include/plaintext.php

   1 <?php
   2 /**
   3  * @file include/plaintext.php
   4  */
   5 use Friendica\App;
   6 use Friendica\Core\PConfig;
   7 use Friendica\Object\Image;
   8 use Friendica\Util\ParseUrl;
   9
  10 require_once "include/bbcode.php";
  11 require_once "include/html2plain.php";
  12 require_once "include/network.php";
  13
  14 /**
  15  * @brief Fetches attachment data that were generated the old way
  16  *
  17  * @param string $body Message body
  18  * @return array
  19  * 'type' -> Message type ("link", "video", "photo")
  20  * 'text' -> Text before the shared message
  21  * 'after' -> Text after the shared message
  22  * 'image' -> Preview image of the message
  23  * 'url' -> Url to the attached message
  24  * 'title' -> Title of the attachment
  25  * 'description' -> Description of the attachment
  26  */
  27 function get_old_attachment_data($body) {
  28
  29         $post = [];
  30
  31         // Simplify image codes
  32         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  33
  34         if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism",$body, $attached,  PREG_SET_ORDER)) {
  35                 foreach ($attached AS $data) {
  36                         if (!in_array($data[1], ["type-link", "type-video", "type-photo"]))
  37                                 continue;
  38
  39                         $post["type"] = substr($data[1], 5);
  40
  41                         $pos = strpos($body, $data[0]);
  42                         if ($pos > 0) {
  43                                 $post["text"] = trim(substr($body, 0, $pos));
  44                                 $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  45                         } else
  46                                 $post["text"] = trim(str_replace($data[0], "", $body));
  47
  48                         $attacheddata = $data[2];
  49
  50                         $URLSearchString = "^\[\]";
  51
  52                         if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  53
  54                                 $picturedata = Image::getInfoFromURL($matches[1]);
  55
  56                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1]))
  57                                         $post["image"] = $matches[1];
  58                                 else
  59                                         $post["preview"] = $matches[1];
  60                         }
  61
  62                         if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  63                                 $post["url"] = $matches[1];
  64                                 $post["title"] = $matches[2];
  65                         }
  66                         if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  67                                 && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  68                                 $post["url"] = $matches[1];
  69                         }
  70
  71                         // Search for description
  72                         if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches))
  73                                 $post["description"] = $matches[1];
  74
  75                 }
  76         }
  77         return $post;
  78 }
  79
  80 /**
  81  * @brief Fetches attachment data that were generated with the "attachment" element
  82  *
  83  * @param string $body Message body
  84  * @return array
  85  * 'type' -> Message type ("link", "video", "photo")
  86  * 'text' -> Text before the shared message
  87  * 'after' -> Text after the shared message
  88  * 'image' -> Preview image of the message
  89  * 'url' -> Url to the attached message
  90  * 'title' -> Title of the attachment
  91  * 'description' -> Description of the attachment
  92  */
  93 function get_attachment_data($body)
  94 {
  95         $data = [];
  96
  97         if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
  98                 return get_old_attachment_data($body);
  99         }
 100
 101         $attributes = $match[2];
 102
 103         $data["text"] = trim($match[1]);
 104
 105         $type = "";
 106         preg_match("/type='(.*?)'/ism", $attributes, $matches);
 107         if (x($matches, 1)) {
 108                 $type = strtolower($matches[1]);
 109         }
 110
 111         preg_match('/type="(.*?)"/ism', $attributes, $matches);
 112         if (x($matches, 1)) {
 113                 $type = strtolower($matches[1]);
 114         }
 115
 116         if ($type == "") {
 117                 return [];
 118         }
 119
 120         if (!in_array($type, ["link", "audio", "photo", "video"])) {
 121                 return [];
 122         }
 123
 124         if ($type != "") {
 125                 $data["type"] = $type;
 126         }
 127
 128         $url = "";
 129         preg_match("/url='(.*?)'/ism", $attributes, $matches);
 130         if (x($matches, 1)) {
 131                 $url = $matches[1];
 132         }
 133
 134         preg_match('/url="(.*?)"/ism', $attributes, $matches);
 135         if (x($matches, 1)) {
 136                 $url = $matches[1];
 137         }
 138
 139         if ($url != "") {
 140                 $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 141         }
 142
 143         $title = "";
 144         preg_match("/title='(.*?)'/ism", $attributes, $matches);
 145         if (x($matches, 1)) {
 146                 $title = $matches[1];
 147         }
 148
 149         preg_match('/title="(.*?)"/ism', $attributes, $matches);
 150         if (x($matches, 1)) {
 151                 $title = $matches[1];
 152         }
 153
 154         if ($title != "") {
 155                 $title = bbcode(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, false, true);
 156                 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 157                 $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 158                 $data["title"] = $title;
 159         }
 160
 161         $image = "";
 162         preg_match("/image='(.*?)'/ism", $attributes, $matches);
 163         if (x($matches, 1)) {
 164                 $image = $matches[1];
 165         }
 166
 167         preg_match('/image="(.*?)"/ism', $attributes, $matches);
 168         if (x($matches, 1)) {
 169                 $image = $matches[1];
 170         }
 171
 172         if ($image != "") {
 173                 $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 174         }
 175
 176         $preview = "";
 177         preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 178         if (x($matches, 1)) {
 179                 $preview = $matches[1];
 180         }
 181
 182         preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 183         if (x($matches, 1)) {
 184                 $preview = $matches[1];
 185         }
 186
 187         if ($preview != "") {
 188                 $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 189         }
 190
 191         $data["description"] = trim($match[3]);
 192
 193         $data["after"] = trim($match[4]);
 194
 195         return $data;
 196 }
 197
 198 function get_attached_data($body, $item = []) {
 199 /*
 200  - text:
 201  - type: link, video, photo
 202  - title:
 203  - url:
 204  - image:
 205  - description:
 206  - (thumbnail)
 207 */
 208
 209         $has_title = !empty($item['title']);
 210         $plink = (!empty($item['plink']) ? $item['plink'] : '');
 211         $post = get_attachment_data($body);
 212
 213         // if nothing is found, it maybe having an image.
 214         if (!isset($post["type"])) {
 215                 // Simplify image codes
 216                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 217
 218                 $URLSearchString = "^\[\]";
 219                 if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures,  PREG_SET_ORDER)) {
 220                         if ((count($pictures) == 1) && !$has_title) {
 221                                 // Checking, if the link goes to a picture
 222                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 223
 224                                 // Workaround:
 225                                 // Sometimes photo posts to the own album are not detected at the start.
 226                                 // So we seem to cannot use the cache for these cases. That's strange.
 227                                 if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/"))
 228                                         $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 229
 230                                 if ($data["type"] == "photo") {
 231                                         $post["type"] = "photo";
 232                                         if (isset($data["images"][0])) {
 233                                                 $post["image"] = $data["images"][0]["src"];
 234                                                 $post["url"] = $data["url"];
 235                                         } else
 236                                                 $post["image"] = $data["url"];
 237
 238                                         $post["preview"] = $pictures[0][2];
 239                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 240                                 } else {
 241                                         $imgdata = Image::getInfoFromURL($pictures[0][1]);
 242                                         if (substr($imgdata["mime"], 0, 6) == "image/") {
 243                                                 $post["type"] = "photo";
 244                                                 $post["image"] = $pictures[0][1];
 245                                                 $post["preview"] = $pictures[0][2];
 246                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 247                                         }
 248                                 }
 249                         } elseif (count($pictures) > 0) {
 250                                 $post["type"] = "link";
 251                                 $post["url"] = $plink;
 252                                 $post["image"] = $pictures[0][2];
 253                                 $post["text"] = $body;
 254                         }
 255                 } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures,  PREG_SET_ORDER)) {
 256                         if ((count($pictures) == 1) && !$has_title) {
 257                                 $post["type"] = "photo";
 258                                 $post["image"] = $pictures[0][1];
 259                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 260                         } elseif (count($pictures) > 0) {
 261                                 $post["type"] = "link";
 262                                 $post["url"] = $plink;
 263                                 $post["image"] = $pictures[0][1];
 264                                 $post["text"] = $body;
 265                         }
 266                 }
 267
 268                 // Test for the external links
 269                 preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1,  PREG_SET_ORDER);
 270                 preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2,  PREG_SET_ORDER);
 271
 272                 $links = array_merge($links1, $links2);
 273
 274                 // If there is only a single one, then use it.
 275                 // This should cover link posts via API.
 276                 if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 277                         $post["type"] = "link";
 278                         $post["text"] = trim($body);
 279                         $post["url"] = $links[0][1];
 280                 }
 281
 282                 // Now count the number of external media links
 283                 preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1,  PREG_SET_ORDER);
 284                 preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2,  PREG_SET_ORDER);
 285                 preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3,  PREG_SET_ORDER);
 286                 preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4,  PREG_SET_ORDER);
 287
 288                 // Add them to the other external links
 289                 $links = array_merge($links, $links1, $links2, $links3, $links4);
 290
 291                 // Are there more than one?
 292                 if (count($links) > 1) {
 293                         // The post will be the type "text", which means a blog post
 294                         unset($post["type"]);
 295                         $post["url"] = $plink;
 296                 }
 297
 298                 if (!isset($post["type"])) {
 299                         $post["type"] = "text";
 300                         $post["text"] = trim($body);
 301                 }
 302         } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 303                 $data = ParseUrl::getSiteinfoCached($post["url"], true);
 304
 305                 if (isset($data["images"][0]))
 306                         $post["image"] = $data["images"][0]["src"];
 307         }
 308
 309         return $post;
 310 }
 311
 312 /**
 313  * Shortens message
 314  *
 315  * @param type $msg
 316  * @param type $limit
 317  * @return type
 318  *
 319  * @todo For Twitter URLs aren't shortened, but they have to be calculated as if.
 320  */
 321 function shortenmsg($msg, $limit)
 322 {
 323         $lines = explode("\n", $msg);
 324         $msg = "";
 325         $recycle = html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 326         $ellipsis = html_entity_decode("&#x2026;", ENT_QUOTES, 'UTF-8');
 327         foreach ($lines AS $row => $line) {
 328                 if (iconv_strlen(trim($msg . "\n" . $line), "UTF-8") <= $limit) {
 329                         $msg = trim($msg . "\n" . $line);
 330                 } elseif (($msg == "") || (($row == 1) && (substr($msg, 0, 4) == $recycle))) {
 331                         // Is the new message empty by now or is it a reshared message?
 332                         $msg = iconv_substr(iconv_substr(trim($msg . "\n" . $line), 0, $limit, "UTF-8"), 0, -3, "UTF-8") . $ellipsis;
 333                 } else {
 334                         break;
 335                 }
 336         }
 337
 338         return $msg;
 339 }
 340
 341 /**
 342  * @brief Convert a message into plaintext for connectors to other networks
 343  *
 344  * @param array $b The message array that is about to be posted
 345  * @param int $limit The maximum number of characters when posting to that network
 346  * @param bool $includedlinks Has an attached link to be included into the message?
 347  * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 348  * @param string $target_network Name of the network where the post should go to.
 349  *
 350  * @return string The converted message
 351  */
 352 function plaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "") {
 353
 354         // Remove the hash tags
 355         $URLSearchString = "^\[\]";
 356         $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 357
 358         // Add an URL element if the text contains a raw link
 359         $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 360
 361         // Remove the abstract
 362         $body = remove_abstract($body);
 363
 364         // At first look at data that is attached via "type-..." stuff
 365         // This will hopefully replaced with a dedicated bbcode later
 366         //$post = get_attached_data($b["body"]);
 367         $post = get_attached_data($body, $b);
 368
 369         if (($b["title"] != "") && ($post["text"] != ""))
 370                 $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 371         elseif ($b["title"] != "")
 372                 $post["text"] = trim($b["title"]);
 373
 374         $abstract = "";
 375
 376         // Fetch the abstract from the given target network
 377         if ($target_network != "") {
 378                 $default_abstract = fetch_abstract($b["body"]);
 379                 $abstract = fetch_abstract($b["body"], $target_network);
 380
 381                 // If we post to a network with no limit we only fetch
 382                 // an abstract exactly for this network
 383                 if (($limit == 0) && ($abstract == $default_abstract))
 384                         $abstract = "";
 385
 386         } else // Try to guess the correct target network
 387                 switch ($htmlmode) {
 388                         case 8:
 389                                 $abstract = fetch_abstract($b["body"], NETWORK_TWITTER);
 390                                 break;
 391                         case 7:
 392                                 $abstract = fetch_abstract($b["body"], NETWORK_STATUSNET);
 393                                 break;
 394                         case 6:
 395                                 $abstract = fetch_abstract($b["body"], NETWORK_APPNET);
 396                                 break;
 397                         default: // We don't know the exact target.
 398                                  // We fetch an abstract since there is a posting limit.
 399                                 if ($limit > 0)
 400                                         $abstract = fetch_abstract($b["body"]);
 401                 }
 402
 403         if ($abstract != "") {
 404                 $post["text"] = $abstract;
 405
 406                 if ($post["type"] == "text") {
 407                         $post["type"] = "link";
 408                         $post["url"] = $b["plink"];
 409                 }
 410         }
 411
 412         $html = bbcode($post["text"].$post["after"], false, false, $htmlmode);
 413         $msg = html2plain($html, 0, true);
 414         $msg = trim(html_entity_decode($msg,ENT_QUOTES,'UTF-8'));
 415
 416         $link = "";
 417         if ($includedlinks) {
 418                 if ($post["type"] == "link")
 419                         $link = $post["url"];
 420                 elseif ($post["type"] == "text")
 421                         $link = $post["url"];
 422                 elseif ($post["type"] == "video")
 423                         $link = $post["url"];
 424                 elseif ($post["type"] == "photo")
 425                         $link = $post["image"];
 426
 427                 if (($msg == "") && isset($post["title"]))
 428                         $msg = trim($post["title"]);
 429
 430                 if (($msg == "") && isset($post["description"]))
 431                         $msg = trim($post["description"]);
 432
 433                 // If the link is already contained in the post, then it neeedn't to be added again
 434                 // But: if the link is beyond the limit, then it has to be added.
 435                 if (($link != "") && strstr($msg, $link)) {
 436                         $pos = strpos($msg, $link);
 437
 438                         // Will the text be shortened in the link?
 439                         // Or is the link the last item in the post?
 440                         if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg))))
 441                                 $msg = trim(str_replace($link, "", $msg));
 442                         elseif (($limit == 0) || ($pos < $limit)) {
 443                                 // The limit has to be increased since it will be shortened - but not now
 444                                 // Only do it with Twitter (htmlmode = 8)
 445                                 if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8))
 446                                         $limit = $limit - 23 + strlen($link);
 447
 448                                 $link = "";
 449
 450                                 if ($post["type"] == "text")
 451                                         unset($post["url"]);
 452                         }
 453                 }
 454         }
 455
 456         if ($limit > 0) {
 457                 // Reduce multiple spaces
 458                 // When posted to a network with limited space, we try to gain space where possible
 459                 while (strpos($msg, "  ") !== false)
 460                         $msg = str_replace("  ", " ", $msg);
 461
 462                 // Twitter is using its own limiter, so we always assume that shortened links will have this length
 463                 if (iconv_strlen($link, "UTF-8") > 0)
 464                         $limit = $limit - 23;
 465
 466                 if (iconv_strlen($msg, "UTF-8") > $limit) {
 467
 468                         if (($post["type"] == "text") && isset($post["url"])) {
 469                                 $post["url"] = $b["plink"];
 470                         } elseif (!isset($post["url"])) {
 471                                 $limit = $limit - 23;
 472                                 $post["url"] = $b["plink"];
 473                         // Which purpose has this line? It is now uncommented, but left as a reminder
 474                         //} elseif (strpos($b["body"], "[share") !== false) {
 475                         //      $post["url"] = $b["plink"];
 476                         } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 477                                 $post["url"] = $b["plink"];
 478                         }
 479                         $msg = shortenmsg($msg, $limit);
 480                 }
 481         }
 482
 483         $post["text"] = trim($msg);
 484
 485         return($post);
 486 }