         }
1306
1307                                         return $return;
1308                                 };
1309
1310                                 // Remove the abstract element. It is a non visible element.
1311                                 $text = self::stripAbstract($text);
1312
1313                                 // Line ending normalisation
1314                                 $text = str_replace("\r\n", "\n", $text);
1315
1316                                 // Move new lines outside of tags
1317                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1318                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1319
1320                                 // Extract the private images which use data urls since preg has issues with
1321                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1322                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1323
1324                                 $extracted = self::extractImagesFromItemBody($text);
1325                                 $text = $extracted['body'];
1326                                 $saved_image = $extracted['images'];
1327
1328                                 // If we find any event code, turn it into an event.
1329                                 // After we're finished processing the bbcode we'll
1330                                 // replace all of the event code with a reformatted version.
1331
1332                                 $ev = Event::fromBBCode($text);
1333
1334                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1335                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1336
1337                                 $text = str_replace("<", "&lt;", $text);
1338                                 $text = str_replace(">", "&gt;", $text);
1339
1340                                 // remove some newlines before the general conversion
1341                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1342                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1343
1344                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1345                                 if (!$try_oembed) {
1346                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1347                                 }
1348
1349                                 // Remove linefeeds inside of the table elements. See issue #6799
1350                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1351                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1352                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1353                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1354                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1355                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1356                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1357                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1358                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1359                                         "[hr]", "[hr]", "[hr]", "[hr]",
1360                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1361                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1362                                 do {
1363                                         $oldtext = $text;
1364                                         $text = str_replace($search, $replace, $text);
1365                                 } while ($oldtext != $text);
1366
1367                                 // Replace these here only once
1368                                 $search = ["\n[table]", "[/table]\n"];
1369                                 $replace = ["[table]", "[/table]"];
1370                                 $text = str_replace($search, $replace, $text);
1371
1372                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1373                                 $text = trim($text, "\n");
1374
1375                                 // removing multiplicated newlines
1376                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1377                                         $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",
1378                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1379                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1380                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1381                                         do {
1382                                                 $oldtext = $text;
1383                                                 $text = str_replace($search, $replace, $text);
1384                                         } while ($oldtext != $text);
1385                                 }
1386
1387                                 // Add HTML new lines
1388                                 $text = str_replace("\n", '<br>', $text);
1389
1390                                 /// @todo Have a closer look at the different html modes
1391                                 // Handle attached links or videos
1392                                 if ($simple_html == self::ACTIVITYPUB) {
1393                                         $text = self::removeAttachment($text);
1394                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1395                                         $text = self::removeAttachment($text, true);
1396                                 } else {
1397                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1398                                 }
1399
1400                                 $nosmile = strpos($text, '[nosmile]') !== false;
1401                                 $text = str_replace('[nosmile]', '', $text);
1402
1403                                 // Replace non graphical smilies for external posts
1404                                 if (!$nosmile) {
1405                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1406                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1407                                         });
1408                                 }
1409
1410                                 // leave open the posibility of [map=something]
1411                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1412                                 if (strpos($text, '[/map]') !== false) {
1413                                         $text = preg_replace_callback(
1414                                                 "/\[map\](.*?)\[\/map\]/ism",
1415                                                 function ($match) use ($simple_html) {
1416                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1417                                                 },
1418                                                 $text
1419                                         );
1420                                 }
1421
1422                                 if (strpos($text, '[map=') !== false) {
1423                                         $text = preg_replace_callback(
1424                                                 "/\[map=(.*?)\]/ism",
1425                                                 function ($match) use ($simple_html) {
1426                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1427                                                 },
1428                                                 $text
1429                                         );
1430                                 }
1431
1432                                 if (strpos($text, '[map]') !== false) {
1433                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1434                                 }
1435
1436                                 // Check for headers
1437                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1438                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1439                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1440                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1441                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1442                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1443
1444                                 // Check for paragraph
1445                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1446
1447                                 // Check for bold text
1448                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1449
1450                                 // Check for Italics text
1451                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1452
1453                                 // Check for Underline text
1454                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1455
1456                                 // Check for strike-through text
1457                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1458
1459                                 // Check for over-line text
1460                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1461
1462                                 // Check for colored text
1463                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1464
1465                                 // Check for sized text
1466                                 // [size=50] --> font-size: 50px (with the unit).
1467                                 if ($simple_html != self::DIASPORA) {
1468                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1469                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1470                                 } else {
1471                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1472                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1473                                 }
1474
1475
1476                                 // Check for centered text
1477                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1478
1479                                 // Check for list text
1480                                 $text = str_replace("[*]", "<li>", $text);
1481
1482                                 // Check for style sheet commands
1483                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1484
1485                                 // Check for CSS classes
1486                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1487
1488                                 // handle nested lists
1489                                 $endlessloop = 0;
1490
1491                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1492                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1493                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1494                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1495                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1496                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1497                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1498                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1499                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1500                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1501                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1502                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1503                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1504                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1505                                 }
1506
1507                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1508                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1509                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1510                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1511
1512                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1513                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1514
1515                                 $text = str_replace('[hr]', '<hr />', $text);
1516
1517                                 if (!$for_plaintext) {
1518                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1519                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1520                                         });
1521                                 }
1522
1523                                 // Check for font change text
1524                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1525
1526                                 // Declare the format for [spoiler] layout
1527                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1528
1529                                 // Check for [spoiler] text
1530                                 // handle nested quotes
1531                                 $endlessloop = 0;
1532                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1533                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1534                                 }
1535
1536                                 // Check for [spoiler=Title] text
1537
1538                                 // handle nested quotes
1539                                 $endlessloop = 0;
1540                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1541                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1542                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1543                                                 $text);
1544                                 }
1545
1546                                 // Declare the format for [quote] layout
1547                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1548
1549                                 // Check for [quote] text
1550                                 // handle nested quotes
1551                                 $endlessloop = 0;
1552                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1553                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1554                                 }
1555
1556                                 // Check for [quote=Author] text
1557
1558                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1559
1560                                 // handle nested quotes
1561                                 $endlessloop = 0;
1562                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1563                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1564                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1565                                                 $text);
1566                                 }
1567
1568
1569                                 // [img=widthxheight]image source[/img]
1570                                 $text = preg_replace_callback(
1571                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1572                                         function ($matches) use ($simple_html) {
1573                                                 if (strpos($matches[3], "data:image/") === 0) {
1574                                                         return $matches[0];
1575                                                 }
1576
1577                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1578                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1579                                         },
1580                                         $text
1581                                 );
1582
1583                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1584                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1585
1586                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1587                                         function ($matches) use ($simple_html) {
1588                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1589                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1590                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1591                                         },
1592                                         $text);
1593
1594                                 // Images
1595                                 // [img]pathtoimage[/img]
1596                                 $text = preg_replace_callback(
1597                                         "/\[img\](.*?)\[\/img\]/ism",
1598                                         function ($matches) use ($simple_html) {
1599                                                 if (strpos($matches[1], "data:image/") === 0) {
1600                                                         return $matches[0];
1601                                                 }
1602
1603                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1604                                                 return "[img]" . $matches[1] . "[/img]";
1605                                         },
1606                                         $text
1607                                 );
1608
1609                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1610                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1611
1612                                 $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);
1613                                 $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>', $text);
1614                                 //$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);
1615
1616                                 // Simplify "video" element
1617                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1618
1619                                 if ($try_oembed) {
1620                                         // html5 video and audio
1621                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1622                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1623
1624                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1625                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1626
1627                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1628                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1629                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1630                                 } else {
1631                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1632                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1633                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1634                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1635                                 }
1636
1637                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1638                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1639
1640                                 // Youtube extensions
1641                                 if ($try_oembed) {
1642                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1643                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1644                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1645                                 }
1646
1647                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1648                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1649                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1650
1651                                 if ($try_oembed) {
1652                                         $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);
1653                                 } else {
1654                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1655                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1656                                 }
1657
1658                                 if ($try_oembed) {
1659                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1660                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1661                                 }
1662
1663                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1664                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1665
1666                                 if ($try_oembed) {
1667                                         $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);
1668                                 } else {
1669                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1670                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1671                                 }
1672
1673                                 // oembed tag
1674                                 $text = OEmbed::BBCode2HTML($text);
1675
1676                                 // Avoid triple linefeeds through oembed
1677                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1678
1679                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1680                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1681                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1682                                 // start which is always required). Allow desc with a missing summary for compatibility.
1683
1684                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1685                                         $sub = Event::getHTML($ev, $simple_html);
1686
1687                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1688                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1689                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1690                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1691                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1692                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1693                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1694                                 }
1695
1696                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1697                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1698                                         // Emojis are always 4 byte Unicode characters
1699                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1700                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1701                                         }
1702                                 }
1703
1704                                 if (!$for_plaintext) {
1705                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1706                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1707                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1708                                         }
1709                                 } else {
1710                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1711                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1712                                 }
1713
1714                                 // Remove all hashtag addresses
1715                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1716                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1717                                 } elseif ($simple_html == self::DIASPORA) {
1718                                         // The ! is converted to @ since Diaspora only understands the @
1719                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1720                                                 '@<a href="$2">$3</a>',
1721                                                 $text);
1722                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1723                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1724                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1725                                                 $text);
1726                                 } elseif (!$simple_html) {
1727                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1728                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1729                                                 $text);
1730                                 }
1731
1732                                 // Bookmarks in red - will be converted to bookmarks in friendica
1733                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1734                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1735                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1736                                                         "[bookmark=$1]$2[/bookmark]", $text);
1737
1738                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1739                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1740                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1741                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1742                                 }
1743
1744                                 // Perform URL Search
1745                                 if ($try_oembed) {
1746                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1747                                 }
1748
1749                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1750
1751                                 // Handle Diaspora posts
1752                                 $text = preg_replace_callback(
1753                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1754                                         function ($match) {
1755                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1756                                         }, $text
1757                                 );
1758
1759                                 $text = preg_replace_callback(
1760                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1761                                         function ($match) {
1762                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1763                                         }, $text
1764                                 );
1765
1766                                 // Server independent link to posts and comments
1767                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1768                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1769                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1770
1771                                 /* Tag conversion
1772                                  * Supports:
1773                                  * - #[url=<anything>]<term>[/url]
1774                                  * - [url=<anything>]#<term>[/url]
1775                                  */
1776                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1777                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1778                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1779                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1780                                                         . XML::escape($matches[1]) . '</a>';
1781                                         } else {
1782                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1783                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1784                                                         . XML::escape($matches[1]) . '</a>';
1785                                         }
1786                                 }, $text);
1787
1788                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1789                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1790                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1791                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1792                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1793
1794                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1795                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1796
1797                                 // Red compatibility, though the link can't be authenticated on Friendica
1798                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1799
1800
1801                                 // we may need to restrict this further if it picks up too many strays
1802                                 // link acct:user@host to a webfinger profile redirector
1803
1804                                 $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);
1805
1806                                 // Perform MAIL Search
1807                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1808                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1809
1810                                 /// @todo What is the meaning of these lines?
1811                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1812                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1813
1814                                 // Currently deactivated, it made problems with " inside of alt texts.
1815                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1816
1817                                 // fix any escaped ampersands that may have been converted into links
1818                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1819
1820                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1821                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1822
1823                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1824
1825                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1826                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1827
1828                                 // sanitize href attributes (only allowlisted protocols URLs)
1829                                 // default value for backward compatibility
1830                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1831
1832                                 // Always allowed protocol even if config isn't set or not including it
1833                                 $allowed_link_protocols[] = '//';
1834                                 $allowed_link_protocols[] = 'http://';
1835                                 $allowed_link_protocols[] = 'https://';
1836                                 $allowed_link_protocols[] = 'redir/';
1837
1838                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1839
1840                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1841                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1842
1843                                 // Shared content
1844                                 $text = self::convertShare(
1845                                         $text,
1846                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1847                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1848                                         }
1849                                 );
1850
1851                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1852
1853                                 return $text;
1854                         }); // Escaped noparse, nobb, pre
1855
1856                         // Remove escaping tags and replace new lines that remain
1857                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1858                                 return str_replace("\n", "<br>", $match[2]);
1859                         }, $text);
1860
1861                         // Additionally, [pre] tags preserve spaces
1862                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1863                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1864                         }, $text);
1865
1866                         return $text;
1867                 }); // Escaped code
1868
1869                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1870                         function ($matches) {
1871                                 if (strpos($matches[2], "\n") !== false) {
1872                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1873                                 } else {
1874                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1875                                 }
1876
1877                                 return $return;
1878                         },
1879                         $text
1880                 );
1881
1882                 // Default iframe allowed domains/path
1883                 $allowedIframeDomains = [
1884                         DI::baseUrl()->getHostname()
1885                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1886                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1887                         'www.youtube.com/embed/',
1888                         'player.vimeo.com/video/',
1889                 ];
1890
1891                 $allowedIframeDomains = array_merge(
1892                         $allowedIframeDomains,
1893                         DI::config()->get('system', 'allowed_oembed') ?
1894                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1895                                 : []
1896                 );
1897
1898                 $text = HTML::purify($text, $allowedIframeDomains);
1899
1900                 return trim($text);
1901         }
1902
1903         /**
1904          * Strips the "abstract" tag from the provided text
1905          *
1906          * @param string $text The text with BBCode
1907          * @return string The same text - but without "abstract" element
1908          */
1909         public static function stripAbstract($text)
1910         {
1911                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1912                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1913
1914                 return $text;
1915         }
1916
1917         /**
1918          * Returns the value of the "abstract" element
1919          *
1920          * @param string $text The text that maybe contains the element
1921          * @param string $addon The addon for which the abstract is meant for
1922          * @return string The abstract
1923          */
1924         public static function getAbstract($text, $addon = '')
1925         {
1926                 $abstract = '';
1927                 $abstracts = [];
1928                 $addon = strtolower($addon);
1929
1930                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1931                         foreach ($results AS $result) {
1932                                 $abstracts[strtolower($result[1])] = $result[2];
1933                         }
1934                 }
1935
1936                 if (isset($abstracts[$addon])) {
1937                         $abstract = $abstracts[$addon];
1938                 }
1939
1940                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1941                         $abstract = $result[1];
1942                 }
1943
1944                 return $abstract;
1945         }
1946
1947         /**
1948          * Callback function to replace a Friendica style mention in a mention for Diaspora
1949          *
1950          * @param array $match Matching values for the callback
1951          *                     [1] = Mention type (! or @)
1952          *                     [2] = Name
1953          *                     [3] = Address
1954          * @return string Replaced mention
1955          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1956          * @throws \ImagickException
1957          */
1958         private static function bbCodeMention2DiasporaCallback($match)
1959         {
1960                 $contact = Contact::getByURL($match[3], false, ['addr']);
1961                 if (empty($contact['addr'])) {
1962                         return $match[0];
1963                 }
1964
1965                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1966                 return $mention;
1967         }
1968
1969         /**
1970          * Converts a BBCode text into Markdown
1971          *
1972          * This function converts a BBCode item body to be sent to Markdown-enabled
1973          * systems like Diaspora and Libertree
1974          *
1975          * @param string $text
1976          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1977          * @return string
1978          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1979          */
1980         public static function toMarkdown($text, $for_diaspora = true)
1981         {
1982                 $original_text = $text;
1983
1984                 // Since Diaspora is creating a summary for links, this function removes them before posting
1985                 if ($for_diaspora) {
1986                         $text = self::removeShareInformation($text);
1987                 }
1988
1989                 /**
1990                  * Transform #tags, strip off the [url] and replace spaces with underscore
1991                  */
1992                 $url_search_string = "^\[\]";
1993                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1994                         function ($matches) {
1995                                 return '#' . str_replace(' ', '_', $matches[2]);
1996                         },
1997                         $text
1998                 );
1999
2000                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2001                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2002
2003                 // Convert it to HTML - don't try oembed
2004                 if ($for_diaspora) {
2005                         $text = self::convert($text, false, self::DIASPORA);
2006
2007                         // Add all tags that maybe were removed
2008                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2009                                 $tagline = '';
2010                                 foreach ($tags[2] as $tag) {
2011                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2012                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2013                                                 $tagline .= '#' . $tag . ' ';
2014                                         }
2015                                 }
2016                                 $text = $text . " " . $tagline;
2017                         }
2018                 } else {
2019                         $text = self::convert($text, false, self::CONNECTORS);
2020                 }
2021
2022                 // If a link is followed by a quote then there should be a newline before it
2023                 // Maybe we should make this newline at every time before a quote.
2024                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2025
2026                 $stamp1 = microtime(true);
2027
2028                 // Now convert HTML to Markdown
2029                 $text = HTML::toMarkdown($text);
2030
2031                 DI::profiler()->saveTimestamp($stamp1, "parser");
2032
2033                 // Libertree has a problem with escaped hashtags.
2034                 $text = str_replace(['\#'], ['#'], $text);
2035
2036                 // Remove any leading or trailing whitespace, as this will mess up
2037                 // the Diaspora signature verification and cause the item to disappear
2038                 $text = trim($text);
2039
2040                 if ($for_diaspora) {
2041                         $url_search_string = "^\[\]";
2042                         $text = preg_replace_callback(
2043                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2044                                 ['self', 'bbCodeMention2DiasporaCallback'],
2045                                 $text
2046                         );
2047                 }
2048
2049                 Hook::callAll('bb2diaspora', $text);
2050
2051                 return $text;
2052         }
2053
2054         /**
2055          * Pull out all #hashtags and @person tags from $string.
2056          *
2057          * We also get @person@domain.com - which would make
2058          * the regex quite complicated as tags can also
2059          * end a sentence. So we'll run through our results
2060          * and strip the period from any tags which end with one.
2061          * Returns array of tags found, or empty array.
2062          *
2063          * @param string $string Post content
2064          *
2065          * @return array List of tag and person names
2066          */
2067         public static function getTags($string)
2068         {
2069                 $ret = [];
2070
2071                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2072                         // Convert hashtag links to hashtags
2073                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2074
2075                         // Force line feeds at bbtags
2076                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2077
2078                         // ignore anything in a bbtag
2079                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2080
2081                         // Match full names against @tags including the space between first and last
2082                         // We will look these up afterward to see if they are full names or not recognisable.
2083
2084                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2085                                 foreach ($matches[1] as $match) {
2086                                         if (strstr($match, ']')) {
2087                                                 // we might be inside a bbcode color tag - leave it alone
2088                                                 continue;
2089                                         }
2090
2091                                         if (substr($match, -1, 1) === '.') {
2092                                                 $ret[] = substr($match, 0, -1);
2093                                         } else {
2094                                                 $ret[] = $match;
2095                                         }
2096                                 }
2097                         }
2098
2099                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2100                         // and #hash tags.
2101
2102                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2103                                 foreach ($matches[1] as $match) {
2104                                         if (strstr($match, ']')) {
2105                                                 // we might be inside a bbcode color tag - leave it alone
2106                                                 continue;
2107                                         }
2108
2109                                         // try not to catch url fragments
2110                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2111                                                 continue;
2112                                         }
2113
2114                                         $ret[] = $match;
2115                                 }
2116                         }
2117                 });
2118
2119                 return array_unique($ret);
2120         }
2121
2122         /**
2123          * Expand tags to URLs
2124          *
2125          * @param string $body
2126          * @return string body with expanded tags
2127          */
2128         public static function expandTags(string $body)
2129         {
2130                 return preg_replace_callback("/([!#@])([^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/",
2131                         function ($match) {
2132                                 switch ($match[1]) {
2133                                         case '!':
2134                                         case '@':
2135                                                 $contact = Contact::getByURL($match[2]);
2136                                                 if (!empty($contact)) {
2137                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2138                                                 } else {
2139                                                         return $match[1] . $match[2];
2140                                                 }
2141                                                 break;
2142                                         case '#':
2143                                                 return $match[1] . '[url=' . 'https://' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2144                                 }
2145                         }, $body);
2146         }
2147
2148         /**
2149          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2150          *
2151          * @param string   $text
2152          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2153          * @param callable $callback
2154          * @return string
2155          * @throws Exception
2156          *@see Strings::performWithEscapedBlocks
2157          *
2158          */
2159         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2160         {
2161                 $tagList = array_map('preg_quote', $tagList);
2162
2163                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2164         }
2165
2166         /**
2167          * Replaces mentions in the provided message body for the provided user and network if any
2168          *
2169          * @param $body
2170          * @param $profile_uid
2171          * @param $network
2172          * @return string
2173          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2174          * @throws \ImagickException
2175          */
2176         public static function setMentions($body, $profile_uid = 0, $network = '')
2177         {
2178                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2179                         $tags = BBCode::getTags($body);
2180
2181                         $tagged = [];
2182                         $inform = '';
2183
2184                         foreach ($tags as $tag) {
2185                                 $tag_type = substr($tag, 0, 1);
2186
2187                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2188                                         continue;
2189                                 }
2190
2191                                 /*
2192                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2193                                  * Robert Johnson should be first in the $tags array
2194                                  */
2195                                 foreach ($tagged as $nextTag) {
2196                                         if (stristr($nextTag, $tag . ' ')) {
2197                                                 continue 2;
2198                                         }
2199                                 }
2200
2201                                 if (($success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network)) && $success['replaced']) {
2202                                         $tagged[] = $tag;
2203                                 }
2204                         }
2205
2206                         return $body;
2207                 });
2208
2209                 return $body;
2210         }
2211
2212         /**
2213          * @param string      $author  Author display name
2214          * @param string      $profile Author profile URL
2215          * @param string      $avatar  Author profile picture URL
2216          * @param string      $link    Post source URL
2217          * @param string      $posted  Post created date
2218          * @param string|null $guid    Post guid (if any)
2219          * @return string
2220          * @TODO Rewrite to handle over whole record array
2221          */
2222         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2223         {
2224                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2225                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2226                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2227                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2228                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2229
2230                 if ($guid) {
2231                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2232                 }
2233
2234                 $header  .= "']";
2235
2236                 return $header;
2237         }
2238
2239         /**
2240          * Returns the BBCode relevant to embed the provided URL in a post body.
2241          * For media type, it will return [img], [video] and [audio] tags.
2242          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2243          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2244          *
2245          * @param string      $url
2246          * @param bool        $tryAttachment
2247          * @param string|null $title
2248          * @param string|null $description
2249          * @param string|null $tags
2250          * @return string
2251          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2252          *@see ParseUrl::getSiteinfoCached
2253          *
2254          */
2255         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2256         {
2257                 DI::logger()->info($url);
2258
2259                 // If there is already some content information submitted we don't
2260                 // need to parse the url for content.
2261                 if (!empty($title) && !empty($description)) {
2262                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2263
2264                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2265
2266                         $str_tags = '';
2267                         if (!empty($tags)) {
2268                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2269                                 if (count($arr_tags)) {
2270                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2271                                 }
2272                         }
2273
2274                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2275
2276                         DI::logger()->info('(unparsed): returns: ' . $result);
2277
2278                         return $result;
2279                 }
2280
2281                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2282
2283                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2284                         switch ($siteinfo['type']) {
2285                                 case 'video':
2286                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2287                                         break;
2288                                 case 'audio':
2289                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2290                                         break;
2291                                 default:
2292                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2293                                         break;
2294                         }
2295
2296                         return $bbcode;
2297                 }
2298
2299                 unset($siteinfo['keywords']);
2300
2301                 // Bypass attachment if parse url for a comment
2302                 if (!$tryAttachment) {
2303                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2304                 }
2305
2306                 // Format it as BBCode attachment
2307                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2308         }
2309 }