src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\Hook;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Logger;
  19 use Friendica\Core\Protocol;
  20 use Friendica\Core\Renderer;
  21 use Friendica\Core\System;
  22 use Friendica\DI;
  23 use Friendica\Model\Contact;
  24 use Friendica\Model\Event;
  25 use Friendica\Model\Photo;
  26 use Friendica\Network\Probe;
  27 use Friendica\Object\Image;
  28 use Friendica\Protocol\Activity;
  29 use Friendica\Util\Images;
  30 use Friendica\Util\Map;
  31 use Friendica\Util\Network;
  32 use Friendica\Util\ParseUrl;
  33 use Friendica\Util\Proxy as ProxyUtils;
  34 use Friendica\Util\Strings;
  35 use Friendica\Util\XML;
  36
  37 class BBCode extends BaseObject
  38 {
  39         /**
  40          * @brief Fetches attachment data that were generated the old way
  41          *
  42          * @param string $body Message body
  43          * @return array
  44          *                     'type' -> Message type ('link', 'video', 'photo')
  45          *                     'text' -> Text before the shared message
  46          *                     'after' -> Text after the shared message
  47          *                     'image' -> Preview image of the message
  48          *                     'url' -> Url to the attached message
  49          *                     'title' -> Title of the attachment
  50          *                     'description' -> Description of the attachment
  51          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  52          */
  53         private static function getOldAttachmentData($body)
  54         {
  55                 $post = [];
  56
  57                 // Simplify image codes
  58                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  59
  60                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  61                         foreach ($attached as $data) {
  62                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  63                                         continue;
  64                                 }
  65
  66                                 $post['type'] = substr($data[1], 5);
  67
  68                                 $pos = strpos($body, $data[0]);
  69                                 if ($pos > 0) {
  70                                         $post['text'] = trim(substr($body, 0, $pos));
  71                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  72                                 } else {
  73                                         $post['text'] = trim(str_replace($data[0], '', $body));
  74                                         $post['after'] = '';
  75                                 }
  76
  77                                 $attacheddata = $data[2];
  78
  79                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
  80
  81                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
  82
  83                                         if ($picturedata) {
  84                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  85                                                         $post['image'] = $matches[1];
  86                                                 } else {
  87                                                         $post['preview'] = $matches[1];
  88                                                 }
  89                                         }
  90                                 }
  91
  92                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  93                                         $post['url'] = $matches[1];
  94                                         $post['title'] = $matches[2];
  95                                 }
  96                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
  97                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  98                                         $post['url'] = $matches[1];
  99                                 }
 100
 101                                 // Search for description
 102                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 103                                         $post['description'] = $matches[1];
 104                                 }
 105                         }
 106                 }
 107                 return $post;
 108         }
 109
 110         /**
 111          * @brief Fetches attachment data that were generated with the "attachment" element
 112          *
 113          * @param string $body Message body
 114          * @return array
 115          *                     'type' -> Message type ('link', 'video', 'photo')
 116          *                     'text' -> Text before the shared message
 117          *                     'after' -> Text after the shared message
 118          *                     'image' -> Preview image of the message
 119          *                     'url' -> Url to the attached message
 120          *                     'title' -> Title of the attachment
 121          *                     'description' -> Description of the attachment
 122          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 123          */
 124         public static function getAttachmentData($body)
 125         {
 126                 $data = [
 127                         'type'        => '',
 128                         'text'        => '',
 129                         'after'       => '',
 130                         'image'       => null,
 131                         'url'         => '',
 132                         'title'       => '',
 133                         'description' => '',
 134                 ];
 135
 136                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 137                         return self::getOldAttachmentData($body);
 138                 }
 139
 140                 $attributes = $match[2];
 141
 142                 $data['text'] = trim($match[1]);
 143
 144                 $type = '';
 145                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 146                 if (!empty($matches[1])) {
 147                         $type = strtolower($matches[1]);
 148                 }
 149
 150                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 151                 if (!empty($matches[1])) {
 152                         $type = strtolower($matches[1]);
 153                 }
 154
 155                 if ($type == '') {
 156                         return [];
 157                 }
 158
 159                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 160                         return [];
 161                 }
 162
 163                 if ($type != '') {
 164                         $data['type'] = $type;
 165                 }
 166
 167                 $url = '';
 168                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 169                 if (!empty($matches[1])) {
 170                         $url = $matches[1];
 171                 }
 172
 173                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 174                 if (!empty($matches[1])) {
 175                         $url = $matches[1];
 176                 }
 177
 178                 if ($url != '') {
 179                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 180                 }
 181
 182                 $title = '';
 183                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 184                 if (!empty($matches[1])) {
 185                         $title = $matches[1];
 186                 }
 187
 188                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 189                 if (!empty($matches[1])) {
 190                         $title = $matches[1];
 191                 }
 192
 193                 if ($title != '') {
 194                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 195                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 196                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 197                         $data['title'] = $title;
 198                 }
 199
 200                 $image = '';
 201                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 202                 if (!empty($matches[1])) {
 203                         $image = $matches[1];
 204                 }
 205
 206                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 207                 if (!empty($matches[1])) {
 208                         $image = $matches[1];
 209                 }
 210
 211                 if ($image != '') {
 212                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 213                 }
 214
 215                 $preview = '';
 216                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 217                 if (!empty($matches[1])) {
 218                         $preview = $matches[1];
 219                 }
 220
 221                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 222                 if (!empty($matches[1])) {
 223                         $preview = $matches[1];
 224                 }
 225
 226                 if ($preview != '') {
 227                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 228                 }
 229
 230                 $data['description'] = trim($match[3]);
 231
 232                 $data['after'] = trim($match[4]);
 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' => $picture[2]];
 258                                 }
 259                         }
 260                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 261                                 $post['image_description'] = $post['images'][0]['description'];
 262                         }
 263                 }
 264
 265                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 266                         foreach ($pictures as $picture) {
 267                                 if (Photo::isLocal($picture[1])) {
 268                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 269                                 }
 270                         }
 271                 }
 272
 273                 // if nothing is found, it maybe having an image.
 274                 if (!isset($post['type'])) {
 275                         // Simplify image codes
 276                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 277                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 278                         $post['text'] = $body;
 279
 280                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 281                                 if ((count($pictures) == 1) && !$has_title) {
 282                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 283                                                 // Replace the preview picture with the real picture
 284                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 285                                                 $data = ['url' => $url, 'type' => 'photo'];
 286                                         } else {
 287                                                 // Checking, if the link goes to a picture
 288                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 289                                         }
 290
 291                                         // Workaround:
 292                                         // Sometimes photo posts to the own album are not detected at the start.
 293                                         // So we seem to cannot use the cache for these cases. That's strange.
 294                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 295                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 296                                         }
 297
 298                                         if ($data['type'] == 'photo') {
 299                                                 $post['type'] = 'photo';
 300                                                 if (isset($data['images'][0])) {
 301                                                         $post['image'] = $data['images'][0]['src'];
 302                                                         $post['url'] = $data['url'];
 303                                                 } else {
 304                                                         $post['image'] = $data['url'];
 305                                                 }
 306
 307                                                 $post['preview'] = $pictures[0][2];
 308                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 309                                         } else {
 310                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 311                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 312                                                         $post['type'] = 'photo';
 313                                                         $post['image'] = $pictures[0][1];
 314                                                         $post['preview'] = $pictures[0][2];
 315                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 316                                                 }
 317                                         }
 318                                 } elseif (count($pictures) > 0) {
 319                                         $post['type'] = 'link';
 320                                         $post['url'] = $plink;
 321                                         $post['image'] = $pictures[0][2];
 322                                         $post['text'] = $body;
 323
 324                                         foreach ($pictures as $picture) {
 325                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 326                                         }
 327                                 }
 328                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 329                                 if ((count($pictures) == 1) && !$has_title) {
 330                                         $post['type'] = 'photo';
 331                                         $post['image'] = $pictures[0][1];
 332                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 333                                 } elseif (count($pictures) > 0) {
 334                                         $post['type'] = 'link';
 335                                         $post['url'] = $plink;
 336                                         $post['image'] = $pictures[0][1];
 337                                         $post['text'] = $body;
 338
 339                                         foreach ($pictures as $picture) {
 340                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 341                                         }
 342                                 }
 343                         }
 344
 345                         // Test for the external links
 346                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 347                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 348
 349                         $links = array_merge($links1, $links2);
 350
 351                         // If there is only a single one, then use it.
 352                         // This should cover link posts via API.
 353                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 354                                 $post['type'] = 'link';
 355                                 $post['url'] = $links[0][1];
 356                         }
 357
 358                         // Simplify "video" element
 359                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 360
 361                         // Now count the number of external media links
 362                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 363                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 364                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 365                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 366
 367                         // Add them to the other external links
 368                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 369
 370                         // Are there more than one?
 371                         if (count($links) > 1) {
 372                                 // The post will be the type "text", which means a blog post
 373                                 unset($post['type']);
 374                                 $post['url'] = $plink;
 375                         }
 376
 377                         if (!isset($post['type'])) {
 378                                 $post['type'] = "text";
 379                                 $post['text'] = trim($body);
 380                         }
 381                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 382                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 383
 384                         if (isset($data['images'][0])) {
 385                                 $post['image'] = $data['images'][0]['src'];
 386                         }
 387                 }
 388
 389                 return $post;
 390         }
 391
 392         /**
 393          * Remove [attachment] BBCode and replaces it with a regular [url]
 394          *
 395          * @param string  $body
 396          * @param boolean $no_link_desc No link description
 397          *
 398          * @return string with replaced body
 399          */
 400         public static function removeAttachment($body, $no_link_desc = false)
 401         {
 402                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 403                         function ($match) use ($no_link_desc) {
 404                                 $attach_data = self::getAttachmentData($match[0]);
 405                                 if (empty($attach_data['url'])) {
 406                                         return $match[0];
 407                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 408                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 409                                 } else {
 410                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 411                                 }
 412                 }, $body);
 413         }
 414
 415         /**
 416          * @brief Converts a BBCode text into plaintext
 417          *
 418          * @param      $text
 419          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 420          *
 421          * @return string
 422          */
 423         public static function toPlaintext($text, $keep_urls = true)
 424         {
 425                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 426
 427                 return $naked_text;
 428         }
 429
 430         private static function proxyUrl($image, $simplehtml = false)
 431         {
 432                 // Only send proxied pictures to API and for internal display
 433                 if (in_array($simplehtml, [false, 2])) {
 434                         return ProxyUtils::proxifyUrl($image);
 435                 } else {
 436                         return $image;
 437                 }
 438         }
 439
 440         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 441         {
 442                 // Suppress "view full size"
 443                 if (intval(Config::get('system', 'no_view_full_size'))) {
 444                         $include_link = false;
 445                 }
 446
 447                 // Picture addresses can contain special characters
 448                 $s = htmlspecialchars_decode($srctext);
 449
 450                 $matches = null;
 451                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 452                 if ($c) {
 453                         foreach ($matches as $mtch) {
 454                                 Logger::log('scale_external_image: ' . $mtch[1]);
 455
 456                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 457                                 if (stristr($mtch[1], $hostname)) {
 458                                         continue;
 459                                 }
 460
 461                                 // $scale_replace, if passed, is an array of two elements. The
 462                                 // first is the name of the full-size image. The second is the
 463                                 // name of a remote, scaled-down version of the full size image.
 464                                 // This allows Friendica to display the smaller remote image if
 465                                 // one exists, while still linking to the full-size image
 466                                 if ($scale_replace) {
 467                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 468                                 } else {
 469                                         $scaled = $mtch[1];
 470                                 }
 471                                 $i = Network::fetchUrl($scaled);
 472                                 if (!$i) {
 473                                         return $srctext;
 474                                 }
 475
 476                                 // guess mimetype from headers or filename
 477                                 $type = Images::guessType($mtch[1], true);
 478
 479                                 if ($i) {
 480                                         $Image = new Image($i, $type);
 481                                         if ($Image->isValid()) {
 482                                                 $orig_width = $Image->getWidth();
 483                                                 $orig_height = $Image->getHeight();
 484
 485                                                 if ($orig_width > 640 || $orig_height > 640) {
 486                                                         $Image->scaleDown(640);
 487                                                         $new_width = $Image->getWidth();
 488                                                         $new_height = $Image->getHeight();
 489                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 490                                                         $s = str_replace(
 491                                                                 $mtch[0],
 492                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 493                                                                 . "\n" . (($include_link)
 494                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 495                                                                         : ''),
 496                                                                 $s
 497                                                         );
 498                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 499                                                 }
 500                                         }
 501                                 }
 502                         }
 503                 }
 504
 505                 // replace the special char encoding
 506                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 507                 return $s;
 508         }
 509
 510         /**
 511          * The purpose of this function is to apply system message length limits to
 512          * imported messages without including any embedded photos in the length
 513          *
 514          * @brief Truncates imported message body string length to max_import_size
 515          * @param string $body
 516          * @return string
 517          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 518          */
 519         public static function limitBodySize($body)
 520         {
 521                 $maxlen = Config::get('config', 'max_import_size', 0);
 522
 523                 // If the length of the body, including the embedded images, is smaller
 524                 // than the maximum, then don't waste time looking for the images
 525                 if ($maxlen && (strlen($body) > $maxlen)) {
 526
 527                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 528
 529                         $orig_body = $body;
 530                         $new_body = '';
 531                         $textlen = 0;
 532
 533                         $img_start = strpos($orig_body, '[img');
 534                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 535                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 536                         while (($img_st_close !== false) && ($img_end !== false)) {
 537
 538                                 $img_st_close++; // make it point to AFTER the closing bracket
 539                                 $img_end += $img_start;
 540                                 $img_end += strlen('[/img]');
 541
 542                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 543                                         // This is an embedded image
 544
 545                                         if (($textlen + $img_start) > $maxlen) {
 546                                                 if ($textlen < $maxlen) {
 547                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 548                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 549                                                         $textlen = $maxlen;
 550                                                 }
 551                                         } else {
 552                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 553                                                 $textlen += $img_start;
 554                                         }
 555
 556                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 557                                 } else {
 558
 559                                         if (($textlen + $img_end) > $maxlen) {
 560                                                 if ($textlen < $maxlen) {
 561                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 562                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 563                                                         $textlen = $maxlen;
 564                                                 }
 565                                         } else {
 566                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 567                                                 $textlen += $img_end;
 568                                         }
 569                                 }
 570                                 $orig_body = substr($orig_body, $img_end);
 571
 572                                 if ($orig_body === false) {
 573                                         // in case the body ends on a closing image tag
 574                                         $orig_body = '';
 575                                 }
 576
 577                                 $img_start = strpos($orig_body, '[img');
 578                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 579                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 580                         }
 581
 582                         if (($textlen + strlen($orig_body)) > $maxlen) {
 583                                 if ($textlen < $maxlen) {
 584                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 585                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 586                                 }
 587                         } else {
 588                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 589                                 $new_body = $new_body . $orig_body;
 590                         }
 591
 592                         return $new_body;
 593                 } else {
 594                         return $body;
 595                 }
 596         }
 597
 598         /**
 599          * Processes [attachment] tags
 600          *
 601          * Note: Can produce a [bookmark] tag in the returned string
 602          *
 603          * @brief Processes [attachment] tags
 604          * @param string   $text
 605          * @param bool|int $simplehtml
 606          * @param bool     $tryoembed
 607          * @return string
 608          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 609          */
 610         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 611         {
 612                 $data = self::getAttachmentData($text);
 613                 if (empty($data) || empty($data['url'])) {
 614                         return $text;
 615                 }
 616
 617                 if (isset($data['title'])) {
 618                         $data['title'] = strip_tags($data['title']);
 619                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 620                 } else {
 621                         $data['title'] = null;
 622                 }
 623
 624                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data['image'])) {
 625                         $data['preview'] = $data['image'];
 626                         $data['image'] = '';
 627                 }
 628
 629                 $return = '';
 630                 try {
 631                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 632                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 633                         } else {
 634                                 throw new Exception('OEmbed is disabled for this attachment.');
 635                         }
 636                 } catch (Exception $e) {
 637                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 638
 639                         if ($simplehtml != 4) {
 640                                 $return = sprintf('<div class="type-%s">', $data['type']);
 641                         }
 642
 643                         if (!empty($data['title']) && !empty($data['url'])) {
 644                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 645                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 646                                 } else {
 647                                         if (!empty($data['image'])) {
 648                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 649                                         } elseif (!empty($data['preview'])) {
 650                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']);
 651                                         }
 652                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 653                                 }
 654                         }
 655
 656                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 657                                 // Sanitize the HTML by converting it to BBCode
 658                                 $bbcode = HTML::toBBCode($data['description']);
 659                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 660                         }
 661
 662                         if (!empty($data['url'])) {
 663                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 664                         }
 665
 666                         if ($simplehtml != 4) {
 667                                 $return .= '</div>';
 668                         }
 669                 }
 670
 671                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 672         }
 673
 674         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 675         {
 676                 $data = self::getAttachmentData($Text);
 677
 678                 if (!$data) {
 679                         return $Text;
 680                 } elseif ($nolink) {
 681                         return $data['text'] . ($data['after'] ?? '');
 682                 }
 683
 684                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 685                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 686                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 687                         $data['title'] = $data['url'];
 688                 } elseif (($text != '') && strstr($title, $text)) {
 689                         $data['text'] = $data['title'];
 690                         $data['title'] = $data['url'];
 691                 }
 692
 693                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 694                         return $data['title'] . $data['after'];
 695                 }
 696
 697                 // If the link already is included in the post, don't add it again
 698                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 699                         return $data['text'] . $data['after'];
 700                 }
 701
 702                 $text = $data['text'];
 703
 704                 if (!empty($data['url']) && !empty($data['title'])) {
 705                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 706                 } elseif (!empty($data['url'])) {
 707                         $text .= "\n[url]" . $data['url'] . '[/url]';
 708                 }
 709
 710                 return $text . "\n" . $data['after'];
 711         }
 712
 713         /**
 714          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 715          *
 716          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 717          * @param array $match Array with the matching values
 718          * @return string reformatted link including HTML codes
 719          */
 720         private static function convertUrlForActivityPubCallback($match)
 721         {
 722                 $url = $match[1];
 723
 724                 if (isset($match[2]) && ($match[1] != $match[2])) {
 725                         return $match[0];
 726                 }
 727
 728                 $parts = parse_url($url);
 729                 if (!isset($parts['scheme'])) {
 730                         return $match[0];
 731                 }
 732
 733                 return self::convertUrlForActivityPub($url);
 734         }
 735
 736         /**
 737          * @brief Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 738          * @param string $url URL that is about to be reformatted
 739          * @return string reformatted link including HTML codes
 740          */
 741         private static function convertUrlForActivityPub($url)
 742         {
 743                 $html = '<a href="%s" target="_blank">%s</a>';
 744                 return sprintf($html, $url, self::getStyledURL($url));
 745         }
 746
 747         /**
 748          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 749          * @param string $url URL that is about to be reformatted
 750          * @return string reformatted link
 751          */
 752         private static function getStyledURL($url)
 753         {
 754                 $parts = parse_url($url);
 755                 $scheme = $parts['scheme'] . '://';
 756                 $styled_url = str_replace($scheme, '', $url);
 757
 758                 if (strlen($styled_url) > 30) {
 759                         $styled_url = substr($styled_url, 0, 30) . "…";
 760                 }
 761
 762                 return $styled_url;
 763         }
 764
 765         /*
 766          * [noparse][i]italic[/i][/noparse] turns into
 767          * [noparse][ i ]italic[ /i ][/noparse],
 768          * to hide them from parser.
 769          */
 770         private static function escapeNoparseCallback($match)
 771         {
 772                 $whole_match = $match[0];
 773                 $captured = $match[1];
 774                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 775                 $new_str = str_replace($captured, $spacefied, $whole_match);
 776                 return $new_str;
 777         }
 778
 779         /*
 780          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 781          * now turns back and the [noparse] tags are trimed
 782          * returning [i]italic[/i]
 783          */
 784         private static function unescapeNoparseCallback($match)
 785         {
 786                 $captured = $match[1];
 787                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 788                 return $unspacefied;
 789         }
 790
 791         /**
 792          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 793          * occurrences
 794          *
 795          * @param string $text        Text to search
 796          * @param string $name        Tag name
 797          * @param int    $occurrences Number of first occurrences to skip
 798          * @return boolean|array
 799          */
 800         public static function getTagPosition($text, $name, $occurrences = 0)
 801         {
 802                 if ($occurrences < 0) {
 803                         $occurrences = 0;
 804                 }
 805
 806                 $start_open = -1;
 807                 for ($i = 0; $i <= $occurrences; $i++) {
 808                         if ($start_open !== false) {
 809                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 810                         }
 811                 }
 812
 813                 if ($start_open === false) {
 814                         return false;
 815                 }
 816
 817                 $start_equal = strpos($text, '=', $start_open);
 818                 $start_close = strpos($text, ']', $start_open);
 819
 820                 if ($start_close === false) {
 821                         return false;
 822                 }
 823
 824                 $start_close++;
 825
 826                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 827
 828                 if ($end_open === false) {
 829                         return false;
 830                 }
 831
 832                 $res = [
 833                         'start' => [
 834                                 'open' => $start_open,
 835                                 'close' => $start_close
 836                         ],
 837                         'end' => [
 838                                 'open' => $end_open,
 839                                 'close' => $end_open + strlen('[/' . $name . ']')
 840                         ],
 841                 ];
 842
 843                 if ($start_equal !== false) {
 844                         $res['start']['equal'] = $start_equal + 1;
 845                 }
 846
 847                 return $res;
 848         }
 849
 850         /**
 851          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 852          *
 853          * @param string $pattern Preg pattern string
 854          * @param string $replace Preg replace string
 855          * @param string $name    BBCode tag name
 856          * @param string $text    Text to search
 857          * @return string
 858          */
 859         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 860         {
 861                 $occurrences = 0;
 862                 $pos = self::getTagPosition($text, $name, $occurrences);
 863                 while ($pos !== false && $occurrences++ < 1000) {
 864                         $start = substr($text, 0, $pos['start']['open']);
 865                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 866                         $end = substr($text, $pos['end']['close']);
 867                         if ($end === false) {
 868                                 $end = '';
 869                         }
 870
 871                         $subject = preg_replace($pattern, $replace, $subject);
 872                         $text = $start . $subject . $end;
 873
 874                         $pos = self::getTagPosition($text, $name, $occurrences);
 875                 }
 876
 877                 return $text;
 878         }
 879
 880         private static function extractImagesFromItemBody($body)
 881         {
 882                 $saved_image = [];
 883                 $orig_body = $body;
 884                 $new_body = '';
 885
 886                 $cnt = 0;
 887                 $img_start = strpos($orig_body, '[img');
 888                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 889                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 890                 while (($img_st_close !== false) && ($img_end !== false)) {
 891                         $img_st_close++; // make it point to AFTER the closing bracket
 892                         $img_end += $img_start;
 893
 894                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 895                                 // This is an embedded image
 896                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 897                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 898
 899                                 $cnt++;
 900                         } else {
 901                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 902                         }
 903
 904                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 905
 906                         if ($orig_body === false) {
 907                                 // in case the body ends on a closing image tag
 908                                 $orig_body = '';
 909                         }
 910
 911                         $img_start = strpos($orig_body, '[img');
 912                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 913                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 914                 }
 915
 916                 $new_body = $new_body . $orig_body;
 917
 918                 return ['body' => $new_body, 'images' => $saved_image];
 919         }
 920
 921         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 922         {
 923                 $newbody = $body;
 924
 925                 $cnt = 0;
 926                 foreach ($images as $image) {
 927                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 928                         // it loops over the array starting from the first element and going sequentially
 929                         // to the last element
 930                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 931                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 932                         $cnt++;
 933                 }
 934
 935                 return $newbody;
 936         }
 937
 938         /**
 939          * This function converts a [share] block to text according to a provided callback function whose signature is:
 940          *
 941          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 942          *
 943          * Where:
 944          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 945          * data lookup
 946          * - $author_contact is a contact record array
 947          * - $content is the inner content of the [share] block
 948          * - $is_quote_share indicates whether there's any content before the [share] block
 949          * - Return value is the string that should replace the [share] block in the provided text
 950          *
 951          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 952          *
 953          * @param  string   $text     A BBCode string
 954          * @param  callable $callback
 955          * @return string The BBCode string with all [share] blocks replaced
 956          */
 957         public static function convertShare($text, callable $callback)
 958         {
 959                 $return = preg_replace_callback(
 960                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 961                         function ($match) use ($callback) {
 962                                 $attribute_string = $match[2];
 963                                 $attributes = [];
 964                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 965                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 966                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 967                                 }
 968
 969                                 // We only call this so that a previously unknown contact can be added.
 970                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 971                                 // This function then can fetch an entry from the contact table.
 972                                 $default['url'] = $attributes['profile'];
 973
 974                                 if (!empty($attributes['author'])) {
 975                                         $default['name'] = $attributes['author'];
 976                                 }
 977
 978                                 if (!empty($attributes['avatar'])) {
 979                                         $default['photo'] = $attributes['avatar'];
 980                                 }
 981
 982                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 983
 984                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 985                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 986
 987                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 988                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 989                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 990
 991                                 if ($attributes['avatar']) {
 992                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 993                                 }
 994
 995                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 996                         },
 997                         $text
 998                 );
 999
1000                 return $return;
1001         }
1002
1003         /**
1004          * Default [share] tag conversion callback
1005          *
1006          * Note: Can produce a [bookmark] tag in the output
1007          *
1008          * @see BBCode::convertShare()
1009          * @param array   $attributes     [share] block attribute values
1010          * @param array   $author_contact Contact row of the shared author
1011          * @param string  $content        Inner content of the [share] block
1012          * @param boolean $is_quote_share Whether there is content before the [share] block
1013          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1014          * @return string
1015          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1016          */
1017         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1018         {
1019                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1020
1021                 switch ($simplehtml) {
1022                         case 1:
1023                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1024                                 break;
1025                         case 2:
1026                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1027                                 break;
1028                         case 3: // Diaspora
1029                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1030                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1031                                 } else {
1032                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1033
1034                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1035                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1036                                         }
1037
1038                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1039
1040                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1041                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1042                                         }
1043                                 }
1044
1045                                 break;
1046                         case 4:
1047                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1048                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1049                                 $headline .= ':</b></p>' . "\n";
1050
1051                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1052
1053                                 break;
1054                         case 5:
1055                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1056                                 break;
1057                         case 7: // statusnet/GNU Social
1058                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1059                                 break;
1060                         case 9: // ActivityPub
1061                                 $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>';
1062                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1063                                 break;
1064                         default:
1065                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1066                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1067                                         try {
1068                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1069                                         } catch (Exception $e) {
1070                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1071                                         }
1072                                 } else {
1073                                         $text = ($is_quote_share? "\n" : '');
1074
1075                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1076                                         $text .= Renderer::replaceMacros($tpl, [
1077                                                 '$profile' => $attributes['profile'],
1078                                                 '$avatar'  => $attributes['avatar'],
1079                                                 '$author'  => $attributes['author'],
1080                                                 '$link'    => $attributes['link'],
1081                                                 '$posted'  => $attributes['posted'],
1082                                                 '$content' => trim($content)
1083                                         ]);
1084                                 }
1085                                 break;
1086                 }
1087
1088                 return $text;
1089         }
1090
1091         private static function removePictureLinksCallback($match)
1092         {
1093                 $cache_key = 'remove:' . $match[1];
1094                 $text = Cache::get($cache_key);
1095
1096                 if (is_null($text)) {
1097                         $a = DI::app();
1098
1099                         $stamp1 = microtime(true);
1100
1101                         $ch = @curl_init($match[1]);
1102                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1103                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1104                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1105                         @curl_exec($ch);
1106                         $curl_info = @curl_getinfo($ch);
1107
1108                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1109
1110                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1111                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1112                         } else {
1113                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1114
1115                                 // if its not a picture then look if its a page that contains a picture link
1116                                 $body = Network::fetchUrl($match[1]);
1117
1118                                 $doc = new DOMDocument();
1119                                 @$doc->loadHTML($body);
1120                                 $xpath = new DOMXPath($doc);
1121                                 $list = $xpath->query("//meta[@name]");
1122                                 foreach ($list as $node) {
1123                                         $attr = [];
1124
1125                                         if ($node->attributes->length) {
1126                                                 foreach ($node->attributes as $attribute) {
1127                                                         $attr[$attribute->name] = $attribute->value;
1128                                                 }
1129                                         }
1130
1131                                         if (strtolower($attr['name']) == 'twitter:image') {
1132                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1133                                         }
1134                                 }
1135                         }
1136                         Cache::set($cache_key, $text);
1137                 }
1138
1139                 return $text;
1140         }
1141
1142         private static function expandLinksCallback($match)
1143         {
1144                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1145                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1146                 } else {
1147                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1148                 }
1149         }
1150
1151         private static function cleanPictureLinksCallback($match)
1152         {
1153                 $a = DI::app();
1154
1155                 // When the picture link is the own photo path then we can avoid fetching the link
1156                 $own_photo_url = preg_quote(Strings::normaliseLink($a->getBaseURL()) . '/photos/');
1157                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1158                         if (!empty($match[3])) {
1159                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1160                         } else {
1161                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1162                         }
1163                         return $text;
1164                 }
1165
1166                 $cache_key = 'clean:' . $match[1];
1167                 $text = Cache::get($cache_key);
1168                 if (!is_null($text)) {
1169                         return $text;
1170                 }
1171
1172                 // Only fetch the header, not the content
1173                 $stamp1 = microtime(true);
1174
1175                 $ch = @curl_init($match[1]);
1176                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1177                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1178                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1179                 @curl_exec($ch);
1180                 $curl_info = @curl_getinfo($ch);
1181
1182                 $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1183
1184                 // if its a link to a picture then embed this picture
1185                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1186                         $text = '[img]' . $match[1] . '[/img]';
1187                 } else {
1188                         if (!empty($match[3])) {
1189                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1190                         } else {
1191                                 $text = '[img]' . $match[2] . '[/img]';
1192                         }
1193
1194                         // if its not a picture then look if its a page that contains a picture link
1195                         $body = Network::fetchUrl($match[1]);
1196
1197                         $doc = new DOMDocument();
1198                         @$doc->loadHTML($body);
1199                         $xpath = new DOMXPath($doc);
1200                         $list = $xpath->query("//meta[@name]");
1201                         foreach ($list as $node) {
1202                                 $attr = [];
1203                                 if ($node->attributes->length) {
1204                                         foreach ($node->attributes as $attribute) {
1205                                                 $attr[$attribute->name] = $attribute->value;
1206                                         }
1207                                 }
1208
1209                                 if (strtolower($attr['name']) == "twitter:image") {
1210                                         if (!empty($match[3])) {
1211                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1212                                         } else {
1213                                                 $text = "[img]" . $attr['content'] . "[/img]";
1214                                         }
1215                                 }
1216                         }
1217                 }
1218                 Cache::set($cache_key, $text);
1219
1220                 return $text;
1221         }
1222
1223         public static function cleanPictureLinks($text)
1224         {
1225                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1226                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1227                 return $return;
1228         }
1229
1230         /**
1231          * @brief Converts a BBCode message to HTML message
1232          *
1233          * BBcode 2 HTML was written by WAY2WEB.net
1234          * extended to work with Mistpark/Friendica - Mike Macgirvin
1235          *
1236          * Simple HTML values meaning:
1237          * - 0: Friendica display
1238          * - 1: Unused
1239          * - 2: Used for Windows Phone push, Friendica API
1240          * - 3: Used before converting to Markdown in bb2diaspora.php
1241          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1242          * - 5: Unused
1243          * - 6: Unused
1244          * - 7: Used for dfrn, OStatus
1245          * - 8: Used for Twitter, WP backlink text setting
1246          * - 9: ActivityPub
1247          *
1248          * @param string $text
1249          * @param bool   $try_oembed
1250          * @param int    $simple_html
1251          * @param bool   $for_plaintext
1252          * @return string
1253          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1254          */
1255         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1256         {
1257                 $a = DI::app();
1258
1259                 /*
1260                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1261                  *
1262                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1263                  * $match[1] = $url
1264                  * $match[2] = $title or absent
1265                  */
1266                 $try_oembed_callback = function ($match)
1267                 {
1268                         $url = $match[1];
1269                         $title = $match[2] ?? null;
1270
1271                         try {
1272                                 $return = OEmbed::getHTML($url, $title);
1273                         } catch (Exception $ex) {
1274                                 $return = $match[0];
1275                         }
1276
1277                         return $return;
1278                 };
1279
1280                 // Extracting code blocks before the whitespace processing and the autolinker
1281                 $codeblocks = [];
1282
1283                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1284                         function ($matches) use (&$codeblocks) {
1285                                 $return = '#codeblock-' . count($codeblocks) . '#';
1286                                 if (strpos($matches[2], "\n") !== false) {
1287                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1288                                 } else {
1289                                         $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1290                                 }
1291
1292                                 return $return;
1293                         },
1294                         $text
1295                 );
1296
1297                 // Hide all [noparse] contained bbtags by spacefying them
1298                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1299
1300                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1301                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1302                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1303
1304                 // Remove the abstract element. It is a non visible element.
1305                 $text = self::stripAbstract($text);
1306
1307                 // Move all spaces out of the tags
1308                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1309                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1310
1311                 // Extract the private images which use data urls since preg has issues with
1312                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1313                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1314
1315                 $extracted = self::extractImagesFromItemBody($text);
1316                 $text = $extracted['body'];
1317                 $saved_image = $extracted['images'];
1318
1319                 // If we find any event code, turn it into an event.
1320                 // After we're finished processing the bbcode we'll
1321                 // replace all of the event code with a reformatted version.
1322
1323                 $ev = Event::fromBBCode($text);
1324
1325                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1326                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1327
1328                 $text = str_replace("<", "&lt;", $text);
1329                 $text = str_replace(">", "&gt;", $text);
1330
1331                 // remove some newlines before the general conversion
1332                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1333                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1334
1335                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1336                 if (!$try_oembed) {
1337                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1338                 }
1339
1340                 // Convert new line chars to html <br /> tags
1341
1342                 // nlbr seems to be hopelessly messed up
1343                 //      $Text = nl2br($Text);
1344
1345                 // We'll emulate it.
1346
1347                 $text = trim($text);
1348                 $text = str_replace("\r\n", "\n", $text);
1349
1350                 // Remove linefeeds inside of the table elements. See issue #6799
1351                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1352                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1353                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1354                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1355                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1356                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1357                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1358                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1359                 do {
1360                         $oldtext = $text;
1361                         $text = str_replace($search, $replace, $text);
1362                 } while ($oldtext != $text);
1363
1364                 // Replace these here only once
1365                 $search = ["\n[table]", "[/table]\n"];
1366                 $replace = ["[table]", "[/table]"];
1367                 $text = str_replace($search, $replace, $text);
1368
1369                 // removing multiplicated newlines
1370                 if (Config::get('system', 'remove_multiplicated_lines')) {
1371                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1372                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1373                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1374                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1375                         do {
1376                                 $oldtext = $text;
1377                                 $text = str_replace($search, $replace, $text);
1378                         } while ($oldtext != $text);
1379                 }
1380
1381                 /// @todo Have a closer look at the different html modes
1382                 // Handle attached links or videos
1383                 if (in_array($simple_html, [9])) {
1384                         $text = self::removeAttachment($text);
1385                 } elseif (!in_array($simple_html, [0, 4])) {
1386                         $text = self::removeAttachment($text, true);
1387                 } else {
1388                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1389                 }
1390
1391                 // leave open the posibility of [map=something]
1392                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1393                 if (strpos($text, '[/map]') !== false) {
1394                         $text = preg_replace_callback(
1395                                 "/\[map\](.*?)\[\/map\]/ism",
1396                                 function ($match) use ($simple_html) {
1397                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1398                                 },
1399                                 $text
1400                         );
1401                 }
1402
1403                 if (strpos($text, '[map=') !== false) {
1404                         $text = preg_replace_callback(
1405                                 "/\[map=(.*?)\]/ism",
1406                                 function ($match) use ($simple_html) {
1407                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1408                                 },
1409                                 $text
1410                         );
1411                 }
1412
1413                 if (strpos($text, '[map]') !== false) {
1414                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1415                 }
1416
1417                 // Check for headers
1418                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1419                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1420                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1421                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1422                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1423                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1424
1425                 // Check for paragraph
1426                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1427
1428                 // Check for bold text
1429                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1430
1431                 // Check for Italics text
1432                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1433
1434                 // Check for Underline text
1435                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1436
1437                 // Check for strike-through text
1438                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1439
1440                 // Check for over-line text
1441                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1442
1443                 // Check for colored text
1444                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1445
1446                 // Check for sized text
1447                 // [size=50] --> font-size: 50px (with the unit).
1448                 if ($simple_html != 3) {
1449                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1450                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1451                 } else {
1452                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1453                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1454                 }
1455
1456
1457                 // Check for centered text
1458                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1459
1460                 // Check for list text
1461                 $text = str_replace("[*]", "<li>", $text);
1462
1463                 // Check for style sheet commands
1464                 $text = preg_replace_callback(
1465                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1466                         function ($match) {
1467                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1468                         },
1469                         $text
1470                 );
1471
1472                 // Check for CSS classes
1473                 $text = preg_replace_callback(
1474                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1475                         function ($match) {
1476                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1477                         },
1478                         $text
1479                 );
1480
1481                 // handle nested lists
1482                 $endlessloop = 0;
1483
1484                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1485                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1486                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1487                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1488                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1489                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1490                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1491                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1492                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1493                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1494                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1495                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1496                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1497                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1498                 }
1499
1500                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1501                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1502                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1503                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1504
1505                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1506                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1507
1508                 $text = str_replace('[hr]', '<hr />', $text);
1509
1510                 if (!$for_plaintext) {
1511                         $escaped = [];
1512
1513                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1514                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1515                                 function ($matches) use (&$escaped) {
1516                                         $return = '{escaped-' . count($escaped) . '}';
1517                                         $escaped[] = $matches[0];
1518
1519                                         return $return;
1520                                 },
1521                                 $text
1522                         );
1523
1524                         // Autolinker for isolated URLs
1525                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1526
1527                         // Restoring escaped blocks
1528                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1529                                 function ($matches) use ($escaped) {
1530                                         return $escaped[intval($matches[1])] ?? $matches[0];
1531                                 },
1532                                 $text
1533                         );
1534                 }
1535
1536                 // This is actually executed in Item::prepareBody()
1537
1538                 $nosmile = strpos($text, '[nosmile]') !== false;
1539                 $text = str_replace('[nosmile]', '', $text);
1540
1541                 // Check for font change text
1542                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1543
1544                 // Declare the format for [spoiler] layout
1545                 $SpoilerLayout = '<details class="spoiler"><summary>' . L10n::t('Click to open/close') . '</summary>$1</details>';
1546
1547                 // Check for [spoiler] text
1548                 // handle nested quotes
1549                 $endlessloop = 0;
1550                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1551                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1552                 }
1553
1554                 // Check for [spoiler=Title] text
1555
1556                 // handle nested quotes
1557                 $endlessloop = 0;
1558                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1559                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1560                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1561                                 $text);
1562                 }
1563
1564                 // Declare the format for [quote] layout
1565                 $QuoteLayout = '<blockquote>$1</blockquote>';
1566
1567                 // Check for [quote] text
1568                 // handle nested quotes
1569                 $endlessloop = 0;
1570                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1571                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1572                 }
1573
1574                 // Check for [quote=Author] text
1575
1576                 $t_wrote = L10n::t('$1 wrote:');
1577
1578                 // handle nested quotes
1579                 $endlessloop = 0;
1580                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1581                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1582                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1583                                 $text);
1584                 }
1585
1586
1587                 // [img=widthxheight]image source[/img]
1588                 $text = preg_replace_callback(
1589                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1590                         function ($matches) use ($simple_html) {
1591                                 if (strpos($matches[3], "data:image/") === 0) {
1592                                         return $matches[0];
1593                                 }
1594
1595                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1596                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1597                         },
1598                         $text
1599                 );
1600
1601                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1602                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1603
1604                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1605                         function ($matches) use ($simple_html) {
1606                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1607                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1608                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1609                         },
1610                         $text);
1611
1612                 // Images
1613                 // [img]pathtoimage[/img]
1614                 $text = preg_replace_callback(
1615                         "/\[img\](.*?)\[\/img\]/ism",
1616                         function ($matches) use ($simple_html) {
1617                                 if (strpos($matches[1], "data:image/") === 0) {
1618                                         return $matches[0];
1619                                 }
1620
1621                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1622                                 return "[img]" . $matches[1] . "[/img]";
1623                         },
1624                         $text
1625                 );
1626
1627                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1628                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1629
1630                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1631                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1632                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1633
1634                 // Simplify "video" element
1635                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1636
1637                 // Try to Oembed
1638                 if ($try_oembed) {
1639                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism", '<video src="$1" controls="controls" width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1640                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3).*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1641
1642                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1643                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1644                 } else {
1645                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1646                                 '<a href="$1" target="_blank">$1</a>', $text);
1647                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1648                                 '<a href="$1" target="_blank">$1</a>', $text);
1649                 }
1650
1651                 // html5 video and audio
1652
1653
1654                 if ($try_oembed) {
1655                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1656                 } else {
1657                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1658                 }
1659
1660                 // Youtube extensions
1661                 if ($try_oembed) {
1662                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1663                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1664                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1665                 }
1666
1667                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1668                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1669                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1670
1671                 if ($try_oembed) {
1672                         $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);
1673                 } else {
1674                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1675                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1676                 }
1677
1678                 if ($try_oembed) {
1679                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1680                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1681                 }
1682
1683                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1684                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1685
1686                 if ($try_oembed) {
1687                         $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);
1688                 } else {
1689                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1690                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1691                 }
1692
1693                 // oembed tag
1694                 $text = OEmbed::BBCode2HTML($text);
1695
1696                 // Avoid triple linefeeds through oembed
1697                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1698
1699                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1700                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1701                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1702                 // start which is always required). Allow desc with a missing summary for compatibility.
1703
1704                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1705                         $sub = Event::getHTML($ev, $simple_html);
1706
1707                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1708                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1709                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1710                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1711                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1712                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1713                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1714                 }
1715
1716                 // Replace non graphical smilies for external posts
1717                 if (!$nosmile && !$for_plaintext) {
1718                         $text = Smilies::replace($text);
1719                 }
1720
1721                 if (!$for_plaintext) {
1722                         if (in_array($simple_html, [7, 9])) {
1723                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1724                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1725                         }
1726                 } else {
1727                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1728                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1729                 }
1730
1731                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1732
1733                 // Remove all hashtag addresses
1734                 if ($simple_html && !in_array($simple_html, [3, 7, 9])) {
1735                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1736                 } elseif ($simple_html == 3) {
1737                         // The ! is converted to @ since Diaspora only understands the @
1738                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1739                                 '@<a href="$2">$3</a>',
1740                                 $text);
1741                 } elseif (in_array($simple_html, [7, 9])) {
1742                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1743                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1744                                 $text);
1745                 } elseif (!$simple_html) {
1746                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1747                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1748                                 $text);
1749                 }
1750
1751                 // Bookmarks in red - will be converted to bookmarks in friendica
1752                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1753                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1754                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1755                                         "[bookmark=$1]$2[/bookmark]", $text);
1756
1757                 if (in_array($simple_html, [2, 6, 7, 8])) {
1758                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1759                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1760                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1761                 }
1762
1763                 if ($simple_html == 5) {
1764                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1765                 }
1766
1767                 // Perform URL Search
1768                 if ($try_oembed) {
1769                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1770                 }
1771
1772                 if ($simple_html == 5) {
1773                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1774                 } else {
1775                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1776                 }
1777
1778                 // Handle Diaspora posts
1779                 $text = preg_replace_callback(
1780                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1781                         function ($match) {
1782                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1783                         }, $text
1784                 );
1785
1786                 $text = preg_replace_callback(
1787                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1788                         function ($match) {
1789                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1790                         }, $text
1791                 );
1792
1793                 // Server independent link to posts and comments
1794                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1795                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1796                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1797
1798                 /* Tag conversion
1799                  * Supports:
1800                  * - #[url=<anything>]<term>[/url]
1801                  * - [url=<anything>]#<term>[/url]
1802                  */
1803                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1804                         return '#<a href="'
1805                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1806                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1807                                 . XML::escape($matches[1])
1808                                 . '</a>';
1809                 }, $text);
1810
1811                 // We need no target="_blank" for local links
1812                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1813                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1814                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1815                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1816
1817                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1818                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1819
1820                 // Red compatibility, though the link can't be authenticated on Friendica
1821                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1822
1823
1824                 // we may need to restrict this further if it picks up too many strays
1825                 // link acct:user@host to a webfinger profile redirector
1826
1827                 $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="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1828
1829                 // Perform MAIL Search
1830                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1831                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1832
1833                 // Unhide all [noparse] contained bbtags unspacefying them
1834                 // and triming the [noparse] tag.
1835
1836                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1837                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1838                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1839
1840                 /// @todo What is the meaning of these lines?
1841                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1842                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1843
1844                 // Currently deactivated, it made problems with " inside of alt texts.
1845                 //$text = preg_replace('/\&quot\;/', '"', $text);
1846
1847                 // fix any escaped ampersands that may have been converted into links
1848                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1849
1850                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1851                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1852
1853                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1854
1855                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1856                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1857
1858                 // sanitize href attributes (only whitelisted protocols URLs)
1859                 // default value for backward compatibility
1860                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', []);
1861
1862                 // Always allowed protocol even if config isn't set or not including it
1863                 $allowed_link_protocols[] = '//';
1864                 $allowed_link_protocols[] = 'http://';
1865                 $allowed_link_protocols[] = 'https://';
1866                 $allowed_link_protocols[] = 'redir/';
1867
1868                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1869
1870                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1871                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1872
1873                 // Shared content
1874                 $text = self::convertShare(
1875                         $text,
1876                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1877                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1878                         }
1879                 );
1880
1881                 if ($saved_image) {
1882                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1883                 }
1884
1885                 // Restore code blocks
1886                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1887                         function ($matches) use ($codeblocks) {
1888                                 $return = $matches[0];
1889                                 if (isset($codeblocks[intval($matches[1])])) {
1890                                         $return = $codeblocks[$matches[1]];
1891                                 }
1892                                 return $return;
1893                         },
1894                         $text
1895                 );
1896
1897                 // Clean up the HTML by loading and saving the HTML with the DOM.
1898                 // Bad structured html can break a whole page.
1899                 // For performance reasons do it only with activated item cache or at export.
1900                 if (!$try_oembed || (get_itemcachepath() != '')) {
1901                         $doc = new DOMDocument();
1902                         $doc->preserveWhiteSpace = false;
1903
1904                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1905
1906                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1907                         $encoding = '<?xml encoding="UTF-8">';
1908                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1909                         $doc->encoding = 'UTF-8';
1910                         $text = $doc->saveHTML();
1911                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1912
1913                         $text = str_replace('<br></li>', '</li>', $text);
1914
1915                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1916                 }
1917
1918                 // Clean up some useless linebreaks in lists
1919                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1920                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1921                 //$Text = str_replace('</li><br />', '</li>', $Text);
1922                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1923                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1924
1925                 Hook::callAll('bbcode', $text);
1926
1927                 return trim($text);
1928         }
1929
1930         /**
1931          * @brief Strips the "abstract" tag from the provided text
1932          *
1933          * @param string $text The text with BBCode
1934          * @return string The same text - but without "abstract" element
1935          */
1936         public static function stripAbstract($text)
1937         {
1938                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1939                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1940
1941                 return $text;
1942         }
1943
1944         /**
1945          * @brief Returns the value of the "abstract" element
1946          *
1947          * @param string $text The text that maybe contains the element
1948          * @param string $addon The addon for which the abstract is meant for
1949          * @return string The abstract
1950          */
1951         public static function getAbstract($text, $addon = '')
1952         {
1953                 $abstract = '';
1954                 $abstracts = [];
1955                 $addon = strtolower($addon);
1956
1957                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1958                         foreach ($results AS $result) {
1959                                 $abstracts[strtolower($result[1])] = $result[2];
1960                         }
1961                 }
1962
1963                 if (isset($abstracts[$addon])) {
1964                         $abstract = $abstracts[$addon];
1965                 }
1966
1967                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1968                         $abstract = $result[1];
1969                 }
1970
1971                 return $abstract;
1972         }
1973
1974         /**
1975          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1976          *
1977          * @param array $match Matching values for the callback
1978          *                     [1] = Mention type (! or @)
1979          *                     [2] = Name
1980          *                     [3] = Address
1981          * @return string Replaced mention
1982          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1983          * @throws \ImagickException
1984          */
1985         private static function bbCodeMention2DiasporaCallback($match)
1986         {
1987                 $contact = Contact::getDetailsByURL($match[3]);
1988
1989                 if (empty($contact['addr'])) {
1990                         $contact = Probe::uri($match[3]);
1991                 }
1992
1993                 if (empty($contact['addr'])) {
1994                         return $match[0];
1995                 }
1996
1997                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1998                 return $mention;
1999         }
2000
2001         /**
2002          * @brief Converts a BBCode text into Markdown
2003          *
2004          * This function converts a BBCode item body to be sent to Markdown-enabled
2005          * systems like Diaspora and Libertree
2006          *
2007          * @param string $text
2008          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2009          * @return string
2010          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2011          */
2012         public static function toMarkdown($text, $for_diaspora = true)
2013         {
2014                 $a = DI::app();
2015
2016                 $original_text = $text;
2017
2018                 // Since Diaspora is creating a summary for links, this function removes them before posting
2019                 if ($for_diaspora) {
2020                         $text = self::removeShareInformation($text);
2021                 }
2022
2023                 /**
2024                  * Transform #tags, strip off the [url] and replace spaces with underscore
2025                  */
2026                 $url_search_string = "^\[\]";
2027                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2028                         function ($matches) {
2029                                 return '#' . str_replace(' ', '_', $matches[2]);
2030                         },
2031                         $text
2032                 );
2033
2034                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2035                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2036
2037                 // Convert it to HTML - don't try oembed
2038                 if ($for_diaspora) {
2039                         $text = self::convert($text, false, 3);
2040
2041                         // Add all tags that maybe were removed
2042                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2043                                 $tagline = '';
2044                                 foreach ($tags[2] as $tag) {
2045                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2046                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2047                                                 $tagline .= '#' . $tag . ' ';
2048                                         }
2049                                 }
2050                                 $text = $text . " " . $tagline;
2051                         }
2052                 } else {
2053                         $text = self::convert($text, false, 4);
2054                 }
2055
2056                 // If a link is followed by a quote then there should be a newline before it
2057                 // Maybe we should make this newline at every time before a quote.
2058                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2059
2060                 $stamp1 = microtime(true);
2061
2062                 // Now convert HTML to Markdown
2063                 $text = HTML::toMarkdown($text);
2064
2065                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
2066
2067                 // Libertree has a problem with escaped hashtags.
2068                 $text = str_replace(['\#'], ['#'], $text);
2069
2070                 // Remove any leading or trailing whitespace, as this will mess up
2071                 // the Diaspora signature verification and cause the item to disappear
2072                 $text = trim($text);
2073
2074                 if ($for_diaspora) {
2075                         $url_search_string = "^\[\]";
2076                         $text = preg_replace_callback(
2077                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2078                                 ['self', 'bbCodeMention2DiasporaCallback'],
2079                                 $text
2080                         );
2081                 }
2082
2083                 Hook::callAll('bb2diaspora', $text);
2084
2085                 return $text;
2086         }
2087
2088         /**
2089      * @brief Pull out all #hashtags and @person tags from $string.
2090      *
2091      * We also get @person@domain.com - which would make
2092      * the regex quite complicated as tags can also
2093      * end a sentence. So we'll run through our results
2094      * and strip the period from any tags which end with one.
2095      * Returns array of tags found, or empty array.
2096      *
2097      * @param string $string Post content
2098      *
2099      * @return array List of tag and person names
2100      */
2101     public static function getTags($string)
2102     {
2103         $ret = [];
2104
2105         // Convert hashtag links to hashtags
2106         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2107
2108         // ignore anything in a code block
2109         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2110
2111         // Force line feeds at bbtags
2112         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2113
2114         // ignore anything in a bbtag
2115         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2116
2117         // Match full names against @tags including the space between first and last
2118         // We will look these up afterward to see if they are full names or not recognisable.
2119
2120         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2121             foreach ($matches[1] as $match) {
2122                 if (strstr($match, ']')) {
2123                     // we might be inside a bbcode color tag - leave it alone
2124                     continue;
2125                 }
2126
2127                 if (substr($match, -1, 1) === '.') {
2128                     $ret[] = substr($match, 0, -1);
2129                 } else {
2130                     $ret[] = $match;
2131                 }
2132             }
2133         }
2134
2135         // Otherwise pull out single word tags. These can be @nickname, @first_last
2136         // and #hash tags.
2137
2138         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2139             foreach ($matches[1] as $match) {
2140                 if (strstr($match, ']')) {
2141                     // we might be inside a bbcode color tag - leave it alone
2142                     continue;
2143                 }
2144                 if (substr($match, -1, 1) === '.') {
2145                     $match = substr($match,0,-1);
2146                 }
2147                 // ignore strictly numeric tags like #1
2148                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2149                     continue;
2150                 }
2151                 // try not to catch url fragments
2152                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2153                     continue;
2154                 }
2155                 $ret[] = $match;
2156             }
2157         }
2158
2159         return $ret;
2160     }
2161 }