include/plaintext.php

   1 <?php
   2
   3 /**
   4  * @file include/plaintext.php
   5  */
   6
   7 use \Friendica\ParseUrl;
   8
   9 require_once("include/Photo.php");
  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 = array();
  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], array("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                                 $picturedata = get_photo_info($matches[1]);
  55
  56                                 if (($picturedata[0] >= 500) AND ($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"] == "") AND (in_array($post["type"], array("link", "video")))
  67                                 AND 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
  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) {
 183 /*
 184  - text:
 185  - type: link, video, photo
 186  - title:
 187  - url:
 188  - image:
 189  - description:
 190  - (thumbnail)
 191 */
 192
 193         $post = get_attachment_data($body);
 194
 195         // if nothing is found, it maybe having an image.
 196         if (!isset($post["type"])) {
 197                 $URLSearchString = "^\[\]";
 198                 if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures,  PREG_SET_ORDER)) {
 199                         if (count($pictures) == 1) {
 200                                 // Checking, if the link goes to a picture
 201                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 202
 203                                 // Workaround:
 204                                 // Sometimes photo posts to the own album are not detected at the start.
 205                                 // So we seem to cannot use the cache for these cases. That's strange.
 206                                 if (($data["type"] != "photo") AND strstr($pictures[0][1], "/photos/"))
 207                                         $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 208
 209                                 if ($data["type"] == "photo") {
 210                                         $post["type"] = "photo";
 211                                         if (isset($data["images"][0])) {
 212                                                 $post["image"] = $data["images"][0]["src"];
 213                                                 $post["url"] = $data["url"];
 214                                         } else
 215                                                 $post["image"] = $data["url"];
 216
 217                                         $post["preview"] = $pictures[0][2];
 218                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 219                                 } else {
 220                                         $imgdata = get_photo_info($pictures[0][1]);
 221                                         if (substr($imgdata["mime"], 0, 6) == "image/") {
 222                                                 $post["type"] = "photo";
 223                                                 $post["image"] = $pictures[0][1];
 224                                                 $post["preview"] = $pictures[0][2];
 225                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 226                                         }
 227                                 }
 228                         } elseif (count($pictures) > 1) {
 229                                 $post["type"] = "link";
 230                                 $post["url"] = $b["plink"];
 231                                 $post["image"] = $pictures[0][2];
 232                                 $post["text"] = $body;
 233                         }
 234                 } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures,  PREG_SET_ORDER)) {
 235                         if (count($pictures) == 1) {
 236                                 $post["type"] = "photo";
 237                                 $post["image"] = $pictures[0][1];
 238                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 239                         } elseif (count($pictures) > 1) {
 240                                 $post["type"] = "link";
 241                                 $post["url"] = $b["plink"];
 242                                 $post["image"] = $pictures[0][1];
 243                                 $post["text"] = $body;
 244                         }
 245                 }
 246
 247                 if (preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links,  PREG_SET_ORDER)) {
 248                         if (count($links) == 1) {
 249                                 $post["type"] = "text";
 250                                 $post["url"] = $links[0][1];
 251                                 $post["text"] = $body;
 252                         }
 253                 }
 254                 if (!isset($post["type"])) {
 255                         $post["type"] = "text";
 256                         $post["text"] = trim($body);
 257                 }
 258         } elseif (isset($post["url"]) AND ($post["type"] == "video")) {
 259                 $data = ParseUrl::getSiteinfoCached($post["url"], true);
 260
 261                 if (isset($data["images"][0]))
 262                         $post["image"] = $data["images"][0]["src"];
 263         }
 264
 265         return($post);
 266 }
 267
 268 function shortenmsg($msg, $limit, $twitter = false) {
 269         /// @TODO
 270         /// For Twitter URLs aren't shortened, but they have to be calculated as if.
 271
 272         $lines = explode("\n", $msg);
 273         $msg = "";
 274         $recycle = html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 275         $ellipsis = html_entity_decode("&#x2026;", ENT_QUOTES, 'UTF-8');
 276         foreach ($lines AS $row=>$line) {
 277                 if (iconv_strlen(trim($msg."\n".$line), "UTF-8") <= $limit)
 278                         $msg = trim($msg."\n".$line);
 279                 // Is the new message empty by now or is it a reshared message?
 280                 elseif (($msg == "") OR (($row == 1) AND (substr($msg, 0, 4) == $recycle)))
 281                         $msg = iconv_substr(iconv_substr(trim($msg."\n".$line), 0, $limit, "UTF-8"), 0, -3, "UTF-8").$ellipsis;
 282                 else
 283                         break;
 284         }
 285         return($msg);
 286 }
 287
 288 /**
 289  * @brief Convert a message into plaintext for connectors to other networks
 290  *
 291  * @param App $a The application class
 292  * @param array $b The message array that is about to be posted
 293  * @param int $limit The maximum number of characters when posting to that network
 294  * @param bool $includedlinks Has an attached link to be included into the message?
 295  * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 296  * @param string $target_network Name of the network where the post should go to.
 297  *
 298  * @return string The converted message
 299  */
 300 function plaintext(App $a, $b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "") {
 301
 302         // Remove the hash tags
 303         $URLSearchString = "^\[\]";
 304         $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 305
 306         // Add an URL element if the text contains a raw link
 307         $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 308
 309         // Remove the abstract
 310         $body = remove_abstract($body);
 311
 312         // At first look at data that is attached via "type-..." stuff
 313         // This will hopefully replaced with a dedicated bbcode later
 314         //$post = get_attached_data($b["body"]);
 315         $post = get_attached_data($body);
 316
 317         if (($b["title"] != "") AND ($post["text"] != ""))
 318                 $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 319         elseif ($b["title"] != "")
 320                 $post["text"] = trim($b["title"]);
 321
 322         $abstract = "";
 323
 324         // Fetch the abstract from the given target network
 325         if ($target_network != "") {
 326                 $default_abstract = fetch_abstract($b["body"]);
 327                 $abstract = fetch_abstract($b["body"], $target_network);
 328
 329                 // If we post to a network with no limit we only fetch
 330                 // an abstract exactly for this network
 331                 if (($limit == 0) AND ($abstract == $default_abstract))
 332                         $abstract = "";
 333
 334         } else // Try to guess the correct target network
 335                 switch ($htmlmode) {
 336                         case 8:
 337                                 $abstract = fetch_abstract($b["body"], NETWORK_TWITTER);
 338                                 break;
 339                         case 7:
 340                                 $abstract = fetch_abstract($b["body"], NETWORK_STATUSNET);
 341                                 break;
 342                         case 6:
 343                                 $abstract = fetch_abstract($b["body"], NETWORK_APPNET);
 344                                 break;
 345                         default: // We don't know the exact target.
 346                                  // We fetch an abstract since there is a posting limit.
 347                                 if ($limit > 0)
 348                                         $abstract = fetch_abstract($b["body"]);
 349                 }
 350
 351         if ($abstract != "") {
 352                 $post["text"] = $abstract;
 353
 354                 if ($post["type"] == "text") {
 355                         $post["type"] = "link";
 356                         $post["url"] = $b["plink"];
 357                 }
 358         }
 359
 360         $html = bbcode($post["text"].$post["after"], false, false, $htmlmode);
 361         $msg = html2plain($html, 0, true);
 362         $msg = trim(html_entity_decode($msg,ENT_QUOTES,'UTF-8'));
 363
 364         $link = "";
 365         if ($includedlinks) {
 366                 if ($post["type"] == "link")
 367                         $link = $post["url"];
 368                 elseif ($post["type"] == "text")
 369                         $link = $post["url"];
 370                 elseif ($post["type"] == "video")
 371                         $link = $post["url"];
 372                 elseif ($post["type"] == "photo")
 373                         $link = $post["image"];
 374
 375                 if (($msg == "") AND isset($post["title"]))
 376                         $msg = trim($post["title"]);
 377
 378                 if (($msg == "") AND isset($post["description"]))
 379                         $msg = trim($post["description"]);
 380
 381                 // If the link is already contained in the post, then it neeedn't to be added again
 382                 // But: if the link is beyond the limit, then it has to be added.
 383                 if (($link != "") AND strstr($msg, $link)) {
 384                         $pos = strpos($msg, $link);
 385
 386                         // Will the text be shortened in the link?
 387                         // Or is the link the last item in the post?
 388                         if (($limit > 0) AND ($pos < $limit) AND (($pos + 23 > $limit) OR ($pos + strlen($link) == strlen($msg))))
 389                                 $msg = trim(str_replace($link, "", $msg));
 390                         elseif (($limit == 0) OR ($pos < $limit)) {
 391                                 // The limit has to be increased since it will be shortened - but not now
 392                                 // Only do it with Twitter (htmlmode = 8)
 393                                 if (($limit > 0) AND (strlen($link) > 23) AND ($htmlmode == 8))
 394                                         $limit = $limit - 23 + strlen($link);
 395
 396                                 $link = "";
 397
 398                                 if ($post["type"] == "text")
 399                                         unset($post["url"]);
 400                         }
 401                 }
 402         }
 403
 404         if ($limit > 0) {
 405                 // Reduce multiple spaces
 406                 // When posted to a network with limited space, we try to gain space where possible
 407                 while (strpos($msg, "  ") !== false)
 408                         $msg = str_replace("  ", " ", $msg);
 409
 410                 // Twitter is using its own limiter, so we always assume that shortened links will have this length
 411                 if (iconv_strlen($link, "UTF-8") > 0)
 412                         $limit = $limit - 23;
 413
 414                 if (iconv_strlen($msg, "UTF-8") > $limit) {
 415
 416                         if (($post["type"] == "text") AND isset($post["url"]))
 417                                 $post["url"] = $b["plink"];
 418                         elseif (!isset($post["url"])) {
 419                                 $limit = $limit - 23;
 420                                 $post["url"] = $b["plink"];
 421                         } elseif (strpos($b["body"], "[share") !== false)
 422                                 $post["url"] = $b["plink"];
 423                         elseif (get_pconfig($b["uid"], "system", "no_intelligent_shortening"))
 424                                 $post["url"] = $b["plink"];
 425
 426                         $msg = shortenmsg($msg, $limit);
 427                 }
 428         }
 429
 430         $post["text"] = trim($msg);
 431
 432         return($post);
 433 }
 434 ?>