include/plaintext.php

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