src/Protocol/Feed.php

   1 <?php
   2 /**
   3  * @file src/Protocol/Feed.php
   4  * 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\Database\DBA;
  15 use Friendica\DI;
  16 use Friendica\Model\Item;
  17 use Friendica\Util\Network;
  18 use Friendica\Util\ParseUrl;
  19 use Friendica\Util\XML;
  20
  21 /**
  22  * This class contain functions to import feeds
  23  *
  24  */
  25 class Feed {
  26         /**
  27          * Read a RSS/RDF/Atom feed and create an item entry for it
  28          *
  29          * @param string $xml      The feed data
  30          * @param array  $importer The user record of the importer
  31          * @param array  $contact  The contact record of the feed
  32          *
  33          * @return array Returns the header and the first item in dry run mode
  34          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  35          */
  36         public static function import($xml, array $importer = [], array $contact = [])
  37         {
  38                 $dryRun = empty($importer) && empty($contact);
  39
  40                 if ($dryRun) {
  41                         Logger::info("Test Atom/RSS feed");
  42                 } else {
  43                         Logger::info("Import Atom/RSS feed '" . $contact["name"] . "' (Contact " . $contact["id"] . ") for user " . $importer["uid"]);
  44                 }
  45
  46                 if (empty($xml)) {
  47                         Logger::info('XML is empty.');
  48                         return [];
  49                 }
  50
  51                 if (!empty($contact['poll'])) {
  52                         $basepath = $contact['poll'];
  53                 } elseif (!empty($contact['url'])) {
  54                         $basepath = $contact['url'];
  55                 } else {
  56                         $basepath = '';
  57                 }
  58
  59                 $doc = new DOMDocument();
  60                 @$doc->loadXML(trim($xml));
  61                 $xpath = new DOMXPath($doc);
  62                 $xpath->registerNamespace('atom', ActivityNamespace::ATOM1);
  63                 $xpath->registerNamespace('dc', "http://purl.org/dc/elements/1.1/");
  64                 $xpath->registerNamespace('content', "http://purl.org/rss/1.0/modules/content/");
  65                 $xpath->registerNamespace('rdf', "http://www.w3.org/1999/02/22-rdf-syntax-ns#");
  66                 $xpath->registerNamespace('rss', "http://purl.org/rss/1.0/");
  67                 $xpath->registerNamespace('media', "http://search.yahoo.com/mrss/");
  68                 $xpath->registerNamespace('poco', ActivityNamespace::POCO);
  69
  70                 $author = [];
  71                 $entries = null;
  72
  73                 // Is it RDF?
  74                 if ($xpath->query('/rdf:RDF/rss:channel')->length > 0) {
  75                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:link/text()');
  76                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:title/text()');
  77
  78                         if (empty($author["author-name"])) {
  79                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:description/text()');
  80                         }
  81                         $entries = $xpath->query('/rdf:RDF/rss:item');
  82                 }
  83
  84                 // Is it Atom?
  85                 if ($xpath->query('/atom:feed')->length > 0) {
  86                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']");
  87                         if (is_object($alternate)) {
  88                                 foreach ($alternate AS $attribute) {
  89                                         if ($attribute->name == "href") {
  90                                                 $author["author-link"] = $attribute->textContent;
  91                                         }
  92                                 }
  93                         }
  94
  95                         if (empty($author["author-link"])) {
  96                                 $self = XML::getFirstAttributes($xpath, "atom:link[@rel='self']");
  97                                 if (is_object($self)) {
  98                                         foreach ($self AS $attribute) {
  99                                                 if ($attribute->name == "href") {
 100                                                         $author["author-link"] = $attribute->textContent;
 101                                                 }
 102                                         }
 103                                 }
 104                         }
 105
 106                         if (empty($author["author-link"])) {
 107                                 $author["author-link"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:id/text()');
 108                         }
 109                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:logo/text()');
 110
 111                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:title/text()');
 112
 113                         if (empty($author["author-name"])) {
 114                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:subtitle/text()');
 115                         }
 116
 117                         if (empty($author["author-name"])) {
 118                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:name/text()');
 119                         }
 120
 121                         $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:displayName/text()');
 122                         if ($value != "") {
 123                                 $author["author-name"] = $value;
 124                         }
 125
 126                         if ($dryRun) {
 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
 140                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:address/poco:formatted/text()');
 141                                 if ($value != "") {
 142                                         $author["author-location"] = $value;
 143                                 }
 144
 145                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:note/text()');
 146                                 if ($value != "") {
 147                                         $author["author-about"] = $value;
 148                                 }
 149
 150                                 $avatar = XML::getFirstAttributes($xpath, "atom:author/atom:link[@rel='avatar']");
 151                                 if (is_object($avatar)) {
 152                                         foreach ($avatar AS $attribute) {
 153                                                 if ($attribute->name == "href") {
 154                                                         $author["author-avatar"] = $attribute->textContent;
 155                                                 }
 156                                         }
 157                                 }
 158                         }
 159
 160                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:updated/text()');
 161
 162                         $author["app"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:generator/text()');
 163
 164                         $entries = $xpath->query('/atom:feed/atom:entry');
 165                 }
 166
 167                 // Is it RSS?
 168                 if ($xpath->query('/rss/channel')->length > 0) {
 169                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rss/channel/link/text()');
 170
 171                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/title/text()');
 172                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/rss/channel/image/url/text()');
 173
 174                         if (empty($author["author-name"])) {
 175                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/copyright/text()');
 176                         }
 177
 178                         if (empty($author["author-name"])) {
 179                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/description/text()');
 180                         }
 181
 182                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/rss/channel/pubDate/text()');
 183
 184                         $author["app"] = XML::getFirstNodeValue($xpath, '/rss/channel/generator/text()');
 185
 186                         $entries = $xpath->query('/rss/channel/item');
 187                 }
 188
 189                 if (!$dryRun) {
 190                         $author["author-link"] = $contact["url"];
 191
 192                         if (empty($author["author-name"])) {
 193                                 $author["author-name"] = $contact["name"];
 194                         }
 195
 196                         $author["author-avatar"] = $contact["thumb"];
 197
 198                         $author["owner-link"] = $contact["url"];
 199                         $author["owner-name"] = $contact["name"];
 200                         $author["owner-avatar"] = $contact["thumb"];
 201                 }
 202
 203                 $header = [];
 204                 $header["uid"] = $importer["uid"] ?? 0;
 205                 $header["network"] = Protocol::FEED;
 206                 $header["wall"] = 0;
 207                 $header["origin"] = 0;
 208                 $header["gravity"] = GRAVITY_PARENT;
 209                 $header["private"] = 2;
 210                 $header["verb"] = Activity::POST;
 211                 $header["object-type"] = Activity\ObjectType::NOTE;
 212
 213                 $header["contact-id"] = $contact["id"] ?? 0;
 214
 215                 if (!is_object($entries)) {
 216                         Logger::info("There are no entries in this feed.");
 217                         return [];
 218                 }
 219
 220                 $items = [];
 221                 // Importing older entries first
 222                 for ($i = $entries->length - 1; $i >= 0; --$i) {
 223                         $entry = $entries->item($i);
 224
 225                         $item = array_merge($header, $author);
 226
 227                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']", $entry);
 228                         if (!is_object($alternate)) {
 229                                 $alternate = XML::getFirstAttributes($xpath, "atom:link", $entry);
 230                         }
 231                         if (is_object($alternate)) {
 232                                 foreach ($alternate AS $attribute) {
 233                                         if ($attribute->name == "href") {
 234                                                 $item["plink"] = $attribute->textContent;
 235                                         }
 236                                 }
 237                         }
 238
 239                         if (empty($item["plink"])) {
 240                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'link/text()', $entry);
 241                         }
 242
 243                         if (empty($item["plink"])) {
 244                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'rss:link/text()', $entry);
 245                         }
 246
 247                         $item["uri"] = XML::getFirstNodeValue($xpath, 'atom:id/text()', $entry);
 248
 249                         if (empty($item["uri"])) {
 250                                 $item["uri"] = XML::getFirstNodeValue($xpath, 'guid/text()', $entry);
 251                         }
 252
 253                         if (empty($item["uri"])) {
 254                                 $item["uri"] = $item["plink"];
 255                         }
 256
 257                         $orig_plink = $item["plink"];
 258
 259                         $item["plink"] = Network::finalUrl($item["plink"]);
 260
 261                         $item["parent-uri"] = $item["uri"];
 262
 263                         if (!$dryRun) {
 264                                 $condition = ["`uid` = ? AND `uri` = ? AND `network` IN (?, ?)",
 265                                         $importer["uid"], $item["uri"], Protocol::FEED, Protocol::DFRN];
 266                                 $previous = Item::selectFirst(['id'], $condition);
 267                                 if (DBA::isResult($previous)) {
 268                                         Logger::info("Item with uri " . $item["uri"] . " for user " . $importer["uid"] . " already existed under id " . $previous["id"]);
 269                                         continue;
 270                                 }
 271                         }
 272
 273                         $item["title"] = XML::getFirstNodeValue($xpath, 'atom:title/text()', $entry);
 274
 275                         if (empty($item["title"])) {
 276                                 $item["title"] = XML::getFirstNodeValue($xpath, 'title/text()', $entry);
 277                         }
 278                         if (empty($item["title"])) {
 279                                 $item["title"] = XML::getFirstNodeValue($xpath, 'rss:title/text()', $entry);
 280                         }
 281
 282                         $item["title"] = html_entity_decode($item["title"], ENT_QUOTES, 'UTF-8');
 283
 284                         $published = XML::getFirstNodeValue($xpath, 'atom:published/text()', $entry);
 285
 286                         if (empty($published)) {
 287                                 $published = XML::getFirstNodeValue($xpath, 'pubDate/text()', $entry);
 288                         }
 289
 290                         if (empty($published)) {
 291                                 $published = XML::getFirstNodeValue($xpath, 'dc:date/text()', $entry);
 292                         }
 293
 294                         $updated = XML::getFirstNodeValue($xpath, 'atom:updated/text()', $entry);
 295
 296                         if (empty($updated) && !empty($published)) {
 297                                 $updated = $published;
 298                         }
 299
 300                         if (empty($published) && !empty($updated)) {
 301                                 $published = $updated;
 302                         }
 303
 304                         if ($published != "") {
 305                                 $item["created"] = $published;
 306                         }
 307
 308                         if ($updated != "") {
 309                                 $item["edited"] = $updated;
 310                         }
 311
 312                         $creator = XML::getFirstNodeValue($xpath, 'author/text()', $entry);
 313
 314                         if (empty($creator)) {
 315                                 $creator = XML::getFirstNodeValue($xpath, 'atom:author/atom:name/text()', $entry);
 316                         }
 317
 318                         if (empty($creator)) {
 319                                 $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 320                         }
 321
 322                         if ($creator != "") {
 323                                 $item["author-name"] = $creator;
 324                         }
 325
 326                         $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 327
 328                         if ($creator != "") {
 329                                 $item["author-name"] = $creator;
 330                         }
 331
 332                         /// @TODO ?
 333                         // <category>Ausland</category>
 334                         // <media:thumbnail width="152" height="76" url="http://www.taz.de/picture/667875/192/14388767.jpg"/>
 335
 336                         $attachments = [];
 337
 338                         $enclosures = $xpath->query("enclosure|atom:link[@rel='enclosure']", $entry);
 339                         foreach ($enclosures AS $enclosure) {
 340                                 $href = "";
 341                                 $length = "";
 342                                 $type = "";
 343
 344                                 foreach ($enclosure->attributes AS $attribute) {
 345                                         if (in_array($attribute->name, ["url", "href"])) {
 346                                                 $href = $attribute->textContent;
 347                                         } elseif ($attribute->name == "length") {
 348                                                 $length = $attribute->textContent;
 349                                         } elseif ($attribute->name == "type") {
 350                                                 $type = $attribute->textContent;
 351                                         }
 352                                 }
 353
 354                                 if (!empty($item["attach"])) {
 355                                         $item["attach"] .= ',';
 356                                 } else {
 357                                         $item["attach"] = '';
 358                                 }
 359
 360                                 $attachments[] = ["link" => $href, "type" => $type, "length" => $length];
 361
 362                                 $item["attach"] .= '[attach]href="' . $href . '" length="' . $length . '" type="' . $type . '"[/attach]';
 363                         }
 364
 365                         $tags = '';
 366                         $categories = $xpath->query("category", $entry);
 367                         foreach ($categories AS $category) {
 368                                 $hashtag = $category->nodeValue;
 369                                 if ($tags != '') {
 370                                         $tags .= ', ';
 371                                 }
 372
 373                                 $taglink = "#[url=" . DI::baseUrl() . "/search?tag=" . $hashtag . "]" . $hashtag . "[/url]";
 374                                 $tags .= $taglink;
 375                         }
 376
 377                         $body = trim(XML::getFirstNodeValue($xpath, 'atom:content/text()', $entry));
 378
 379                         if (empty($body)) {
 380                                 $body = trim(XML::getFirstNodeValue($xpath, 'content:encoded/text()', $entry));
 381                         }
 382
 383                         $summary = trim(XML::getFirstNodeValue($xpath, 'atom:summary/text()', $entry));
 384
 385                         if (empty($summary)) {
 386                                 $summary = trim(XML::getFirstNodeValue($xpath, 'description/text()', $entry));
 387                         }
 388
 389                         if (empty($body)) {
 390                                 $body = $summary;
 391                                 $summary = '';
 392                         }
 393
 394                         if ($body == $summary) {
 395                                 $summary = '';
 396                         }
 397
 398                         // remove the content of the title if it is identically to the body
 399                         // This helps with auto generated titles e.g. from tumblr
 400                         if (self::titleIsBody($item["title"], $body)) {
 401                                 $item["title"] = "";
 402                         }
 403                         $item["body"] = HTML::toBBCode($body, $basepath);
 404
 405                         if (($item["body"] == '') && ($item["title"] != '')) {
 406                                 $item["body"] = $item["title"];
 407                                 $item["title"] = '';
 408                         }
 409
 410                         $preview = '';
 411                         if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] < 3)) {
 412                                 // Handle enclosures and treat them as preview picture
 413                                 foreach ($attachments AS $attachment) {
 414                                         if ($attachment["type"] == "image/jpeg") {
 415                                                 $preview = $attachment["link"];
 416                                         }
 417                                 }
 418
 419                                 // Remove a possible link to the item itself
 420                                 $item["body"] = str_replace($item["plink"], '', $item["body"]);
 421                                 $item["body"] = trim(preg_replace('/\[url\=\](\w+.*?)\[\/url\]/i', '', $item["body"]));
 422
 423                                 // Replace the content when the title is longer than the body
 424                                 $replace = (strlen($item["title"]) > strlen($item["body"]));
 425
 426                                 // Replace it, when there is an image in the body
 427                                 if (strstr($item["body"], '[/img]')) {
 428                                         $replace = true;
 429                                 }
 430
 431                                 // Replace it, when there is a link in the body
 432                                 if (strstr($item["body"], '[/url]')) {
 433                                         $replace = true;
 434                                 }
 435
 436                                 if ($replace) {
 437                                         $item["body"] = trim($item["title"]);
 438                                 }
 439
 440                                 $data = ParseUrl::getSiteinfoCached($item['plink'], true);
 441                                 if (!empty($data['text']) && !empty($data['title']) && (mb_strlen($item['body']) < mb_strlen($data['text']))) {
 442                                         // When the fetched page info text is longer than the body, we do try to enhance the body
 443                                         if (!empty($item['body']) && (strpos($data['title'], $item['body']) === false) && (strpos($data['text'], $item['body']) === false)) {
 444                                                 // The body is not part of the fetched page info title or page info text. So we add the text to the body
 445                                                 $item['body'] .= "\n\n" . $data['text'];
 446                                         } else {
 447                                                 // Else we replace the body with the page info text
 448                                                 $item['body'] = $data['text'];
 449                                         }
 450                                 }
 451
 452                                 // We always strip the title since it will be added in the page information
 453                                 $item["title"] = "";
 454                                 $item["body"] = $item["body"] . add_page_info($item["plink"], false, $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 455                                 $item["tag"] = add_page_keywords($item["plink"], $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 456                                 $item["object-type"] = Activity\ObjectType::BOOKMARK;
 457                                 unset($item["attach"]);
 458                         } else {
 459                                 if (!empty($summary)) {
 460                                         $item["body"] = '[abstract]' . HTML::toBBCode($summary, $basepath) . "[/abstract]\n" . $item["body"];
 461                                 }
 462
 463                                 if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] == 3)) {
 464                                         if (!empty($tags)) {
 465                                                 $item["tag"] = $tags;
 466                                         } else {
 467                                                 // @todo $preview is never set in this case, is it intended? - @MrPetovan 2018-02-13
 468                                                 $item["tag"] = add_page_keywords($item["plink"], $preview, true, $contact["ffi_keyword_blacklist"]);
 469                                         }
 470                                         $item["body"] .= "\n" . $item['tag'];
 471                                 }
 472
 473                                 // Add the link to the original feed entry if not present in feed
 474                                 if (($item['plink'] != '') && !strstr($item["body"], $item['plink'])) {
 475                                         $item["body"] .= "[hr][url]" . $item['plink'] . "[/url]";
 476                                 }
 477                         }
 478
 479                         if ($dryRun) {
 480                                 $items[] = $item;
 481                                 break;
 482                         } else {
 483                                 Logger::info("Stored feed: " . print_r($item, true));
 484
 485                                 $notify = Item::isRemoteSelf($contact, $item);
 486
 487                                 // Distributed items should have a well formatted URI.
 488                                 // Additionally we have to avoid conflicts with identical URI between imported feeds and these items.
 489                                 if ($notify) {
 490                                         $item['guid'] = Item::guidFromUri($orig_plink, DI::baseUrl()->getHostname());
 491                                         unset($item['uri']);
 492                                         unset($item['parent-uri']);
 493
 494                                         // Set the delivery priority for "remote self" to "medium"
 495                                         $notify = PRIORITY_MEDIUM;
 496                                 }
 497
 498                                 $id = Item::insert($item, false, $notify);
 499
 500                                 Logger::info("Feed for contact " . $contact["url"] . " stored under id " . $id);
 501                         }
 502                 }
 503
 504                 return ["header" => $author, "items" => $items];
 505         }
 506
 507         private static function titleIsBody($title, $body)
 508         {
 509                 $title = strip_tags($title);
 510                 $title = trim($title);
 511                 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 512                 $title = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $title);
 513
 514                 $body = strip_tags($body);
 515                 $body = trim($body);
 516                 $body = html_entity_decode($body, ENT_QUOTES, 'UTF-8');
 517                 $body = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $body);
 518
 519                 if (strlen($title) < strlen($body)) {
 520                         $body = substr($body, 0, strlen($title));
 521                 }
 522
 523                 if (($title != $body) && (substr($title, -3) == "...")) {
 524                         $pos = strrpos($title, "...");
 525                         if ($pos > 0) {
 526                                 $title = substr($title, 0, $pos);
 527                                 $body = substr($body, 0, $pos);
 528                         }
 529                 }
 530                 return ($title == $body);
 531         }
 532 }