src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Addon;
  15 use Friendica\Core\Cache;
  16 use Friendica\Core\Config;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Protocol;
  19 use Friendica\Core\System;
  20 use Friendica\Model\Contact;
  21 use Friendica\Model\Event;
  22 use Friendica\Network\Probe;
  23 use Friendica\Object\Image;
  24 use Friendica\Util\Map;
  25 use Friendica\Util\Network;
  26 use Friendica\Util\ParseUrl;
  27 use League\HTMLToMarkdown\HtmlConverter;
  28
  29 require_once "mod/proxy.php";
  30
  31 class BBCode extends BaseObject
  32 {
  33         /**
  34          * @brief Fetches attachment data that were generated the old way
  35          *
  36          * @param string $body Message body
  37          * @return array
  38          * 'type' -> Message type ("link", "video", "photo")
  39          * 'text' -> Text before the shared message
  40          * 'after' -> Text after the shared message
  41          * 'image' -> Preview image of the message
  42          * 'url' -> Url to the attached message
  43          * 'title' -> Title of the attachment
  44          * 'description' -> Description of the attachment
  45          */
  46         private static function getOldAttachmentData($body)
  47         {
  48                 $post = [];
  49
  50                 // Simplify image codes
  51                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  52
  53                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  54                         foreach ($attached as $data) {
  55                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  56                                         continue;
  57                                 }
  58
  59                                 $post["type"] = substr($data[1], 5);
  60
  61                                 $pos = strpos($body, $data[0]);
  62                                 if ($pos > 0) {
  63                                         $post["text"] = trim(substr($body, 0, $pos));
  64                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  65                                 } else {
  66                                         $post["text"] = trim(str_replace($data[0], "", $body));
  67                                 }
  68
  69                                 $attacheddata = $data[2];
  70
  71                                 $URLSearchString = "^\[\]";
  72
  73                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  74
  75                                         $picturedata = Image::getInfoFromURL($matches[1]);
  76
  77                                         if ($picturedata) {
  78                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  79                                                         $post["image"] = $matches[1];
  80                                                 } else {
  81                                                         $post["preview"] = $matches[1];
  82                                                 }
  83                                         }
  84                                 }
  85
  86                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  87                                         $post["url"] = $matches[1];
  88                                         $post["title"] = $matches[2];
  89                                 }
  90                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  91                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  92                                         $post["url"] = $matches[1];
  93                                 }
  94
  95                                 // Search for description
  96                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  97                                         $post["description"] = $matches[1];
  98                                 }
  99                         }
 100                 }
 101                 return $post;
 102         }
 103
 104         /**
 105          * @brief Fetches attachment data that were generated with the "attachment" element
 106          *
 107          * @param string $body Message body
 108          * @return array
 109          * 'type' -> Message type ("link", "video", "photo")
 110          * 'text' -> Text before the shared message
 111          * 'after' -> Text after the shared message
 112          * 'image' -> Preview image of the message
 113          * 'url' -> Url to the attached message
 114          * 'title' -> Title of the attachment
 115          * 'description' -> Description of the attachment
 116          */
 117         public static function getAttachmentData($body)
 118         {
 119                 $data = [];
 120
 121                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 122                         return self::getOldAttachmentData($body);
 123                 }
 124
 125                 $attributes = $match[2];
 126
 127                 $data["text"] = trim($match[1]);
 128
 129                 $type = "";
 130                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 131                 if (x($matches, 1)) {
 132                         $type = strtolower($matches[1]);
 133                 }
 134
 135                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 136                 if (x($matches, 1)) {
 137                         $type = strtolower($matches[1]);
 138                 }
 139
 140                 if ($type == "") {
 141                         return [];
 142                 }
 143
 144                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 145                         return [];
 146                 }
 147
 148                 if ($type != "") {
 149                         $data["type"] = $type;
 150                 }
 151
 152                 $url = "";
 153                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 154                 if (x($matches, 1)) {
 155                         $url = $matches[1];
 156                 }
 157
 158                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 159                 if (x($matches, 1)) {
 160                         $url = $matches[1];
 161                 }
 162
 163                 if ($url != "") {
 164                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 165                 }
 166
 167                 $title = "";
 168                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 169                 if (x($matches, 1)) {
 170                         $title = $matches[1];
 171                 }
 172
 173                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 174                 if (x($matches, 1)) {
 175                         $title = $matches[1];
 176                 }
 177
 178                 if ($title != "") {
 179                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 180                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 181                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 182                         $data["title"] = $title;
 183                 }
 184
 185                 $image = "";
 186                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 187                 if (x($matches, 1)) {
 188                         $image = $matches[1];
 189                 }
 190
 191                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 192                 if (x($matches, 1)) {
 193                         $image = $matches[1];
 194                 }
 195
 196                 if ($image != "") {
 197                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 198                 }
 199
 200                 $preview = "";
 201                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 202                 if (x($matches, 1)) {
 203                         $preview = $matches[1];
 204                 }
 205
 206                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 207                 if (x($matches, 1)) {
 208                         $preview = $matches[1];
 209                 }
 210
 211                 if ($preview != "") {
 212                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 213                 }
 214
 215                 $data["description"] = trim($match[3]);
 216
 217                 $data["after"] = trim($match[4]);
 218
 219                 return $data;
 220         }
 221
 222         public static function getAttachedData($body, $item = [])
 223         {
 224                 /*
 225                 - text:
 226                 - type: link, video, photo
 227                 - title:
 228                 - url:
 229                 - image:
 230                 - description:
 231                 - (thumbnail)
 232                 */
 233
 234                 $has_title = !empty($item['title']);
 235                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 236                 $post = self::getAttachmentData($body);
 237
 238                 // if nothing is found, it maybe having an image.
 239                 if (!isset($post["type"])) {
 240                         // Simplify image codes
 241                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 242
 243                         $URLSearchString = "^\[\]";
 244
 245                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 246
 247                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 248                                 if ((count($pictures) == 1) && !$has_title) {
 249                                         // Checking, if the link goes to a picture
 250                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 251
 252                                         // Workaround:
 253                                         // Sometimes photo posts to the own album are not detected at the start.
 254                                         // So we seem to cannot use the cache for these cases. That's strange.
 255                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 256                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 257                                         }
 258
 259                                         if ($data["type"] == "photo") {
 260                                                 $post["type"] = "photo";
 261                                                 if (isset($data["images"][0])) {
 262                                                         $post["image"] = $data["images"][0]["src"];
 263                                                         $post["url"] = $data["url"];
 264                                                 } else {
 265                                                         $post["image"] = $data["url"];
 266                                                 }
 267
 268                                                 $post["preview"] = $pictures[0][2];
 269                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 270                                         } else {
 271                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 272                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 273                                                         $post["type"] = "photo";
 274                                                         $post["image"] = $pictures[0][1];
 275                                                         $post["preview"] = $pictures[0][2];
 276                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 277                                                 }
 278                                         }
 279                                 } elseif (count($pictures) > 0) {
 280                                         $post["type"] = "link";
 281                                         $post["url"] = $plink;
 282                                         $post["image"] = $pictures[0][2];
 283                                         $post["text"] = $body;
 284                                 }
 285                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 286                                 if ((count($pictures) == 1) && !$has_title) {
 287                                         $post["type"] = "photo";
 288                                         $post["image"] = $pictures[0][1];
 289                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 290                                 } elseif (count($pictures) > 0) {
 291                                         $post["type"] = "link";
 292                                         $post["url"] = $plink;
 293                                         $post["image"] = $pictures[0][1];
 294                                         $post["text"] = $body;
 295                                 }
 296                         }
 297
 298                         // Test for the external links
 299                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 300                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 301
 302                         $links = array_merge($links1, $links2);
 303
 304                         // If there is only a single one, then use it.
 305                         // This should cover link posts via API.
 306                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 307                                 $post["type"] = "link";
 308                                 $post["text"] = trim($body);
 309                                 $post["url"] = $links[0][1];
 310                         }
 311
 312                         // Now count the number of external media links
 313                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 314                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 315                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 316                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 317
 318                         // Add them to the other external links
 319                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 320
 321                         // Are there more than one?
 322                         if (count($links) > 1) {
 323                                 // The post will be the type "text", which means a blog post
 324                                 unset($post["type"]);
 325                                 $post["url"] = $plink;
 326                         }
 327
 328                         if (!isset($post["type"])) {
 329                                 $post["type"] = "text";
 330                                 $post["text"] = trim($body);
 331                         }
 332                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 333                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 334
 335                         if (isset($data["images"][0])) {
 336                                 $post["image"] = $data["images"][0]["src"];
 337                         }
 338                 }
 339
 340                 return $post;
 341         }
 342
 343         /**
 344          * @brief Converts a BBCode text into plaintext
 345          *
 346          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 347          *
 348          * @return string
 349          */
 350         public static function toPlaintext($text, $keep_urls = true)
 351         {
 352                 $naked_text = preg_replace('/\[(.+?)\]/','', $text);
 353                 if (!$keep_urls) {
 354                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 355                 }
 356
 357                 return $naked_text;
 358         }
 359
 360         private static function proxyUrl($image, $simplehtml = false)
 361         {
 362                 // Only send proxied pictures to API and for internal display
 363                 if (in_array($simplehtml, [false, 2])) {
 364                         return proxy_url($image);
 365                 } else {
 366                         return $image;
 367                 }
 368         }
 369
 370         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 371         {
 372                 // Suppress "view full size"
 373                 if (intval(Config::get('system', 'no_view_full_size'))) {
 374                         $include_link = false;
 375                 }
 376
 377                 // Picture addresses can contain special characters
 378                 $s = htmlspecialchars_decode($srctext);
 379
 380                 $matches = null;
 381                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 382                 if ($c) {
 383                         foreach ($matches as $mtch) {
 384                                 logger('scale_external_image: ' . $mtch[1]);
 385
 386                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 387                                 if (stristr($mtch[1], $hostname)) {
 388                                         continue;
 389                                 }
 390
 391                                 // $scale_replace, if passed, is an array of two elements. The
 392                                 // first is the name of the full-size image. The second is the
 393                                 // name of a remote, scaled-down version of the full size image.
 394                                 // This allows Friendica to display the smaller remote image if
 395                                 // one exists, while still linking to the full-size image
 396                                 if ($scale_replace) {
 397                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 398                                 } else {
 399                                         $scaled = $mtch[1];
 400                                 }
 401                                 $i = Network::fetchUrl($scaled);
 402                                 if (!$i) {
 403                                         return $srctext;
 404                                 }
 405
 406                                 // guess mimetype from headers or filename
 407                                 $type = Image::guessType($mtch[1], true);
 408
 409                                 if ($i) {
 410                                         $Image = new Image($i, $type);
 411                                         if ($Image->isValid()) {
 412                                                 $orig_width = $Image->getWidth();
 413                                                 $orig_height = $Image->getHeight();
 414
 415                                                 if ($orig_width > 640 || $orig_height > 640) {
 416                                                         $Image->scaleDown(640);
 417                                                         $new_width = $Image->getWidth();
 418                                                         $new_height = $Image->getHeight();
 419                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 420                                                         $s = str_replace(
 421                                                                 $mtch[0],
 422                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 423                                                                 . "\n" . (($include_link)
 424                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 425                                                                         : ''),
 426                                                                 $s
 427                                                         );
 428                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 429                                                 }
 430                                         }
 431                                 }
 432                         }
 433                 }
 434
 435                 // replace the special char encoding
 436                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 437                 return $s;
 438         }
 439
 440         /**
 441          * The purpose of this function is to apply system message length limits to
 442          * imported messages without including any embedded photos in the length
 443          *
 444          * @brief Truncates imported message body string length to max_import_size
 445          * @param string $body
 446          * @return string
 447          */
 448         public static function limitBodySize($body)
 449         {
 450                 $maxlen = Config::get('config', 'max_import_size', 0);
 451
 452                 // If the length of the body, including the embedded images, is smaller
 453                 // than the maximum, then don't waste time looking for the images
 454                 if ($maxlen && (strlen($body) > $maxlen)) {
 455
 456                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
 457
 458                         $orig_body = $body;
 459                         $new_body = '';
 460                         $textlen = 0;
 461
 462                         $img_start = strpos($orig_body, '[img');
 463                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 464                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 465                         while (($img_st_close !== false) && ($img_end !== false)) {
 466
 467                                 $img_st_close++; // make it point to AFTER the closing bracket
 468                                 $img_end += $img_start;
 469                                 $img_end += strlen('[/img]');
 470
 471                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 472                                         // This is an embedded image
 473
 474                                         if (($textlen + $img_start) > $maxlen) {
 475                                                 if ($textlen < $maxlen) {
 476                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
 477                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 478                                                         $textlen = $maxlen;
 479                                                 }
 480                                         } else {
 481                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 482                                                 $textlen += $img_start;
 483                                         }
 484
 485                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 486                                 } else {
 487
 488                                         if (($textlen + $img_end) > $maxlen) {
 489                                                 if ($textlen < $maxlen) {
 490                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
 491                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 492                                                         $textlen = $maxlen;
 493                                                 }
 494                                         } else {
 495                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 496                                                 $textlen += $img_end;
 497                                         }
 498                                 }
 499                                 $orig_body = substr($orig_body, $img_end);
 500
 501                                 if ($orig_body === false) {
 502                                         // in case the body ends on a closing image tag
 503                                         $orig_body = '';
 504                                 }
 505
 506                                 $img_start = strpos($orig_body, '[img');
 507                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 508                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 509                         }
 510
 511                         if (($textlen + strlen($orig_body)) > $maxlen) {
 512                                 if ($textlen < $maxlen) {
 513                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
 514                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 515                                 }
 516                         } else {
 517                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
 518                                 $new_body = $new_body . $orig_body;
 519                         }
 520
 521                         return $new_body;
 522                 } else {
 523                         return $body;
 524                 }
 525         }
 526
 527         /**
 528          * Processes [attachment] tags
 529          *
 530          * Note: Can produce a [bookmark] tag in the returned string
 531          *
 532          * @brief Processes [attachment] tags
 533          * @param string $return
 534          * @param bool|int $simplehtml
 535          * @param bool $tryoembed
 536          * @return string
 537          */
 538         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 539         {
 540                 $data = self::getAttachmentData($return);
 541                 if (empty($data) || empty($data["url"])) {
 542                         return $return;
 543                 }
 544
 545                 if (isset($data["title"])) {
 546                         $data["title"] = strip_tags($data["title"]);
 547                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 548                 }
 549
 550                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 551                         $data["preview"] = $data["image"];
 552                         $data["image"] = "";
 553                 }
 554
 555                 $return = '';
 556                 if ($simplehtml == 7) {
 557                         $return = self::convertUrlForOStatus($data["url"]);
 558                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 559                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 560                 } else {
 561                         try {
 562                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 563                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 564                                 } else {
 565                                         throw new Exception('OEmbed is disabled for this attachment.');
 566                                 }
 567                         } catch (Exception $e) {
 568                                 $data["title"] = defaults($data, 'title', $data['url']);
 569
 570                                 if ($simplehtml != 4) {
 571                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 572                                 }
 573
 574                                 if (!empty($data["image"])) {
 575                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 576                                 } elseif (!empty($data["preview"])) {
 577                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], self::proxyUrl($data["preview"], $simplehtml), $data["title"]);
 578                                 }
 579
 580                                 if (($data["type"] == "photo") && !empty($data["url"]) && !empty($data["image"])) {
 581                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 582                                 } else {
 583                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 584                                 }
 585
 586                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
 587                                         // Sanitize the HTML by converting it to BBCode
 588                                         $bbcode = HTML::toBBCode($data["description"]);
 589                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 590                                 }
 591                                 if ($data["type"] == "link") {
 592                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 593                                 }
 594
 595                                 if ($simplehtml != 4) {
 596                                         $return .= '</div>';
 597                                 }
 598                         }
 599                 }
 600
 601                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
 602         }
 603
 604         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 605         {
 606                 $data = self::getAttachmentData($Text);
 607
 608                 if (!$data) {
 609                         return $Text;
 610                 } elseif ($nolink) {
 611                         return $data["text"] . $data["after"];
 612                 }
 613
 614                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 615                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 616                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 617                         $data["title"] = $data["url"];
 618                 } elseif (($text != "") && strstr($title, $text)) {
 619                         $data["text"] = $data["title"];
 620                         $data["title"] = $data["url"];
 621                 }
 622
 623                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
 624                         return $data["title"] . $data["after"];
 625                 }
 626
 627                 // If the link already is included in the post, don't add it again
 628                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
 629                         return $data["text"] . $data["after"];
 630                 }
 631
 632                 $text = $data["text"];
 633
 634                 if (($data["url"] != "") && ($data["title"] != "")) {
 635                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 636                 } elseif (($data["url"] != "")) {
 637                         $text .= "\n[url]" . $data["url"] . "[/url]";
 638                 }
 639
 640                 return $text . "\n" . $data["after"];
 641         }
 642
 643         /**
 644          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 645          *
 646          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 647          * @param array $match Array with the matching values
 648          * @return string reformatted link including HTML codes
 649          */
 650         private static function convertUrlForOStatusCallback($match)
 651         {
 652                 $url = $match[1];
 653
 654                 if (isset($match[2]) && ($match[1] != $match[2])) {
 655                         return $match[0];
 656                 }
 657
 658                 $parts = parse_url($url);
 659                 if (!isset($parts['scheme'])) {
 660                         return $match[0];
 661                 }
 662
 663                 return self::convertUrlForOStatus($url);
 664         }
 665
 666         /**
 667          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 668          * @param string $url URL that is about to be reformatted
 669          * @return string reformatted link including HTML codes
 670          */
 671         private static function convertUrlForOStatus($url)
 672         {
 673                 $parts = parse_url($url);
 674                 $scheme = $parts['scheme'] . '://';
 675                 $styled_url = str_replace($scheme, '', $url);
 676
 677                 if (strlen($styled_url) > 30) {
 678                         $styled_url = substr($styled_url, 0, 30) . "…";
 679                 }
 680
 681                 $html = '<a href="%s" target="_blank">%s</a>';
 682
 683                 return sprintf($html, $url, $styled_url);
 684         }
 685
 686         /*
 687          * [noparse][i]italic[/i][/noparse] turns into
 688          * [noparse][ i ]italic[ /i ][/noparse],
 689          * to hide them from parser.
 690          */
 691         private static function escapeNoparseCallback($match)
 692         {
 693                 $whole_match = $match[0];
 694                 $captured = $match[1];
 695                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 696                 $new_str = str_replace($captured, $spacefied, $whole_match);
 697                 return $new_str;
 698         }
 699
 700         /*
 701          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 702          * now turns back and the [noparse] tags are trimed
 703          * returning [i]italic[/i]
 704          */
 705         private static function unescapeNoparseCallback($match)
 706         {
 707                 $captured = $match[1];
 708                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 709                 return $unspacefied;
 710         }
 711
 712         /**
 713          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 714          * occurrences
 715          *
 716          * @param string $text        Text to search
 717          * @param string $name        Tag name
 718          * @param int    $occurrences Number of first occurrences to skip
 719          * @return boolean|array
 720          */
 721         public static function getTagPosition($text, $name, $occurrences = 0)
 722         {
 723                 if ($occurrences < 0) {
 724                         $occurrences = 0;
 725                 }
 726
 727                 $start_open = -1;
 728                 for ($i = 0; $i <= $occurrences; $i++) {
 729                         if ($start_open !== false) {
 730                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 731                         }
 732                 }
 733
 734                 if ($start_open === false) {
 735                         return false;
 736                 }
 737
 738                 $start_equal = strpos($text, '=', $start_open);
 739                 $start_close = strpos($text, ']', $start_open);
 740
 741                 if ($start_close === false) {
 742                         return false;
 743                 }
 744
 745                 $start_close++;
 746
 747                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 748
 749                 if ($end_open === false) {
 750                         return false;
 751                 }
 752
 753                 $res = [
 754                         'start' => [
 755                                 'open' => $start_open,
 756                                 'close' => $start_close
 757                         ],
 758                         'end' => [
 759                                 'open' => $end_open,
 760                                 'close' => $end_open + strlen('[/' . $name . ']')
 761                         ],
 762                 ];
 763
 764                 if ($start_equal !== false) {
 765                         $res['start']['equal'] = $start_equal + 1;
 766                 }
 767
 768                 return $res;
 769         }
 770
 771         /**
 772          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 773          *
 774          * @param type $pattern Preg pattern string
 775          * @param type $replace Preg replace string
 776          * @param type $name    BBCode tag name
 777          * @param type $text    Text to search
 778          * @return string
 779          */
 780         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 781         {
 782                 $occurrences = 0;
 783                 $pos = self::getTagPosition($text, $name, $occurrences);
 784                 while ($pos !== false && $occurrences++ < 1000) {
 785                         $start = substr($text, 0, $pos['start']['open']);
 786                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 787                         $end = substr($text, $pos['end']['close']);
 788                         if ($end === false) {
 789                                 $end = '';
 790                         }
 791
 792                         $subject = preg_replace($pattern, $replace, $subject);
 793                         $text = $start . $subject . $end;
 794
 795                         $pos = self::getTagPosition($text, $name, $occurrences);
 796                 }
 797
 798                 return $text;
 799         }
 800
 801         private static function extractImagesFromItemBody($body)
 802         {
 803                 $saved_image = [];
 804                 $orig_body = $body;
 805                 $new_body = '';
 806
 807                 $cnt = 0;
 808                 $img_start = strpos($orig_body, '[img');
 809                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 810                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 811                 while (($img_st_close !== false) && ($img_end !== false)) {
 812                         $img_st_close++; // make it point to AFTER the closing bracket
 813                         $img_end += $img_start;
 814
 815                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 816                                 // This is an embedded image
 817                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 818                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 819
 820                                 $cnt++;
 821                         } else {
 822                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 823                         }
 824
 825                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 826
 827                         if ($orig_body === false) {
 828                                 // in case the body ends on a closing image tag
 829                                 $orig_body = '';
 830                         }
 831
 832                         $img_start = strpos($orig_body, '[img');
 833                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 834                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 835                 }
 836
 837                 $new_body = $new_body . $orig_body;
 838
 839                 return ['body' => $new_body, 'images' => $saved_image];
 840         }
 841
 842         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 843         {
 844                 $newbody = $body;
 845
 846                 $cnt = 0;
 847                 foreach ($images as $image) {
 848                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 849                         // it loops over the array starting from the first element and going sequentially
 850                         // to the last element
 851                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 852                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 853                         $cnt++;
 854                 }
 855
 856                 return $newbody;
 857         }
 858
 859         /**
 860          * Processes [share] tags
 861          *
 862          * Note: Can produce a [bookmark] tag in the output
 863          *
 864          * @brief Processes [share] tags
 865          * @param array    $share      preg_match_callback result array
 866          * @param bool|int $simplehtml
 867          * @return string
 868          */
 869         private static function convertShare($share, $simplehtml)
 870         {
 871                 $attributes = $share[2];
 872
 873                 $author = "";
 874                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
 875                 if (x($matches, 1)) {
 876                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
 877                 }
 878
 879                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
 880                 if (x($matches, 1)) {
 881                         $author = $matches[1];
 882                 }
 883
 884                 $profile = "";
 885                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
 886                 if (x($matches, 1)) {
 887                         $profile = $matches[1];
 888                 }
 889
 890                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
 891                 if (x($matches, 1)) {
 892                         $profile = $matches[1];
 893                 }
 894
 895                 $avatar = "";
 896                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
 897                 if (x($matches, 1)) {
 898                         $avatar = $matches[1];
 899                 }
 900
 901                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
 902                 if (x($matches, 1)) {
 903                         $avatar = $matches[1];
 904                 }
 905
 906                 $link = "";
 907                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
 908                 if (x($matches, 1)) {
 909                         $link = $matches[1];
 910                 }
 911
 912                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
 913                 if (x($matches, 1)) {
 914                         $link = $matches[1];
 915                 }
 916
 917                 $posted = "";
 918
 919                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
 920                 if (x($matches, 1)) {
 921                         $posted = $matches[1];
 922                 }
 923
 924                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
 925                 if (x($matches, 1)) {
 926                         $posted = $matches[1];
 927                 }
 928
 929                 // We only call this so that a previously unknown contact can be added.
 930                 // This is important for the function "Model\Contact::getDetailsByURL()".
 931                 // This function then can fetch an entry from the contact table.
 932                 Contact::getIdForURL($profile, 0, true);
 933
 934                 $data = Contact::getDetailsByURL($profile);
 935
 936                 if (x($data, "name") && x($data, "addr")) {
 937                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
 938                 } else {
 939                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
 940                 }
 941
 942                 if (x($data, "addr")) {
 943                         $userid = $data["addr"];
 944                 } else {
 945                         $userid = Protocol::formatMention($profile, $author);
 946                 }
 947
 948                 if (x($data, "name")) {
 949                         $author = $data["name"];
 950                 }
 951
 952                 if (x($data, "micro")) {
 953                         $avatar = $data["micro"];
 954                 }
 955
 956                 $preshare = trim($share[1]);
 957                 if ($preshare != "") {
 958                         $preshare .= "<br />";
 959                 }
 960
 961                 switch ($simplehtml) {
 962                         case 1:
 963                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
 964                                 break;
 965                         case 2:
 966                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
 967                                 break;
 968                         case 3: // Diaspora
 969                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
 970
 971                                 $text = trim($share[1]);
 972
 973                                 if ($text != "") {
 974                                         $text .= "<hr />";
 975                                 }
 976
 977                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
 978                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
 979                                 } else {
 980                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
 981
 982                                         if ($link != "") {
 983                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
 984                                         }
 985                                 }
 986
 987                                 break;
 988                         case 4:
 989                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 990                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
 991                                 $headline .= ":</b><br />";
 992
 993                                 $text = trim($share[1]);
 994
 995                                 if ($text != "") {
 996                                         $text .= "<hr />";
 997                                 }
 998
 999                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1000
1001                                 break;
1002                         case 5:
1003                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1004                                 break;
1005                         case 6: // app.net
1006                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1007                                 break;
1008                         case 7: // statusnet/GNU Social
1009                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1010                                 break;
1011                         case 8: // twitter
1012                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1013                                 break;
1014                         case 9: // Google+/Facebook
1015                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1016
1017                                 if ($link != "") {
1018                                         $text .= "<br /><br />" . $link;
1019                                 }
1020                                 break;
1021                         default:
1022                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1023                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1024                                         try {
1025                                                 $oembed = OEmbed::getHTML($link, $preshare);
1026                                         } catch (Exception $e) {
1027                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1028                                         }
1029
1030                                         $text = $preshare . $oembed;
1031                                 } else {
1032                                         $text = trim($share[1]) . "\n";
1033
1034                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1035
1036                                         $tpl = get_markup_template('shared_content.tpl');
1037                                         $text .= replace_macros($tpl, [
1038                                                 '$profile' => $profile,
1039                                                 '$avatar' => $avatar,
1040                                                 '$author' => $author,
1041                                                 '$link' => $link,
1042                                                 '$posted' => $posted,
1043                                                 '$content' => trim($share[3])
1044                                         ]);
1045                                 }
1046                                 break;
1047                 }
1048
1049                 return $text;
1050         }
1051
1052         private static function removePictureLinksCallback($match)
1053         {
1054                 $text = Cache::get($match[1]);
1055
1056                 if (is_null($text)) {
1057                         $a = self::getApp();
1058
1059                         $stamp1 = microtime(true);
1060
1061                         $ch = @curl_init($match[1]);
1062                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1063                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1064                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1065                         @curl_exec($ch);
1066                         $curl_info = @curl_getinfo($ch);
1067
1068                         $a->save_timestamp($stamp1, "network");
1069
1070                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1071                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1072                         } else {
1073                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1074
1075                                 // if its not a picture then look if its a page that contains a picture link
1076                                 $body = Network::fetchUrl($match[1]);
1077
1078                                 $doc = new DOMDocument();
1079                                 @$doc->loadHTML($body);
1080                                 $xpath = new DOMXPath($doc);
1081                                 $list = $xpath->query("//meta[@name]");
1082                                 foreach ($list as $node) {
1083                                         $attr = [];
1084
1085                                         if ($node->attributes->length) {
1086                                                 foreach ($node->attributes as $attribute) {
1087                                                         $attr[$attribute->name] = $attribute->value;
1088                                                 }
1089                                         }
1090
1091                                         if (strtolower($attr["name"]) == "twitter:image") {
1092                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1093                                         }
1094                                 }
1095                         }
1096                         Cache::set($match[1], $text);
1097                 }
1098
1099                 return $text;
1100         }
1101
1102         private static function expandLinksCallback($match)
1103         {
1104                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1105                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1106                 } else {
1107                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1108                 }
1109         }
1110
1111         private static function cleanPictureLinksCallback($match)
1112         {
1113                 $text = Cache::get($match[1]);
1114
1115                 if (is_null($text)) {
1116                         $a = self::getApp();
1117
1118                         $stamp1 = microtime(true);
1119
1120                         $ch = @curl_init($match[1]);
1121                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1122                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1123                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1124                         @curl_exec($ch);
1125                         $curl_info = @curl_getinfo($ch);
1126
1127                         $a->save_timestamp($stamp1, "network");
1128
1129                         // if its a link to a picture then embed this picture
1130                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1131                                 $text = "[img]" . $match[1] . "[/img]";
1132                         } else {
1133                                 $text = "[img]" . $match[2] . "[/img]";
1134
1135                                 // if its not a picture then look if its a page that contains a picture link
1136                                 $body = Network::fetchUrl($match[1]);
1137
1138                                 $doc = new DOMDocument();
1139                                 @$doc->loadHTML($body);
1140                                 $xpath = new DOMXPath($doc);
1141                                 $list = $xpath->query("//meta[@name]");
1142                                 foreach ($list as $node) {
1143                                         $attr = [];
1144                                         if ($node->attributes->length) {
1145                                                 foreach ($node->attributes as $attribute) {
1146                                                         $attr[$attribute->name] = $attribute->value;
1147                                                 }
1148                                         }
1149
1150                                         if (strtolower($attr["name"]) == "twitter:image") {
1151                                                 $text = "[img]" . $attr["content"] . "[/img]";
1152                                         }
1153                                 }
1154                         }
1155                         Cache::set($match[1], $text);
1156                 }
1157
1158                 return $text;
1159         }
1160
1161         public static function cleanPictureLinks($text)
1162         {
1163                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1164                 return $return;
1165         }
1166
1167         private static function textHighlightCallback($match)
1168         {
1169                 // Fallback in case the language doesn't exist
1170                 $return = '[code]' . $match[2] . '[/code]';
1171
1172                 if (in_array(strtolower($match[1]),
1173                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1174                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1175                 ) {
1176                         $return = text_highlight($match[2], strtolower($match[1]));
1177                 }
1178
1179                 return $return;
1180         }
1181
1182         /**
1183          * @brief Converts a BBCode message to HTML message
1184          *
1185          * BBcode 2 HTML was written by WAY2WEB.net
1186          * extended to work with Mistpark/Friendica - Mike Macgirvin
1187          *
1188          * Simple HTML values meaning:
1189          * - 0: Friendica display
1190          * - 1: Unused
1191          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1192          * - 3: Used before converting to Markdown in bb2diaspora.php
1193          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1194          * - 5: Unused
1195          * - 6: Used for Appnet
1196          * - 7: Used for dfrn, OStatus
1197          * - 8: Used for WP backlink text setting
1198          *
1199          * @param string $text
1200          * @param bool   $try_oembed
1201          * @param int    $simple_html
1202          * @param bool   $for_plaintext
1203          * @return string
1204          */
1205         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1206         {
1207                 $a = self::getApp();
1208
1209                 /*
1210                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1211                  *
1212                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1213                  * $match[1] = $url
1214                  * $match[2] = $title or absent
1215                  */
1216                 $try_oembed_callback = function ($match)
1217                 {
1218                         $url = $match[1];
1219                         $title = defaults($match, 2, null);
1220
1221                         try {
1222                                 $return = OEmbed::getHTML($url, $title);
1223                         } catch (Exception $ex) {
1224                                 $return = $match[0];
1225                         }
1226
1227                         return $return;
1228                 };
1229
1230                 // Hide all [noparse] contained bbtags by spacefying them
1231                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1232
1233                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1234                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1235                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1236
1237                 // Remove the abstract element. It is a non visible element.
1238                 $text = self::stripAbstract($text);
1239
1240                 // Move all spaces out of the tags
1241                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1242                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1243
1244                 // Extract the private images which use data urls since preg has issues with
1245                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1246                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1247
1248                 $extracted = self::extractImagesFromItemBody($text);
1249                 $text = $extracted['body'];
1250                 $saved_image = $extracted['images'];
1251
1252                 // If we find any event code, turn it into an event.
1253                 // After we're finished processing the bbcode we'll
1254                 // replace all of the event code with a reformatted version.
1255
1256                 $ev = Event::fromBBCode($text);
1257
1258                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1259                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1260
1261                 $text = str_replace("<", "&lt;", $text);
1262                 $text = str_replace(">", "&gt;", $text);
1263
1264                 // remove some newlines before the general conversion
1265                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1266                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1267
1268                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1269                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1270
1271                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1272                 if (!$try_oembed) {
1273                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1274                 }
1275
1276                 // Check for [code] text here, before the linefeeds are messed with.
1277                 // The highlighter will unescape and re-escape the content.
1278                 if (strpos($text, '[code=') !== false) {
1279                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1280                 }
1281                 // Convert new line chars to html <br /> tags
1282
1283                 // nlbr seems to be hopelessly messed up
1284                 //      $Text = nl2br($Text);
1285
1286                 // We'll emulate it.
1287
1288                 $text = trim($text);
1289                 $text = str_replace("\r\n", "\n", $text);
1290
1291                 // removing multiplicated newlines
1292                 if (Config::get("system", "remove_multiplicated_lines")) {
1293                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1294                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1295                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1296                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1297                         do {
1298                                 $oldtext = $text;
1299                                 $text = str_replace($search, $replace, $text);
1300                         } while ($oldtext != $text);
1301                 }
1302
1303                 // Set up the parameters for a URL search string
1304                 $URLSearchString = "^\[\]";
1305                 // Set up the parameters for a MAIL search string
1306                 $MAILSearchString = $URLSearchString;
1307
1308                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1309                 if (!$for_plaintext) {
1310                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1311                         // Currently disabled, since the function is too greedy
1312                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1313                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1314                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1315                         if ($simple_html == 7) {
1316                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1317                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1318                         }
1319                 } else {
1320                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1321                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1322                 }
1323
1324
1325                 // Handle attached links or videos
1326                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1327
1328                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1329
1330                 // Remove all hashtag addresses
1331                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1332                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1333                 } elseif ($simple_html == 3) {
1334                         // The ! is converted to @ since Diaspora only understands the @
1335                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1336                                 '@<a href="$2">$3</a>',
1337                                 $text);
1338                 } elseif ($simple_html == 7) {
1339                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1340                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1341                                 $text);
1342                 } elseif (!$simple_html) {
1343                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1344                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1345                                 $text);
1346                 }
1347
1348                 // Bookmarks in red - will be converted to bookmarks in friendica
1349                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1350                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1351                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1352                                         "[bookmark=$1]$2[/bookmark]", $text);
1353
1354                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1355                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1356                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1357                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1358                 }
1359
1360                 if ($simple_html == 5) {
1361                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1362                 }
1363
1364                 // Perform URL Search
1365                 if ($try_oembed) {
1366                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1367                 }
1368
1369                 if ($simple_html == 5) {
1370                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1371                 } else {
1372                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1373                 }
1374
1375                 // Handle Diaspora posts
1376                 $text = preg_replace_callback(
1377                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1378                         function ($match) {
1379                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1380                         }, $text
1381                 );
1382
1383                 $text = preg_replace_callback(
1384                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1385                         function ($match) {
1386                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1387                         }, $text
1388                 );
1389
1390                 // Server independent link to posts and comments
1391                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1392                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1393                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1394
1395                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1396                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1397
1398                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1399                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1400
1401                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1402                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1403                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1404
1405                 // Red compatibility, though the link can't be authenticated on Friendica
1406                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1407
1408
1409                 // we may need to restrict this further if it picks up too many strays
1410                 // link acct:user@host to a webfinger profile redirector
1411
1412                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1413
1414                 // Perform MAIL Search
1415                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1416                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1417
1418                 // leave open the posibility of [map=something]
1419                 // this is replaced in prepare_body() which has knowledge of the item location
1420
1421                 if (strpos($text, '[/map]') !== false) {
1422                         $text = preg_replace_callback(
1423                                 "/\[map\](.*?)\[\/map\]/ism",
1424                                 function ($match) use ($simple_html) {
1425                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1426                                 },
1427                                 $text
1428                         );
1429                 }
1430                 if (strpos($text, '[map=') !== false) {
1431                         $text = preg_replace_callback(
1432                                 "/\[map=(.*?)\]/ism",
1433                                 function ($match) use ($simple_html) {
1434                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1435                                 },
1436                                 $text
1437                         );
1438                 }
1439                 if (strpos($text, '[map]') !== false) {
1440                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1441                 }
1442
1443                 // Check for headers
1444                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1445                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1446                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1447                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1448                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1449                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1450
1451                 // Check for paragraph
1452                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1453
1454                 // Check for bold text
1455                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1456
1457                 // Check for Italics text
1458                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1459
1460                 // Check for Underline text
1461                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1462
1463                 // Check for strike-through text
1464                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1465
1466                 // Check for over-line text
1467                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1468
1469                 // Check for colored text
1470                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1471
1472                 // Check for sized text
1473                 // [size=50] --> font-size: 50px (with the unit).
1474                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1475                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1476
1477                 // Check for centered text
1478                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1479
1480                 // Check for list text
1481                 $text = str_replace("[*]", "<li>", $text);
1482
1483                 // Check for style sheet commands
1484                 $text = preg_replace_callback(
1485                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1486                         function ($match) {
1487                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1488                         },
1489                         $text
1490                 );
1491
1492                 // Check for CSS classes
1493                 $text = preg_replace_callback(
1494                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1495                         function ($match) {
1496                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1497                         },
1498                         $text
1499                 );
1500
1501                 // handle nested lists
1502                 $endlessloop = 0;
1503
1504                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1505                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1506                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1507                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1508                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1509                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1510                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1511                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1512                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1513                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1514                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1515                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1516                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1517                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1518                 }
1519
1520                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1521                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1522                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1523                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1524
1525                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1526                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1527
1528                 $text = str_replace('[hr]', '<hr />', $text);
1529
1530                 // This is actually executed in prepare_body()
1531
1532                 $text = str_replace('[nosmile]', '', $text);
1533
1534                 // Check for font change text
1535                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1536
1537                 // Declare the format for [code] layout
1538
1539                 $CodeLayout = '<code>$1</code>';
1540                 // Check for [code] text
1541                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1542
1543                 // Declare the format for [spoiler] layout
1544                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
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=Author] text
1554
1555                 $t_wrote = L10n::t('$1 wrote:');
1556
1557                 // handle nested quotes
1558                 $endlessloop = 0;
1559                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1560                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1561                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1562                                                  $text);
1563                 }
1564
1565                 // Declare the format for [quote] layout
1566                 $QuoteLayout = '<blockquote>$1</blockquote>';
1567
1568                 // Check for [quote] text
1569                 // handle nested quotes
1570                 $endlessloop = 0;
1571                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1572                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1573                 }
1574
1575                 // Check for [quote=Author] text
1576
1577                 $t_wrote = L10n::t('$1 wrote:');
1578
1579                 // handle nested quotes
1580                 $endlessloop = 0;
1581                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1582                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1583                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1584                                                  $text);
1585                 }
1586
1587
1588                 // [img=widthxheight]image source[/img]
1589                 $text = preg_replace_callback(
1590                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1591                         function ($matches) use ($simple_html) {
1592                                 if (strpos($matches[3], "data:image/") === 0) {
1593                                         return $matches[0];
1594                                 }
1595
1596                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1597                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1598                         },
1599                         $text
1600                 );
1601
1602                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1603                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1604
1605                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1606                         function ($matches) use ($simple_html) {
1607                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1608                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1609                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1610                         },
1611                         $text);
1612
1613                 // Images
1614                 // [img]pathtoimage[/img]
1615                 $text = preg_replace_callback(
1616                         "/\[img\](.*?)\[\/img\]/ism",
1617                         function ($matches) use ($simple_html) {
1618                                 if (strpos($matches[1], "data:image/") === 0) {
1619                                         return $matches[0];
1620                                 }
1621
1622                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1623                                 return "[img]" . $matches[1] . "[/img]";
1624                         },
1625                         $text
1626                 );
1627
1628                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1629                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1630
1631                 // Shared content
1632                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1633                         function ($match) use ($simple_html) {
1634                                 return self::convertShare($match, $simple_html);
1635                         }, $text);
1636
1637                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1638                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1639                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1640
1641                 // Try to Oembed
1642                 if ($try_oembed) {
1643                         $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);
1644                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3).*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1645
1646                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1647                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1648                 } else {
1649                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1650                                                 '<a href="$1" target="_blank">$1</a>', $text);
1651                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1652                                                 '<a href="$1" target="_blank">$1</a>', $text);
1653                 }
1654
1655                 // html5 video and audio
1656
1657
1658                 if ($try_oembed) {
1659                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1660                 } else {
1661                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1662                 }
1663
1664                 // Youtube extensions
1665                 if ($try_oembed) {
1666                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1667                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1668                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1669                 }
1670
1671                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1672                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1673                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1674
1675                 if ($try_oembed) {
1676                         $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);
1677                 } else {
1678                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1679                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1680                 }
1681
1682                 if ($try_oembed) {
1683                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1684                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1685                 }
1686
1687                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1688                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1689
1690                 if ($try_oembed) {
1691                         $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);
1692                 } else {
1693                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1694                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1695                 }
1696
1697                 // oembed tag
1698                 $text = OEmbed::BBCode2HTML($text);
1699
1700                 // Avoid triple linefeeds through oembed
1701                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1702
1703                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1704                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1705                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1706                 // start which is always required). Allow desc with a missing summary for compatibility.
1707
1708                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1709                         $sub = Event::getHTML($ev, $simple_html);
1710
1711                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1712                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1713                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1714                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1715                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1716                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1717                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1718                 }
1719
1720                 // Replace non graphical smilies for external posts
1721                 if ($simple_html) {
1722                         $text = Smilies::replace($text, false, true);
1723                 }
1724
1725                 // Replace inline code blocks
1726                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1727                         function ($match) use ($simple_html) {
1728                                 $return = '<key>' . $match[1] . '</key>';
1729                                 // Use <code> for Diaspora inline code blocks
1730                                 if ($simple_html === 3) {
1731                                         $return = '<code>' . $match[1] . '</code>';
1732                                 }
1733                                 return $return;
1734                         }
1735                 , $text);
1736
1737                 // Unhide all [noparse] contained bbtags unspacefying them
1738                 // and triming the [noparse] tag.
1739
1740                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1741                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1742                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1743
1744                 /// @todo What is the meaning of these lines?
1745                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1746                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1747
1748                 // Currently deactivated, it made problems with " inside of alt texts.
1749                 //$text = preg_replace('/\&quot\;/', '"', $text);
1750
1751                 // fix any escaped ampersands that may have been converted into links
1752                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1753
1754                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1755                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1756                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1757                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1758
1759                 // sanitize href attributes (only whitelisted protocols URLs)
1760                 // default value for backward compatibility
1761                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1762
1763                 // Always allowed protocol even if config isn't set or not including it
1764                 $allowed_link_protocols[] = 'http';
1765                 $allowed_link_protocols[] = 'redir/';
1766
1767                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1768                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1769
1770                 if ($saved_image) {
1771                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1772                 }
1773
1774                 // Clean up the HTML by loading and saving the HTML with the DOM.
1775                 // Bad structured html can break a whole page.
1776                 // For performance reasons do it only with ativated item cache or at export.
1777                 if (!$try_oembed || (get_itemcachepath() != "")) {
1778                         $doc = new DOMDocument();
1779                         $doc->preserveWhiteSpace = false;
1780
1781                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1782
1783                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1784                         $encoding = '<?xml encoding="UTF-8">';
1785                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1786                         $doc->encoding = 'UTF-8';
1787                         $text = $doc->saveHTML();
1788                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1789
1790                         $text = str_replace('<br></li>', '</li>', $text);
1791
1792                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1793                 }
1794
1795                 // Clean up some useless linebreaks in lists
1796                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1797                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1798                 //$Text = str_replace('</li><br />', '</li>', $Text);
1799                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1800                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1801
1802                 Addon::callHooks('bbcode', $text);
1803
1804                 return trim($text);
1805         }
1806
1807         /**
1808          * @brief Strips the "abstract" tag from the provided text
1809          *
1810          * @param string $text The text with BBCode
1811          * @return string The same text - but without "abstract" element
1812          */
1813         public static function stripAbstract($text)
1814         {
1815                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1816                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1817
1818                 return $text;
1819         }
1820
1821         /**
1822          * @brief Returns the value of the "abstract" element
1823          *
1824          * @param string $text The text that maybe contains the element
1825          * @param string $addon The addon for which the abstract is meant for
1826          * @return string The abstract
1827          */
1828         public static function getAbstract($text, $addon = "")
1829         {
1830                 $abstract = "";
1831                 $abstracts = [];
1832                 $addon = strtolower($addon);
1833
1834                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1835                         foreach ($results AS $result) {
1836                                 $abstracts[strtolower($result[1])] = $result[2];
1837                         }
1838                 }
1839
1840                 if (isset($abstracts[$addon])) {
1841                         $abstract = $abstracts[$addon];
1842                 }
1843
1844                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1845                         $abstract = $result[1];
1846                 }
1847
1848                 return $abstract;
1849         }
1850
1851         /**
1852          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1853          *
1854          * @param array $match Matching values for the callback
1855          * @return string Replaced mention
1856          */
1857         private static function bbCodeMention2DiasporaCallback($match)
1858         {
1859                 $contact = Contact::getDetailsByURL($match[3]);
1860
1861                 if (empty($contact['addr'])) {
1862                         $contact = Probe::uri($match[3]);
1863                 }
1864
1865                 if (empty($contact['addr'])) {
1866                         return $match[0];
1867                 }
1868
1869                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1870                 return $mention;
1871         }
1872
1873         /**
1874          * @brief Converts a BBCode text into Markdown
1875          *
1876          * This function converts a BBCode item body to be sent to Markdown-enabled
1877          * systems like Diaspora and Libertree
1878          *
1879          * @param string $text
1880          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1881          * @return string
1882          */
1883         public static function toMarkdown($text, $for_diaspora = true)
1884         {
1885                 $a = self::getApp();
1886
1887                 $original_text = $text;
1888
1889                 // Since Diaspora is creating a summary for links, this function removes them before posting
1890                 if ($for_diaspora) {
1891                         $text = self::removeShareInformation($text);
1892                 }
1893
1894                 /**
1895                  * Transform #tags, strip off the [url] and replace spaces with underscore
1896                  */
1897                 $url_search_string = "^\[\]";
1898                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1899                         function ($matches) {
1900                                 return '#' . str_replace(' ', '_', $matches[2]);
1901                         },
1902                         $text
1903                 );
1904
1905                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1906                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1907
1908                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
1909                 $codeblocks = [];
1910
1911                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
1912                         function ($matches) use (&$codeblocks) {
1913                                 $return = $matches[0];
1914                                 if (strpos($matches[2], "\n") !== false) {
1915                                         $return = '#codeblock-' . count($codeblocks) . '#';
1916
1917                                         $prefix = '````' . $matches[1] . PHP_EOL;
1918                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
1919                                 }
1920                                 return $return;
1921                         },
1922                         $text
1923                 );
1924
1925                 // Convert it to HTML - don't try oembed
1926                 if ($for_diaspora) {
1927                         $text = self::convert($text, false, 3);
1928
1929                         // Add all tags that maybe were removed
1930                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1931                                 $tagline = "";
1932                                 foreach ($tags[2] as $tag) {
1933                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1934                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1935                                                 $tagline .= '#' . $tag . ' ';
1936                                         }
1937                                 }
1938                                 $text = $text . " " . $tagline;
1939                         }
1940                 } else {
1941                         $text = self::convert($text, false, 4);
1942                 }
1943
1944                 // mask some special HTML chars from conversation to markdown
1945                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1946
1947                 // If a link is followed by a quote then there should be a newline before it
1948                 // Maybe we should make this newline at every time before a quote.
1949                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1950
1951                 $stamp1 = microtime(true);
1952
1953                 // Now convert HTML to Markdown
1954                 $converter = new HtmlConverter();
1955                 $text = $converter->convert($text);
1956
1957                 // unmask the special chars back to HTML
1958                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1959
1960                 $a->save_timestamp($stamp1, "parser");
1961
1962                 // Libertree has a problem with escaped hashtags.
1963                 $text = str_replace(['\#'], ['#'], $text);
1964
1965                 // Remove any leading or trailing whitespace, as this will mess up
1966                 // the Diaspora signature verification and cause the item to disappear
1967                 $text = trim($text);
1968
1969                 if ($for_diaspora) {
1970                         $url_search_string = "^\[\]";
1971                         $text = preg_replace_callback(
1972                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
1973                                 ['self', 'bbCodeMention2DiasporaCallback'],
1974                                 $text
1975                         );
1976                 }
1977
1978                 // Restore code blocks
1979                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1980                         function ($matches) use ($codeblocks) {
1981                                 $return = '';
1982                                 if (isset($codeblocks[intval($matches[1])])) {
1983                                         $return = $codeblocks[$matches[1]];
1984                                 }
1985                                 return $return;
1986                         },
1987                         $text
1988                 );
1989
1990                 Addon::callHooks('bb2diaspora', $text);
1991
1992                 return $text;
1993         }
1994 }