src/Protocol/Feed.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2020, Friendica
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Protocol;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Friendica\Content\Text\HTML;
  27 use Friendica\Core\Logger;
  28 use Friendica\Core\Protocol;
  29 use Friendica\Database\DBA;
  30 use Friendica\DI;
  31 use Friendica\Model\Item;
  32 use Friendica\Model\Tag;
  33 use Friendica\Util\Network;
  34 use Friendica\Util\ParseUrl;
  35 use Friendica\Util\XML;
  36
  37 /**
  38  * This class contain functions to import feeds (RSS/RDF/Atom)
  39  */
  40 class Feed {
  41         /**
  42          * Read a RSS/RDF/Atom feed and create an item entry for it
  43          *
  44          * @param string $xml      The feed data
  45          * @param array  $importer The user record of the importer
  46          * @param array  $contact  The contact record of the feed
  47          *
  48          * @return array Returns the header and the first item in dry run mode
  49          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  50          */
  51         public static function import($xml, array $importer = [], array $contact = [])
  52         {
  53                 $dryRun = empty($importer) && empty($contact);
  54
  55                 if ($dryRun) {
  56                         Logger::info("Test Atom/RSS feed");
  57                 } else {
  58                         Logger::info("Import Atom/RSS feed '" . $contact["name"] . "' (Contact " . $contact["id"] . ") for user " . $importer["uid"]);
  59                 }
  60
  61                 if (empty($xml)) {
  62                         Logger::info('XML is empty.');
  63                         return [];
  64                 }
  65
  66                 if (!empty($contact['poll'])) {
  67                         $basepath = $contact['poll'];
  68                 } elseif (!empty($contact['url'])) {
  69                         $basepath = $contact['url'];
  70                 } else {
  71                         $basepath = '';
  72                 }
  73
  74                 $doc = new DOMDocument();
  75                 @$doc->loadXML(trim($xml));
  76                 $xpath = new DOMXPath($doc);
  77                 $xpath->registerNamespace('atom', ActivityNamespace::ATOM1);
  78                 $xpath->registerNamespace('dc', "http://purl.org/dc/elements/1.1/");
  79                 $xpath->registerNamespace('content', "http://purl.org/rss/1.0/modules/content/");
  80                 $xpath->registerNamespace('rdf', "http://www.w3.org/1999/02/22-rdf-syntax-ns#");
  81                 $xpath->registerNamespace('rss', "http://purl.org/rss/1.0/");
  82                 $xpath->registerNamespace('media', "http://search.yahoo.com/mrss/");
  83                 $xpath->registerNamespace('poco', ActivityNamespace::POCO);
  84
  85                 $author = [];
  86                 $entries = null;
  87
  88                 // Is it RDF?
  89                 if ($xpath->query('/rdf:RDF/rss:channel')->length > 0) {
  90                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:link/text()');
  91                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:title/text()');
  92
  93                         if (empty($author["author-name"])) {
  94                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:description/text()');
  95                         }
  96                         $entries = $xpath->query('/rdf:RDF/rss:item');
  97                 }
  98
  99                 // Is it Atom?
 100                 if ($xpath->query('/atom:feed')->length > 0) {
 101                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']");
 102                         if (is_object($alternate)) {
 103                                 foreach ($alternate AS $attribute) {
 104                                         if ($attribute->name == "href") {
 105                                                 $author["author-link"] = $attribute->textContent;
 106                                         }
 107                                 }
 108                         }
 109
 110                         if (empty($author["author-link"])) {
 111                                 $self = XML::getFirstAttributes($xpath, "atom:link[@rel='self']");
 112                                 if (is_object($self)) {
 113                                         foreach ($self AS $attribute) {
 114                                                 if ($attribute->name == "href") {
 115                                                         $author["author-link"] = $attribute->textContent;
 116                                                 }
 117                                         }
 118                                 }
 119                         }
 120
 121                         if (empty($author["author-link"])) {
 122                                 $author["author-link"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:id/text()');
 123                         }
 124                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:logo/text()');
 125
 126                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:title/text()');
 127
 128                         if (empty($author["author-name"])) {
 129                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:subtitle/text()');
 130                         }
 131
 132                         if (empty($author["author-name"])) {
 133                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:name/text()');
 134                         }
 135
 136                         $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:displayName/text()');
 137                         if ($value != "") {
 138                                 $author["author-name"] = $value;
 139                         }
 140
 141                         if ($dryRun) {
 142                                 $author["author-id"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:id/text()');
 143
 144                                 // See https://tools.ietf.org/html/rfc4287#section-3.2.2
 145                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/atom:uri/text()');
 146                                 if ($value != "") {
 147                                         $author["author-link"] = $value;
 148                                 }
 149
 150                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:preferredUsername/text()');
 151                                 if ($value != "") {
 152                                         $author["author-nick"] = $value;
 153                                 }
 154
 155                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:address/poco:formatted/text()');
 156                                 if ($value != "") {
 157                                         $author["author-location"] = $value;
 158                                 }
 159
 160                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:note/text()');
 161                                 if ($value != "") {
 162                                         $author["author-about"] = $value;
 163                                 }
 164
 165                                 $avatar = XML::getFirstAttributes($xpath, "atom:author/atom:link[@rel='avatar']");
 166                                 if (is_object($avatar)) {
 167                                         foreach ($avatar AS $attribute) {
 168                                                 if ($attribute->name == "href") {
 169                                                         $author["author-avatar"] = $attribute->textContent;
 170                                                 }
 171                                         }
 172                                 }
 173                         }
 174
 175                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:updated/text()');
 176
 177                         $author["app"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:generator/text()');
 178
 179                         $entries = $xpath->query('/atom:feed/atom:entry');
 180                 }
 181
 182                 // Is it RSS?
 183                 if ($xpath->query('/rss/channel')->length > 0) {
 184                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rss/channel/link/text()');
 185
 186                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/title/text()');
 187                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/rss/channel/image/url/text()');
 188
 189                         if (empty($author["author-name"])) {
 190                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/copyright/text()');
 191                         }
 192
 193                         if (empty($author["author-name"])) {
 194                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/description/text()');
 195                         }
 196
 197                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/rss/channel/pubDate/text()');
 198
 199                         $author["app"] = XML::getFirstNodeValue($xpath, '/rss/channel/generator/text()');
 200
 201                         $entries = $xpath->query('/rss/channel/item');
 202                 }
 203
 204                 if (!$dryRun) {
 205                         $author["author-link"] = $contact["url"];
 206
 207                         if (empty($author["author-name"])) {
 208                                 $author["author-name"] = $contact["name"];
 209                         }
 210
 211                         $author["author-avatar"] = $contact["thumb"];
 212
 213                         $author["owner-link"] = $contact["url"];
 214                         $author["owner-name"] = $contact["name"];
 215                         $author["owner-avatar"] = $contact["thumb"];
 216                 }
 217
 218                 $header = [];
 219                 $header["uid"] = $importer["uid"] ?? 0;
 220                 $header["network"] = Protocol::FEED;
 221                 $header["wall"] = 0;
 222                 $header["origin"] = 0;
 223                 $header["gravity"] = GRAVITY_PARENT;
 224                 $header["private"] = Item::PUBLIC;
 225                 $header["verb"] = Activity::POST;
 226                 $header["object-type"] = Activity\ObjectType::NOTE;
 227
 228                 $header["contact-id"] = $contact["id"] ?? 0;
 229
 230                 if (!is_object($entries)) {
 231                         Logger::info("There are no entries in this feed.");
 232                         return [];
 233                 }
 234
 235                 $items = [];
 236
 237                 // Limit the number of items that are about to be fetched
 238                 $total_items = ($entries->length - 1);
 239                 $max_items = DI::config()->get('system', 'max_feed_items');
 240                 if (($max_items > 0) && ($total_items > $max_items)) {
 241                         $total_items = $max_items;
 242                 }
 243
 244                 // Importing older entries first
 245                 for ($i = $total_items; $i >= 0; --$i) {
 246                         $entry = $entries->item($i);
 247
 248                         $item = array_merge($header, $author);
 249
 250                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']", $entry);
 251                         if (!is_object($alternate)) {
 252                                 $alternate = XML::getFirstAttributes($xpath, "atom:link", $entry);
 253                         }
 254                         if (is_object($alternate)) {
 255                                 foreach ($alternate AS $attribute) {
 256                                         if ($attribute->name == "href") {
 257                                                 $item["plink"] = $attribute->textContent;
 258                                         }
 259                                 }
 260                         }
 261
 262                         if (empty($item["plink"])) {
 263                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'link/text()', $entry);
 264                         }
 265
 266                         if (empty($item["plink"])) {
 267                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'rss:link/text()', $entry);
 268                         }
 269
 270                         $item["uri"] = XML::getFirstNodeValue($xpath, 'atom:id/text()', $entry);
 271
 272                         if (empty($item["uri"])) {
 273                                 $item["uri"] = XML::getFirstNodeValue($xpath, 'guid/text()', $entry);
 274                         }
 275
 276                         if (empty($item["uri"])) {
 277                                 $item["uri"] = $item["plink"];
 278                         }
 279
 280                         $orig_plink = $item["plink"];
 281
 282                         $item["plink"] = Network::finalUrl($item["plink"]);
 283
 284                         $item["parent-uri"] = $item["uri"];
 285
 286                         if (!$dryRun) {
 287                                 $condition = ["`uid` = ? AND `uri` = ? AND `network` IN (?, ?)",
 288                                         $importer["uid"], $item["uri"], Protocol::FEED, Protocol::DFRN];
 289                                 $previous = Item::selectFirst(['id'], $condition);
 290                                 if (DBA::isResult($previous)) {
 291                                         Logger::info("Item with uri " . $item["uri"] . " for user " . $importer["uid"] . " already existed under id " . $previous["id"]);
 292                                         continue;
 293                                 }
 294                         }
 295
 296                         $item["title"] = XML::getFirstNodeValue($xpath, 'atom:title/text()', $entry);
 297
 298                         if (empty($item["title"])) {
 299                                 $item["title"] = XML::getFirstNodeValue($xpath, 'title/text()', $entry);
 300                         }
 301                         if (empty($item["title"])) {
 302                                 $item["title"] = XML::getFirstNodeValue($xpath, 'rss:title/text()', $entry);
 303                         }
 304
 305                         $item["title"] = html_entity_decode($item["title"], ENT_QUOTES, 'UTF-8');
 306
 307                         $published = XML::getFirstNodeValue($xpath, 'atom:published/text()', $entry);
 308
 309                         if (empty($published)) {
 310                                 $published = XML::getFirstNodeValue($xpath, 'pubDate/text()', $entry);
 311                         }
 312
 313                         if (empty($published)) {
 314                                 $published = XML::getFirstNodeValue($xpath, 'dc:date/text()', $entry);
 315                         }
 316
 317                         $updated = XML::getFirstNodeValue($xpath, 'atom:updated/text()', $entry);
 318
 319                         if (empty($updated) && !empty($published)) {
 320                                 $updated = $published;
 321                         }
 322
 323                         if (empty($published) && !empty($updated)) {
 324                                 $published = $updated;
 325                         }
 326
 327                         if ($published != "") {
 328                                 $item["created"] = $published;
 329                         }
 330
 331                         if ($updated != "") {
 332                                 $item["edited"] = $updated;
 333                         }
 334
 335                         $creator = XML::getFirstNodeValue($xpath, 'author/text()', $entry);
 336
 337                         if (empty($creator)) {
 338                                 $creator = XML::getFirstNodeValue($xpath, 'atom:author/atom:name/text()', $entry);
 339                         }
 340
 341                         if (empty($creator)) {
 342                                 $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 343                         }
 344
 345                         if ($creator != "") {
 346                                 $item["author-name"] = $creator;
 347                         }
 348
 349                         $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 350
 351                         if ($creator != "") {
 352                                 $item["author-name"] = $creator;
 353                         }
 354
 355                         /// @TODO ?
 356                         // <category>Ausland</category>
 357                         // <media:thumbnail width="152" height="76" url="http://www.taz.de/picture/667875/192/14388767.jpg"/>
 358
 359                         $attachments = [];
 360
 361                         $enclosures = $xpath->query("enclosure|atom:link[@rel='enclosure']", $entry);
 362                         foreach ($enclosures AS $enclosure) {
 363                                 $href = "";
 364                                 $length = "";
 365                                 $type = "";
 366
 367                                 foreach ($enclosure->attributes AS $attribute) {
 368                                         if (in_array($attribute->name, ["url", "href"])) {
 369                                                 $href = $attribute->textContent;
 370                                         } elseif ($attribute->name == "length") {
 371                                                 $length = $attribute->textContent;
 372                                         } elseif ($attribute->name == "type") {
 373                                                 $type = $attribute->textContent;
 374                                         }
 375                                 }
 376
 377                                 if (!empty($item["attach"])) {
 378                                         $item["attach"] .= ',';
 379                                 } else {
 380                                         $item["attach"] = '';
 381                                 }
 382
 383                                 $attachments[] = ["link" => $href, "type" => $type, "length" => $length];
 384
 385                                 $item["attach"] .= '[attach]href="' . $href . '" length="' . $length . '" type="' . $type . '"[/attach]';
 386                         }
 387
 388                         $taglist = [];
 389                         $categories = $xpath->query("category", $entry);
 390                         foreach ($categories AS $category) {
 391                                 $taglist[] = $category->nodeValue;
 392                         }
 393
 394                         $body = trim(XML::getFirstNodeValue($xpath, 'atom:content/text()', $entry));
 395
 396                         if (empty($body)) {
 397                                 $body = trim(XML::getFirstNodeValue($xpath, 'content:encoded/text()', $entry));
 398                         }
 399
 400                         $summary = trim(XML::getFirstNodeValue($xpath, 'atom:summary/text()', $entry));
 401
 402                         if (empty($summary)) {
 403                                 $summary = trim(XML::getFirstNodeValue($xpath, 'description/text()', $entry));
 404                         }
 405
 406                         if (empty($body)) {
 407                                 $body = $summary;
 408                                 $summary = '';
 409                         }
 410
 411                         if ($body == $summary) {
 412                                 $summary = '';
 413                         }
 414
 415                         // remove the content of the title if it is identically to the body
 416                         // This helps with auto generated titles e.g. from tumblr
 417                         if (self::titleIsBody($item["title"], $body)) {
 418                                 $item["title"] = "";
 419                         }
 420                         $item["body"] = HTML::toBBCode($body, $basepath);
 421
 422                         if (($item["body"] == '') && ($item["title"] != '')) {
 423                                 $item["body"] = $item["title"];
 424                                 $item["title"] = '';
 425                         }
 426
 427                         $preview = '';
 428                         if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] < 3)) {
 429                                 // Handle enclosures and treat them as preview picture
 430                                 foreach ($attachments AS $attachment) {
 431                                         if ($attachment["type"] == "image/jpeg") {
 432                                                 $preview = $attachment["link"];
 433                                         }
 434                                 }
 435
 436                                 // Remove a possible link to the item itself
 437                                 $item["body"] = str_replace($item["plink"], '', $item["body"]);
 438                                 $item["body"] = trim(preg_replace('/\[url\=\](\w+.*?)\[\/url\]/i', '', $item["body"]));
 439
 440                                 // Replace the content when the title is longer than the body
 441                                 $replace = (strlen($item["title"]) > strlen($item["body"]));
 442
 443                                 // Replace it, when there is an image in the body
 444                                 if (strstr($item["body"], '[/img]')) {
 445                                         $replace = true;
 446                                 }
 447
 448                                 // Replace it, when there is a link in the body
 449                                 if (strstr($item["body"], '[/url]')) {
 450                                         $replace = true;
 451                                 }
 452
 453                                 if ($replace) {
 454                                         $item["body"] = trim($item["title"]);
 455                                 }
 456
 457                                 $data = ParseUrl::getSiteinfoCached($item['plink'], true);
 458                                 if (!empty($data['text']) && !empty($data['title']) && (mb_strlen($item['body']) < mb_strlen($data['text']))) {
 459                                         // When the fetched page info text is longer than the body, we do try to enhance the body
 460                                         if (!empty($item['body']) && (strpos($data['title'], $item['body']) === false) && (strpos($data['text'], $item['body']) === false)) {
 461                                                 // The body is not part of the fetched page info title or page info text. So we add the text to the body
 462                                                 $item['body'] .= "\n\n" . $data['text'];
 463                                         } else {
 464                                                 // Else we replace the body with the page info text
 465                                                 $item['body'] = $data['text'];
 466                                         }
 467                                 }
 468
 469                                 // We always strip the title since it will be added in the page information
 470                                 $item["title"] = "";
 471                                 $item["body"] = $item["body"] . add_page_info($item["plink"], false, $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 472                                 $taglist = get_page_keywords($item["plink"], $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 473                                 $item["object-type"] = Activity\ObjectType::BOOKMARK;
 474                                 unset($item["attach"]);
 475                         } else {
 476                                 if (!empty($summary)) {
 477                                         $item["body"] = '[abstract]' . HTML::toBBCode($summary, $basepath) . "[/abstract]\n" . $item["body"];
 478                                 }
 479
 480                                 if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] == 3)) {
 481                                         if (empty($taglist)) {
 482                                                 $taglist = get_page_keywords($item["plink"], $preview, true, $contact["ffi_keyword_blacklist"]);
 483                                         }
 484                                         $item["body"] .= "\n" . self::tagToString($taglist);
 485                                 } else {
 486                                         $taglist = [];
 487                                 }
 488
 489                                 // Add the link to the original feed entry if not present in feed
 490                                 if (($item['plink'] != '') && !strstr($item["body"], $item['plink'])) {
 491                                         $item["body"] .= "[hr][url]" . $item['plink'] . "[/url]";
 492                                 }
 493                         }
 494
 495                         if ($dryRun) {
 496                                 $items[] = $item;
 497                                 break;
 498                         } else {
 499                                 Logger::info("Stored feed: " . print_r($item, true));
 500
 501                                 $notify = Item::isRemoteSelf($contact, $item);
 502
 503                                 // Distributed items should have a well formatted URI.
 504                                 // Additionally we have to avoid conflicts with identical URI between imported feeds and these items.
 505                                 if ($notify) {
 506                                         $item['guid'] = Item::guidFromUri($orig_plink, DI::baseUrl()->getHostname());
 507                                         unset($item['uri']);
 508                                         unset($item['parent-uri']);
 509
 510                                         // Set the delivery priority for "remote self" to "medium"
 511                                         $notify = PRIORITY_MEDIUM;
 512                                 }
 513
 514                                 $id = Item::insert($item, $notify);
 515
 516                                 Logger::info("Feed for contact " . $contact["url"] . " stored under id " . $id);
 517
 518                                 if (!empty($id) && !empty($taglist)) {
 519                                         $feeditem = Item::selectFirst(['uri-id'], ['id' => $id]);
 520                                         foreach ($taglist as $tag) {
 521                                                 Tag::store($feeditem['uri-id'], Tag::HASHTAG, $tag);
 522                                         }
 523                                 }
 524                         }
 525                 }
 526
 527                 return ["header" => $author, "items" => $items];
 528         }
 529
 530         /**
 531          * Convert a tag array to a tag string
 532          *
 533          * @param array $tags
 534          * @return string tag string
 535          */
 536         private static function tagToString(array $tags)
 537         {
 538                 $tagstr = '';
 539
 540                 foreach ($tags as $tag) {
 541                         if ($tagstr != "") {
 542                                 $tagstr .= ", ";
 543                         }
 544
 545                         $tagstr .= "#[url=" . DI::baseUrl() . "/search?tag=" . urlencode($tag) . "]" . $tag . "[/url]";
 546                 }
 547
 548                 return $tagstr;
 549         }
 550
 551         private static function titleIsBody($title, $body)
 552         {
 553                 $title = strip_tags($title);
 554                 $title = trim($title);
 555                 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 556                 $title = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $title);
 557
 558                 $body = strip_tags($body);
 559                 $body = trim($body);
 560                 $body = html_entity_decode($body, ENT_QUOTES, 'UTF-8');
 561                 $body = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $body);
 562
 563                 if (strlen($title) < strlen($body)) {
 564                         $body = substr($body, 0, strlen($title));
 565                 }
 566
 567                 if (($title != $body) && (substr($title, -3) == "...")) {
 568                         $pos = strrpos($title, "...");
 569                         if ($pos > 0) {
 570                                 $title = substr($title, 0, $pos);
 571                                 $body = substr($body, 0, $pos);
 572                         }
 573                 }
 574                 return ($title == $body);
 575         }
 576 }