src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5 namespace Friendica\Content\Text;
   6
   7 use Friendica\App;
   8 use Friendica\Content\Text\Plaintext;
   9 use Friendica\Core\Config;
  10 use Friendica\Core\L10n;
  11 use Friendica\Core\PConfig;
  12 use Friendica\Core\System;
  13 use Friendica\Object\Image;
  14 use Friendica\Util\ParseUrl;
  15 use Friendica\Content\OEmbed;
  16
  17 require_once "include/bbcode.php";
  18 require_once "include/html2plain.php";
  19
  20 class BBCode
  21 {
  22         /**
  23          * @brief Fetches attachment data that were generated the old way
  24          *
  25          * @param string $body Message body
  26          * @return array
  27          * 'type' -> Message type ("link", "video", "photo")
  28          * 'text' -> Text before the shared message
  29          * 'after' -> Text after the shared message
  30          * 'image' -> Preview image of the message
  31          * 'url' -> Url to the attached message
  32          * 'title' -> Title of the attachment
  33          * 'description' -> Description of the attachment
  34          */
  35         private static function getOldAttachmentData($body)
  36         {
  37                 $post = [];
  38
  39                 // Simplify image codes
  40                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  41
  42                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  43                         foreach ($attached as $data) {
  44                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  45                                         continue;
  46                                 }
  47
  48                                 $post["type"] = substr($data[1], 5);
  49
  50                                 $pos = strpos($body, $data[0]);
  51                                 if ($pos > 0) {
  52                                         $post["text"] = trim(substr($body, 0, $pos));
  53                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  54                                 } else {
  55                                         $post["text"] = trim(str_replace($data[0], "", $body));
  56                                 }
  57
  58                                 $attacheddata = $data[2];
  59
  60                                 $URLSearchString = "^\[\]";
  61
  62                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  63
  64                                         $picturedata = Image::getInfoFromURL($matches[1]);
  65
  66                                         if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  67                                                 $post["image"] = $matches[1];
  68                                         } else {
  69                                                 $post["preview"] = $matches[1];
  70                                         }
  71                                 }
  72
  73                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  74                                         $post["url"] = $matches[1];
  75                                         $post["title"] = $matches[2];
  76                                 }
  77                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  78                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  79                                         $post["url"] = $matches[1];
  80                                 }
  81
  82                                 // Search for description
  83                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  84                                         $post["description"] = $matches[1];
  85                                 }
  86                         }
  87                 }
  88                 return $post;
  89         }
  90
  91         /**
  92          * @brief Fetches attachment data that were generated with the "attachment" element
  93          *
  94          * @param string $body Message body
  95          * @return array
  96          * 'type' -> Message type ("link", "video", "photo")
  97          * 'text' -> Text before the shared message
  98          * 'after' -> Text after the shared message
  99          * 'image' -> Preview image of the message
 100          * 'url' -> Url to the attached message
 101          * 'title' -> Title of the attachment
 102          * 'description' -> Description of the attachment
 103          */
 104         public static function getAttachmentData($body)
 105         {
 106                 $data = [];
 107
 108                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 109                         return self::getOldAttachmentData($body);
 110                 }
 111
 112                 $attributes = $match[2];
 113
 114                 $data["text"] = trim($match[1]);
 115
 116                 $type = "";
 117                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 118                 if (x($matches, 1)) {
 119                         $type = strtolower($matches[1]);
 120                 }
 121
 122                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 123                 if (x($matches, 1)) {
 124                         $type = strtolower($matches[1]);
 125                 }
 126
 127                 if ($type == "") {
 128                         return [];
 129                 }
 130
 131                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 132                         return [];
 133                 }
 134
 135                 if ($type != "") {
 136                         $data["type"] = $type;
 137                 }
 138
 139                 $url = "";
 140                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 141                 if (x($matches, 1)) {
 142                         $url = $matches[1];
 143                 }
 144
 145                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 146                 if (x($matches, 1)) {
 147                         $url = $matches[1];
 148                 }
 149
 150                 if ($url != "") {
 151                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 152                 }
 153
 154                 $title = "";
 155                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 156                 if (x($matches, 1)) {
 157                         $title = $matches[1];
 158                 }
 159
 160                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 161                 if (x($matches, 1)) {
 162                         $title = $matches[1];
 163                 }
 164
 165                 if ($title != "") {
 166                         $title = bbcode(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, false, true);
 167                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 168                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 169                         $data["title"] = $title;
 170                 }
 171
 172                 $image = "";
 173                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 174                 if (x($matches, 1)) {
 175                         $image = $matches[1];
 176                 }
 177
 178                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 179                 if (x($matches, 1)) {
 180                         $image = $matches[1];
 181                 }
 182
 183                 if ($image != "") {
 184                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 185                 }
 186
 187                 $preview = "";
 188                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 189                 if (x($matches, 1)) {
 190                         $preview = $matches[1];
 191                 }
 192
 193                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 194                 if (x($matches, 1)) {
 195                         $preview = $matches[1];
 196                 }
 197
 198                 if ($preview != "") {
 199                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 200                 }
 201
 202                 $data["description"] = trim($match[3]);
 203
 204                 $data["after"] = trim($match[4]);
 205
 206                 return $data;
 207         }
 208
 209         public static function getAttachedData($body, $item = [])
 210         {
 211                 /*
 212                 - text:
 213                 - type: link, video, photo
 214                 - title:
 215                 - url:
 216                 - image:
 217                 - description:
 218                 - (thumbnail)
 219                 */
 220
 221                 $has_title = !empty($item['title']);
 222                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 223                 $post = self::getAttachmentData($body);
 224
 225                 // if nothing is found, it maybe having an image.
 226                 if (!isset($post["type"])) {
 227                         // Simplify image codes
 228                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 229
 230                         $URLSearchString = "^\[\]";
 231                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 232                                 if ((count($pictures) == 1) && !$has_title) {
 233                                         // Checking, if the link goes to a picture
 234                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 235
 236                                         // Workaround:
 237                                         // Sometimes photo posts to the own album are not detected at the start.
 238                                         // So we seem to cannot use the cache for these cases. That's strange.
 239                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 240                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 241                                         }
 242
 243                                         if ($data["type"] == "photo") {
 244                                                 $post["type"] = "photo";
 245                                                 if (isset($data["images"][0])) {
 246                                                         $post["image"] = $data["images"][0]["src"];
 247                                                         $post["url"] = $data["url"];
 248                                                 } else {
 249                                                         $post["image"] = $data["url"];
 250                                                 }
 251
 252                                                 $post["preview"] = $pictures[0][2];
 253                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 254                                         } else {
 255                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 256                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 257                                                         $post["type"] = "photo";
 258                                                         $post["image"] = $pictures[0][1];
 259                                                         $post["preview"] = $pictures[0][2];
 260                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 261                                                 }
 262                                         }
 263                                 } elseif (count($pictures) > 0) {
 264                                         $post["type"] = "link";
 265                                         $post["url"] = $plink;
 266                                         $post["image"] = $pictures[0][2];
 267                                         $post["text"] = $body;
 268                                 }
 269                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 270                                 if ((count($pictures) == 1) && !$has_title) {
 271                                         $post["type"] = "photo";
 272                                         $post["image"] = $pictures[0][1];
 273                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 274                                 } elseif (count($pictures) > 0) {
 275                                         $post["type"] = "link";
 276                                         $post["url"] = $plink;
 277                                         $post["image"] = $pictures[0][1];
 278                                         $post["text"] = $body;
 279                                 }
 280                         }
 281
 282                         // Test for the external links
 283                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 284                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 285
 286                         $links = array_merge($links1, $links2);
 287
 288                         // If there is only a single one, then use it.
 289                         // This should cover link posts via API.
 290                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 291                                 $post["type"] = "link";
 292                                 $post["text"] = trim($body);
 293                                 $post["url"] = $links[0][1];
 294                         }
 295
 296                         // Now count the number of external media links
 297                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 298                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 299                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 300                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 301
 302                         // Add them to the other external links
 303                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 304
 305                         // Are there more than one?
 306                         if (count($links) > 1) {
 307                                 // The post will be the type "text", which means a blog post
 308                                 unset($post["type"]);
 309                                 $post["url"] = $plink;
 310                         }
 311
 312                         if (!isset($post["type"])) {
 313                                 $post["type"] = "text";
 314                                 $post["text"] = trim($body);
 315                         }
 316                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 317                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 318
 319                         if (isset($data["images"][0])) {
 320                                 $post["image"] = $data["images"][0]["src"];
 321                         }
 322                 }
 323
 324                 return $post;
 325         }
 326
 327         /**
 328          * @brief Convert a message into plaintext for connectors to other networks
 329          *
 330          * @param array $b The message array that is about to be posted
 331          * @param int $limit The maximum number of characters when posting to that network
 332          * @param bool $includedlinks Has an attached link to be included into the message?
 333          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 334          * @param string $target_network Name of the network where the post should go to.
 335          *
 336          * @return string The converted message
 337          */
 338         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 339         {
 340                 // Remove the hash tags
 341                 $URLSearchString = "^\[\]";
 342                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 343
 344                 // Add an URL element if the text contains a raw link
 345                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 346
 347                 // Remove the abstract
 348                 $body = remove_abstract($body);
 349
 350                 // At first look at data that is attached via "type-..." stuff
 351                 // This will hopefully replaced with a dedicated bbcode later
 352                 //$post = self::getAttachedData($b["body"]);
 353                 $post = self::getAttachedData($body, $b);
 354
 355                 if (($b["title"] != "") && ($post["text"] != "")) {
 356                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 357                 } elseif ($b["title"] != "") {
 358                         $post["text"] = trim($b["title"]);
 359                 }
 360
 361                 $abstract = "";
 362
 363                 // Fetch the abstract from the given target network
 364                 if ($target_network != "") {
 365                         $default_abstract = fetch_abstract($b["body"]);
 366                         $abstract = fetch_abstract($b["body"], $target_network);
 367
 368                         // If we post to a network with no limit we only fetch
 369                         // an abstract exactly for this network
 370                         if (($limit == 0) && ($abstract == $default_abstract)) {
 371                                 $abstract = "";
 372                         }
 373                 } else {// Try to guess the correct target network
 374                         switch ($htmlmode) {
 375                                 case 8:
 376                                         $abstract = fetch_abstract($b["body"], NETWORK_TWITTER);
 377                                         break;
 378                                 case 7:
 379                                         $abstract = fetch_abstract($b["body"], NETWORK_STATUSNET);
 380                                         break;
 381                                 case 6:
 382                                         $abstract = fetch_abstract($b["body"], NETWORK_APPNET);
 383                                         break;
 384                                 default: // We don't know the exact target.
 385                                         // We fetch an abstract since there is a posting limit.
 386                                         if ($limit > 0) {
 387                                                 $abstract = fetch_abstract($b["body"]);
 388                                         }
 389                         }
 390                 }
 391
 392                 if ($abstract != "") {
 393                         $post["text"] = $abstract;
 394
 395                         if ($post["type"] == "text") {
 396                                 $post["type"] = "link";
 397                                 $post["url"] = $b["plink"];
 398                         }
 399                 }
 400
 401                 $html = bbcode($post["text"].$post["after"], false, false, $htmlmode);
 402                 $msg = html2plain($html, 0, true);
 403                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 404
 405                 $link = "";
 406                 if ($includedlinks) {
 407                         if ($post["type"] == "link") {
 408                                 $link = $post["url"];
 409                         } elseif ($post["type"] == "text") {
 410                                 $link = $post["url"];
 411                         } elseif ($post["type"] == "video") {
 412                                 $link = $post["url"];
 413                         } elseif ($post["type"] == "photo") {
 414                                 $link = $post["image"];
 415                         }
 416
 417                         if (($msg == "") && isset($post["title"])) {
 418                                 $msg = trim($post["title"]);
 419                         }
 420
 421                         if (($msg == "") && isset($post["description"])) {
 422                                 $msg = trim($post["description"]);
 423                         }
 424
 425                         // If the link is already contained in the post, then it neeedn't to be added again
 426                         // But: if the link is beyond the limit, then it has to be added.
 427                         if (($link != "") && strstr($msg, $link)) {
 428                                 $pos = strpos($msg, $link);
 429
 430                                 // Will the text be shortened in the link?
 431                                 // Or is the link the last item in the post?
 432                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 433                                         $msg = trim(str_replace($link, "", $msg));
 434                                 } elseif (($limit == 0) || ($pos < $limit)) {
 435                                         // The limit has to be increased since it will be shortened - but not now
 436                                         // Only do it with Twitter (htmlmode = 8)
 437                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 438                                                 $limit = $limit - 23 + strlen($link);
 439                                         }
 440
 441                                         $link = "";
 442
 443                                         if ($post["type"] == "text") {
 444                                                 unset($post["url"]);
 445                                         }
 446                                 }
 447                         }
 448                 }
 449
 450                 if ($limit > 0) {
 451                         // Reduce multiple spaces
 452                         // When posted to a network with limited space, we try to gain space where possible
 453                         while (strpos($msg, "  ") !== false) {
 454                                 $msg = str_replace("  ", " ", $msg);
 455                         }
 456
 457                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 458                         if (iconv_strlen($link, "UTF-8") > 0) {
 459                                 $limit = $limit - 23;
 460                         }
 461
 462                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 463                                 if (($post["type"] == "text") && isset($post["url"])) {
 464                                         $post["url"] = $b["plink"];
 465                                 } elseif (!isset($post["url"])) {
 466                                         $limit = $limit - 23;
 467                                         $post["url"] = $b["plink"];
 468                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 469                                 //} elseif (strpos($b["body"], "[share") !== false) {
 470                                 //      $post["url"] = $b["plink"];
 471                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 472                                         $post["url"] = $b["plink"];
 473                                 }
 474                                 $msg = Plaintext::shorten($msg, $limit);
 475                         }
 476                 }
 477
 478                 $post["text"] = trim($msg);
 479
 480                 return($post);
 481         }
 482
 483         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 484         {
 485                 // Suppress "view full size"
 486                 if (intval(Config::get('system', 'no_view_full_size'))) {
 487                         $include_link = false;
 488                 }
 489
 490                 // Picture addresses can contain special characters
 491                 $s = htmlspecialchars_decode($srctext);
 492
 493                 $matches = null;
 494                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 495                 if ($c) {
 496                         foreach ($matches as $mtch) {
 497                                 logger('scale_external_image: ' . $mtch[1]);
 498
 499                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 500                                 if (stristr($mtch[1], $hostname)) {
 501                                         continue;
 502                                 }
 503
 504                                 // $scale_replace, if passed, is an array of two elements. The
 505                                 // first is the name of the full-size image. The second is the
 506                                 // name of a remote, scaled-down version of the full size image.
 507                                 // This allows Friendica to display the smaller remote image if
 508                                 // one exists, while still linking to the full-size image
 509                                 if ($scale_replace) {
 510                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 511                                 } else {
 512                                         $scaled = $mtch[1];
 513                                 }
 514                                 $i = OEmbed::fetchURL($scaled);
 515                                 if (! $i) {
 516                                         return $srctext;
 517                                 }
 518
 519                                 // guess mimetype from headers or filename
 520                                 $type = Image::guessType($mtch[1], true);
 521
 522                                 if ($i) {
 523                                         $Image = new Image($i, $type);
 524                                         if ($Image->isValid()) {
 525                                                 $orig_width = $Image->getWidth();
 526                                                 $orig_height = $Image->getHeight();
 527
 528                                                 if ($orig_width > 640 || $orig_height > 640) {
 529                                                         $Image->scaleDown(640);
 530                                                         $new_width = $Image->getWidth();
 531                                                         $new_height = $Image->getHeight();
 532                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 533                                                         $s = str_replace(
 534                                                                 $mtch[0],
 535                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 536                                                                 . "\n" . (($include_link)
 537                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 538                                                                         : ''),
 539                                                                 $s
 540                                                         );
 541                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 542                                                 }
 543                                         }
 544                                 }
 545                         }
 546                 }
 547
 548                 // replace the special char encoding
 549                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 550                 return $s;
 551         }
 552 }