src/Protocol/Feed.php

   1 <?php
   2 /**
   3  * @file src/Protocol/Feed.php
   4  * @brief Imports RSS/RDF/Atom feeds
   5  *
   6  */
   7 namespace Friendica\Protocol;
   8
   9 use DOMDocument;
  10 use DOMXPath;
  11 use Friendica\Content\Text\HTML;
  12 use Friendica\Core\Logger;
  13 use Friendica\Core\Protocol;
  14 use Friendica\Core\System;
  15 use Friendica\Database\DBA;
  16 use Friendica\DI;
  17 use Friendica\Model\Item;
  18 use Friendica\Protocol\ActivityNamespace;
  19 use Friendica\Util\ParseUrl;
  20 use Friendica\Util\Network;
  21 use Friendica\Util\XML;
  22
  23 /**
  24  * @brief This class contain functions to import feeds
  25  *
  26  */
  27 class Feed {
  28         /**
  29          * @brief Read a RSS/RDF/Atom feed and create an item entry for it
  30          *
  31          * @param string $xml      The feed data
  32          * @param array  $importer The user record of the importer
  33          * @param array  $contact  The contact record of the feed
  34          * @param string $hub      Unused dummy value for compatibility reasons
  35          * @param bool   $simulate If enabled, no data is imported
  36          *
  37          * @return array In simulation mode it returns the header and the first item
  38          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  39          */
  40         public static function import($xml, $importer, &$contact, &$hub, $simulate = false) {
  41
  42                 $a = \get_app();
  43
  44                 if (!$simulate) {
  45                         Logger::log("Import Atom/RSS feed '".$contact["name"]."' (Contact ".$contact["id"].") for user ".$importer["uid"], Logger::DEBUG);
  46                 } else {
  47                         Logger::log("Test Atom/RSS feed", Logger::DEBUG);
  48                 }
  49                 if (empty($xml)) {
  50                         Logger::log('XML is empty.', Logger::DEBUG);
  51                         return;
  52                 }
  53
  54                 if (!empty($contact['poll'])) {
  55                         $basepath = $contact['poll'];
  56                 } elseif (!empty($contact['url'])) {
  57                         $basepath = $contact['url'];
  58                 } else {
  59                         $basepath = '';
  60                 }
  61
  62                 $doc = new DOMDocument();
  63                 @$doc->loadXML(trim($xml));
  64                 $xpath = new DOMXPath($doc);
  65                 $xpath->registerNamespace('atom', ActivityNamespace::ATOM1);
  66                 $xpath->registerNamespace('dc', "http://purl.org/dc/elements/1.1/");
  67                 $xpath->registerNamespace('content', "http://purl.org/rss/1.0/modules/content/");
  68                 $xpath->registerNamespace('rdf', "http://www.w3.org/1999/02/22-rdf-syntax-ns#");
  69                 $xpath->registerNamespace('rss', "http://purl.org/rss/1.0/");
  70                 $xpath->registerNamespace('media', "http://search.yahoo.com/mrss/");
  71                 $xpath->registerNamespace('poco', ActivityNamespace::POCO);
  72
  73                 $author = [];
  74                 $entries = null;
  75
  76                 // Is it RDF?
  77                 if ($xpath->query('/rdf:RDF/rss:channel')->length > 0) {
  78                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:link/text()');
  79                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:title/text()');
  80
  81                         if (empty($author["author-name"])) {
  82                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:description/text()');
  83                         }
  84                         $entries = $xpath->query('/rdf:RDF/rss:item');
  85                 }
  86
  87                 // Is it Atom?
  88                 if ($xpath->query('/atom:feed')->length > 0) {
  89                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']");
  90                         if (is_object($alternate)) {
  91                                 foreach ($alternate AS $attribute) {
  92                                         if ($attribute->name == "href") {
  93                                                 $author["author-link"] = $attribute->textContent;
  94                                         }
  95                                 }
  96                         }
  97
  98                         if (empty($author["author-link"])) {
  99                                 $self = XML::getFirstAttributes($xpath, "atom:link[@rel='self']");
 100                                 if (is_object($self)) {
 101                                         foreach ($self AS $attribute) {
 102                                                 if ($attribute->name == "href") {
 103                                                         $author["author-link"] = $attribute->textContent;
 104                                                 }
 105                                         }
 106                                 }
 107                         }
 108
 109                         if (empty($author["author-link"])) {
 110                                 $author["author-link"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:id/text()');
 111                         }
 112                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:logo/text()');
 113
 114                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:title/text()');
 115
 116                         if (empty($author["author-name"])) {
 117                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:subtitle/text()');
 118                         }
 119                         if (empty($author["author-name"])) {
 120                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:name/text()');
 121                         }
 122                         $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:displayName/text()');
 123                         if ($value != "") {
 124                                 $author["author-name"] = $value;
 125                         }
 126                         if ($simulate) {
 127                                 $author["author-id"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:id/text()');
 128
 129                                 // See https://tools.ietf.org/html/rfc4287#section-3.2.2
 130                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/atom:uri/text()');
 131                                 if ($value != "") {
 132                                         $author["author-link"] = $value;
 133                                 }
 134
 135                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:preferredUsername/text()');
 136                                 if ($value != "") {
 137                                         $author["author-nick"] = $value;
 138                                 }
 139                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:address/poco:formatted/text()');
 140                                 if ($value != "") {
 141                                         $author["author-location"] = $value;
 142                                 }
 143                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:note/text()');
 144                                 if ($value != "") {
 145                                         $author["author-about"] = $value;
 146                                 }
 147                                 $avatar = XML::getFirstAttributes($xpath, "atom:author/atom:link[@rel='avatar']");
 148                                 if (is_object($avatar)) {
 149                                         foreach ($avatar AS $attribute) {
 150                                                 if ($attribute->name == "href") {
 151                                                         $author["author-avatar"] = $attribute->textContent;
 152                                                 }
 153                                         }
 154                                 }
 155                         }
 156
 157                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:updated/text()');
 158
 159                         $author["app"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:generator/text()');
 160
 161                         $entries = $xpath->query('/atom:feed/atom:entry');
 162                 }
 163
 164                 // Is it RSS?
 165                 if ($xpath->query('/rss/channel')->length > 0) {
 166                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rss/channel/link/text()');
 167
 168                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/title/text()');
 169                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/rss/channel/image/url/text()');
 170
 171                         if (empty($author["author-name"])) {
 172                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/copyright/text()');
 173                         }
 174                         if (empty($author["author-name"])) {
 175                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/description/text()');
 176                         }
 177                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/rss/channel/pubDate/text()');
 178
 179                         $author["app"] = XML::getFirstNodeValue($xpath, '/rss/channel/generator/text()');
 180
 181                         $entries = $xpath->query('/rss/channel/item');
 182                 }
 183
 184                 if (!$simulate) {
 185                         $author["author-link"] = $contact["url"];
 186
 187                         if (empty($author["author-name"])) {
 188                                 $author["author-name"] = $contact["name"];
 189                         }
 190                         $author["author-avatar"] = $contact["thumb"];
 191
 192                         $author["owner-link"] = $contact["url"];
 193                         $author["owner-name"] = $contact["name"];
 194                         $author["owner-avatar"] = $contact["thumb"];
 195                 }
 196
 197                 $header = [];
 198                 $header["uid"] = $importer["uid"];
 199                 $header["network"] = Protocol::FEED;
 200                 $header["wall"] = 0;
 201                 $header["origin"] = 0;
 202                 $header["gravity"] = GRAVITY_PARENT;
 203                 $header["private"] = 2;
 204                 $header["verb"] = Activity::POST;
 205                 $header["object-type"] = Activity\ObjectType::NOTE;
 206
 207                 $header["contact-id"] = $contact["id"];
 208
 209                 if (!is_object($entries)) {
 210                         Logger::log("There are no entries in this feed.", Logger::DEBUG);
 211                         return;
 212                 }
 213
 214                 $items = [];
 215                 // Importing older entries first
 216                 for($i = $entries->length - 1; $i >= 0;--$i) {
 217                         $entry = $entries->item($i);
 218
 219                         $item = array_merge($header, $author);
 220
 221                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']", $entry);
 222                         if (!is_object($alternate)) {
 223                                 $alternate = XML::getFirstAttributes($xpath, "atom:link", $entry);
 224                         }
 225                         if (is_object($alternate)) {
 226                                 foreach ($alternate AS $attribute) {
 227                                         if ($attribute->name == "href") {
 228                                                 $item["plink"] = $attribute->textContent;
 229                                         }
 230                                 }
 231                         }
 232                         if (empty($item["plink"])) {
 233                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'link/text()', $entry);
 234                         }
 235                         if (empty($item["plink"])) {
 236                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'rss:link/text()', $entry);
 237                         }
 238
 239                         $item["uri"] = XML::getFirstNodeValue($xpath, 'atom:id/text()', $entry);
 240
 241                         if (empty($item["uri"])) {
 242                                 $item["uri"] = XML::getFirstNodeValue($xpath, 'guid/text()', $entry);
 243                         }
 244                         if (empty($item["uri"])) {
 245                                 $item["uri"] = $item["plink"];
 246                         }
 247
 248                         $orig_plink = $item["plink"];
 249
 250                         $item["plink"] = Network::finalUrl($item["plink"]);
 251
 252                         $item["parent-uri"] = $item["uri"];
 253
 254                         if (!$simulate) {
 255                                 $condition = ["`uid` = ? AND `uri` = ? AND `network` IN (?, ?)",
 256                                         $importer["uid"], $item["uri"], Protocol::FEED, Protocol::DFRN];
 257                                 $previous = Item::selectFirst(['id'], $condition);
 258                                 if (DBA::isResult($previous)) {
 259                                         Logger::log("Item with uri ".$item["uri"]." for user ".$importer["uid"]." already existed under id ".$previous["id"], Logger::DEBUG);
 260                                         continue;
 261                                 }
 262                         }
 263
 264                         $item["title"] = XML::getFirstNodeValue($xpath, 'atom:title/text()', $entry);
 265
 266                         if (empty($item["title"])) {
 267                                 $item["title"] = XML::getFirstNodeValue($xpath, 'title/text()', $entry);
 268                         }
 269                         if (empty($item["title"])) {
 270                                 $item["title"] = XML::getFirstNodeValue($xpath, 'rss:title/text()', $entry);
 271                         }
 272
 273                         $item["title"] = html_entity_decode($item["title"], ENT_QUOTES, 'UTF-8');
 274
 275                         $published = XML::getFirstNodeValue($xpath, 'atom:published/text()', $entry);
 276
 277                         if (empty($published)) {
 278                                 $published = XML::getFirstNodeValue($xpath, 'pubDate/text()', $entry);
 279                         }
 280                         if (empty($published)) {
 281                                 $published = XML::getFirstNodeValue($xpath, 'dc:date/text()', $entry);
 282                         }
 283                         $updated = XML::getFirstNodeValue($xpath, 'atom:updated/text()', $entry);
 284
 285                         if (empty($updated) && !empty($published)) {
 286                                 $updated = $published;
 287                         }
 288
 289                         if (empty($published) && !empty($updated)) {
 290                                 $published = $updated;
 291                         }
 292
 293                         if ($published != "") {
 294                                 $item["created"] = $published;
 295                         }
 296                         if ($updated != "") {
 297                                 $item["edited"] = $updated;
 298                         }
 299                         $creator = XML::getFirstNodeValue($xpath, 'author/text()', $entry);
 300
 301                         if (empty($creator)) {
 302                                 $creator = XML::getFirstNodeValue($xpath, 'atom:author/atom:name/text()', $entry);
 303                         }
 304                         if (empty($creator)) {
 305                                 $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 306                         }
 307                         if ($creator != "") {
 308                                 $item["author-name"] = $creator;
 309                         }
 310                         $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 311
 312                         if ($creator != "") {
 313                                 $item["author-name"] = $creator;
 314                         }
 315
 316                         /// @TODO ?
 317                         // <category>Ausland</category>
 318                         // <media:thumbnail width="152" height="76" url="http://www.taz.de/picture/667875/192/14388767.jpg"/>
 319
 320                         $attachments = [];
 321
 322                         $enclosures = $xpath->query("enclosure|atom:link[@rel='enclosure']", $entry);
 323                         foreach ($enclosures AS $enclosure) {
 324                                 $href = "";
 325                                 $length = "";
 326                                 $type = "";
 327
 328                                 foreach ($enclosure->attributes AS $attribute) {
 329                                         if (in_array($attribute->name, ["url", "href"])) {
 330                                                 $href = $attribute->textContent;
 331                                         } elseif ($attribute->name == "length") {
 332                                                 $length = $attribute->textContent;
 333                                         } elseif ($attribute->name == "type") {
 334                                                 $type = $attribute->textContent;
 335                                         }
 336                                 }
 337                                 if (!empty($item["attach"])) {
 338                                         $item["attach"] .= ',';
 339                                 } else {
 340                                         $item["attach"] = '';
 341                                 }
 342
 343                                 $attachments[] = ["link" => $href, "type" => $type, "length" => $length];
 344
 345                                 $item["attach"] .= '[attach]href="'.$href.'" length="'.$length.'" type="'.$type.'"[/attach]';
 346                         }
 347
 348                         $tags = '';
 349                         $categories = $xpath->query("category", $entry);
 350                         foreach ($categories AS $category) {
 351                                 $hashtag = $category->nodeValue;
 352                                 if ($tags != '') {
 353                                         $tags .= ', ';
 354                                 }
 355
 356                                 $taglink = "#[url=" . System::baseUrl() . "/search?tag=" . $hashtag . "]" . $hashtag . "[/url]";
 357                                 $tags .= $taglink;
 358                         }
 359
 360                         $body = trim(XML::getFirstNodeValue($xpath, 'atom:content/text()', $entry));
 361
 362                         if (empty($body)) {
 363                                 $body = trim(XML::getFirstNodeValue($xpath, 'content:encoded/text()', $entry));
 364                         }
 365
 366                         $summary = trim(XML::getFirstNodeValue($xpath, 'atom:summary/text()', $entry));
 367
 368                         if (empty($summary)) {
 369                                 $summary = trim(XML::getFirstNodeValue($xpath, 'description/text()', $entry));
 370                         }
 371
 372                         if (empty($body)) {
 373                                 $body = $summary;
 374                                 $summary = '';
 375                         }
 376
 377                         if ($body == $summary) {
 378                                 $summary = '';
 379                         }
 380
 381                         // remove the content of the title if it is identically to the body
 382                         // This helps with auto generated titles e.g. from tumblr
 383                         if (self::titleIsBody($item["title"], $body)) {
 384                                 $item["title"] = "";
 385                         }
 386                         $item["body"] = HTML::toBBCode($body, $basepath);
 387
 388                         if (($item["body"] == '') && ($item["title"] != '')) {
 389                                 $item["body"] = $item["title"];
 390                                 $item["title"] = '';
 391                         }
 392
 393                         $preview = '';
 394                         if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] < 3)) {
 395                                 // Handle enclosures and treat them as preview picture
 396                                 foreach ($attachments AS $attachment) {
 397                                         if ($attachment["type"] == "image/jpeg") {
 398                                                 $preview = $attachment["link"];
 399                                         }
 400                                 }
 401
 402                                 // Remove a possible link to the item itself
 403                                 $item["body"] = str_replace($item["plink"], '', $item["body"]);
 404                                 $item["body"] = trim(preg_replace('/\[url\=\](\w+.*?)\[\/url\]/i', '', $item["body"]));
 405
 406                                 // Replace the content when the title is longer than the body
 407                                 $replace = (strlen($item["title"]) > strlen($item["body"]));
 408
 409                                 // Replace it, when there is an image in the body
 410                                 if (strstr($item["body"], '[/img]')) {
 411                                         $replace = true;
 412                                 }
 413
 414                                 // Replace it, when there is a link in the body
 415                                 if (strstr($item["body"], '[/url]')) {
 416                                         $replace = true;
 417                                 }
 418
 419                                 if ($replace) {
 420                                         $item["body"] = trim($item["title"]);
 421                                 }
 422
 423                                 $data = ParseUrl::getSiteinfoCached($item['plink'], true);
 424                                 if (!empty($data['text']) && !empty($data['title']) && (mb_strlen($item['body']) < mb_strlen($data['text']))) {
 425                                         // When the fetched page info text is longer than the body, we do try to enhance the body
 426                                         if (!empty($item['body']) && (strpos($data['title'], $item['body']) === false) && (strpos($data['text'], $item['body']) === false)) {
 427                                                 // The body is not part of the fetched page info title or page info text. So we add the text to the body
 428                                                 $item['body'] .= "\n\n" . $data['text'];
 429                                         } else {
 430                                                 // Else we replace the body with the page info text
 431                                                 $item['body'] = $data['text'];
 432                                         }
 433                                 }
 434
 435                                 // We always strip the title since it will be added in the page information
 436                                 $item["title"] = "";
 437                                 $item["body"] = $item["body"].add_page_info($item["plink"], false, $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 438                                 $item["tag"] = add_page_keywords($item["plink"], $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 439                                 $item["object-type"] = Activity\ObjectType::BOOKMARK;
 440                                 unset($item["attach"]);
 441                         } else {
 442                                 if (!empty($summary)) {
 443                                         $item["body"] = '[abstract]' . HTML::toBBCode($summary, $basepath) . "[/abstract]\n" . $item["body"];
 444                                 }
 445
 446                                 if ($contact["fetch_further_information"] == 3) {
 447                                         if (!empty($tags)) {
 448                                                 $item["tag"] = $tags;
 449                                         } else {
 450                                                 // @todo $preview is never set in this case, is it intended? - @MrPetovan 2018-02-13
 451                                                 $item["tag"] = add_page_keywords($item["plink"], $preview, true, $contact["ffi_keyword_blacklist"]);
 452                                         }
 453                                         $item["body"] .= "\n".$item['tag'];
 454                                 }
 455                                 // Add the link to the original feed entry if not present in feed
 456                                 if (($item['plink'] != '') && !strstr($item["body"], $item['plink'])) {
 457                                         $item["body"] .= "[hr][url]".$item['plink']."[/url]";
 458                                 }
 459                         }
 460
 461                         if (!$simulate) {
 462                                 Logger::log("Stored feed: ".print_r($item, true), Logger::DEBUG);
 463
 464                                 $notify = Item::isRemoteSelf($contact, $item);
 465
 466                                 // Distributed items should have a well formatted URI.
 467                                 // Additionally we have to avoid conflicts with identical URI between imported feeds and these items.
 468                                 if ($notify) {
 469                                         $item['guid'] = Item::guidFromUri($orig_plink, DI::baseUrl()->getHostname()());
 470                                         unset($item['uri']);
 471                                         unset($item['parent-uri']);
 472
 473                                         // Set the delivery priority for "remote self" to "medium"
 474                                         $notify = PRIORITY_MEDIUM;
 475                                 }
 476
 477                                 $id = Item::insert($item, false, $notify);
 478
 479                                 Logger::log("Feed for contact ".$contact["url"]." stored under id ".$id);
 480                         } else {
 481                                 $items[] = $item;
 482                         }
 483                         if ($simulate) {
 484                                 break;
 485                         }
 486                 }
 487
 488                 if ($simulate) {
 489                         return ["header" => $author, "items" => $items];
 490                 }
 491         }
 492
 493         private static function titleIsBody($title, $body)
 494         {
 495                 $title = strip_tags($title);
 496                 $title = trim($title);
 497                 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 498                 $title = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $title);
 499
 500                 $body = strip_tags($body);
 501                 $body = trim($body);
 502                 $body = html_entity_decode($body, ENT_QUOTES, 'UTF-8');
 503                 $body = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $body);
 504
 505                 if (strlen($title) < strlen($body)) {
 506                         $body = substr($body, 0, strlen($title));
 507                 }
 508
 509                 if (($title != $body) && (substr($title, -3) == "...")) {
 510                         $pos = strrpos($title, "...");
 511                         if ($pos > 0) {
 512                                 $title = substr($title, 0, $pos);
 513                                 $body = substr($body, 0, $pos);
 514                         }
 515                 }
 516                 return ($title == $body);
 517         }
 518 }