src/Content/Text/BBCode.php

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