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\Util\Network;
  33 use Friendica\Util\ParseUrl;
  34 use Friendica\Util\XML;
  35
  36 /**
  37  * This class contain functions to import feeds (RSS/RDF/Atom)
  38  */
  39 class Feed {
  40         /**
  41          * Read a RSS/RDF/Atom feed and create an item entry for it
  42          *
  43          * @param string $xml      The feed data
  44          * @param array  $importer The user record of the importer
  45          * @param array  $contact  The contact record of the feed
  46          *
  47          * @return array Returns the header and the first item in dry run mode
  48          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  49          */
  50         public static function import($xml, array $importer = [], array $contact = [])
  51         {
  52                 $dryRun = empty($importer) && empty($contact);
  53
  54                 if ($dryRun) {
  55                         Logger::info("Test Atom/RSS feed");
  56                 } else {
  57                         Logger::info("Import Atom/RSS feed '" . $contact["name"] . "' (Contact " . $contact["id"] . ") for user " . $importer["uid"]);
  58                 }
  59
  60                 if (empty($xml)) {
  61                         Logger::info('XML is empty.');
  62                         return [];
  63                 }
  64
  65                 if (!empty($contact['poll'])) {
  66                         $basepath = $contact['poll'];
  67                 } elseif (!empty($contact['url'])) {
  68                         $basepath = $contact['url'];
  69                 } else {
  70                         $basepath = '';
  71                 }
  72
  73                 $doc = new DOMDocument();
  74                 @$doc->loadXML(trim($xml));
  75                 $xpath = new DOMXPath($doc);
  76                 $xpath->registerNamespace('atom', ActivityNamespace::ATOM1);
  77                 $xpath->registerNamespace('dc', "http://purl.org/dc/elements/1.1/");
  78                 $xpath->registerNamespace('content', "http://purl.org/rss/1.0/modules/content/");
  79                 $xpath->registerNamespace('rdf', "http://www.w3.org/1999/02/22-rdf-syntax-ns#");
  80                 $xpath->registerNamespace('rss', "http://purl.org/rss/1.0/");
  81                 $xpath->registerNamespace('media', "http://search.yahoo.com/mrss/");
  82                 $xpath->registerNamespace('poco', ActivityNamespace::POCO);
  83
  84                 $author = [];
  85                 $entries = null;
  86
  87                 // Is it RDF?
  88                 if ($xpath->query('/rdf:RDF/rss:channel')->length > 0) {
  89                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:link/text()');
  90                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:title/text()');
  91
  92                         if (empty($author["author-name"])) {
  93                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rdf:RDF/rss:channel/rss:description/text()');
  94                         }
  95                         $entries = $xpath->query('/rdf:RDF/rss:item');
  96                 }
  97
  98                 // Is it Atom?
  99                 if ($xpath->query('/atom:feed')->length > 0) {
 100                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']");
 101                         if (is_object($alternate)) {
 102                                 foreach ($alternate AS $attribute) {
 103                                         if ($attribute->name == "href") {
 104                                                 $author["author-link"] = $attribute->textContent;
 105                                         }
 106                                 }
 107                         }
 108
 109                         if (empty($author["author-link"])) {
 110                                 $self = XML::getFirstAttributes($xpath, "atom:link[@rel='self']");
 111                                 if (is_object($self)) {
 112                                         foreach ($self AS $attribute) {
 113                                                 if ($attribute->name == "href") {
 114                                                         $author["author-link"] = $attribute->textContent;
 115                                                 }
 116                                         }
 117                                 }
 118                         }
 119
 120                         if (empty($author["author-link"])) {
 121                                 $author["author-link"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:id/text()');
 122                         }
 123                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:logo/text()');
 124
 125                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:title/text()');
 126
 127                         if (empty($author["author-name"])) {
 128                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:subtitle/text()');
 129                         }
 130
 131                         if (empty($author["author-name"])) {
 132                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:name/text()');
 133                         }
 134
 135                         $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:displayName/text()');
 136                         if ($value != "") {
 137                                 $author["author-name"] = $value;
 138                         }
 139
 140                         if ($dryRun) {
 141                                 $author["author-id"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:author/atom:id/text()');
 142
 143                                 // See https://tools.ietf.org/html/rfc4287#section-3.2.2
 144                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/atom:uri/text()');
 145                                 if ($value != "") {
 146                                         $author["author-link"] = $value;
 147                                 }
 148
 149                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:preferredUsername/text()');
 150                                 if ($value != "") {
 151                                         $author["author-nick"] = $value;
 152                                 }
 153
 154                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:address/poco:formatted/text()');
 155                                 if ($value != "") {
 156                                         $author["author-location"] = $value;
 157                                 }
 158
 159                                 $value = XML::getFirstNodeValue($xpath, 'atom:author/poco:note/text()');
 160                                 if ($value != "") {
 161                                         $author["author-about"] = $value;
 162                                 }
 163
 164                                 $avatar = XML::getFirstAttributes($xpath, "atom:author/atom:link[@rel='avatar']");
 165                                 if (is_object($avatar)) {
 166                                         foreach ($avatar AS $attribute) {
 167                                                 if ($attribute->name == "href") {
 168                                                         $author["author-avatar"] = $attribute->textContent;
 169                                                 }
 170                                         }
 171                                 }
 172                         }
 173
 174                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:updated/text()');
 175
 176                         $author["app"] = XML::getFirstNodeValue($xpath, '/atom:feed/atom:generator/text()');
 177
 178                         $entries = $xpath->query('/atom:feed/atom:entry');
 179                 }
 180
 181                 // Is it RSS?
 182                 if ($xpath->query('/rss/channel')->length > 0) {
 183                         $author["author-link"] = XML::getFirstNodeValue($xpath, '/rss/channel/link/text()');
 184
 185                         $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/title/text()');
 186                         $author["author-avatar"] = XML::getFirstNodeValue($xpath, '/rss/channel/image/url/text()');
 187
 188                         if (empty($author["author-name"])) {
 189                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/copyright/text()');
 190                         }
 191
 192                         if (empty($author["author-name"])) {
 193                                 $author["author-name"] = XML::getFirstNodeValue($xpath, '/rss/channel/description/text()');
 194                         }
 195
 196                         $author["edited"] = $author["created"] = XML::getFirstNodeValue($xpath, '/rss/channel/pubDate/text()');
 197
 198                         $author["app"] = XML::getFirstNodeValue($xpath, '/rss/channel/generator/text()');
 199
 200                         $entries = $xpath->query('/rss/channel/item');
 201                 }
 202
 203                 if (!$dryRun) {
 204                         $author["author-link"] = $contact["url"];
 205
 206                         if (empty($author["author-name"])) {
 207                                 $author["author-name"] = $contact["name"];
 208                         }
 209
 210                         $author["author-avatar"] = $contact["thumb"];
 211
 212                         $author["owner-link"] = $contact["url"];
 213                         $author["owner-name"] = $contact["name"];
 214                         $author["owner-avatar"] = $contact["thumb"];
 215                 }
 216
 217                 $header = [];
 218                 $header["uid"] = $importer["uid"] ?? 0;
 219                 $header["network"] = Protocol::FEED;
 220                 $header["wall"] = 0;
 221                 $header["origin"] = 0;
 222                 $header["gravity"] = GRAVITY_PARENT;
 223                 $header["private"] = 2;
 224                 $header["verb"] = Activity::POST;
 225                 $header["object-type"] = Activity\ObjectType::NOTE;
 226
 227                 $header["contact-id"] = $contact["id"] ?? 0;
 228
 229                 if (!is_object($entries)) {
 230                         Logger::info("There are no entries in this feed.");
 231                         return [];
 232                 }
 233
 234                 $items = [];
 235                 // Importing older entries first
 236                 for ($i = $entries->length - 1; $i >= 0; --$i) {
 237                         $entry = $entries->item($i);
 238
 239                         $item = array_merge($header, $author);
 240
 241                         $alternate = XML::getFirstAttributes($xpath, "atom:link[@rel='alternate']", $entry);
 242                         if (!is_object($alternate)) {
 243                                 $alternate = XML::getFirstAttributes($xpath, "atom:link", $entry);
 244                         }
 245                         if (is_object($alternate)) {
 246                                 foreach ($alternate AS $attribute) {
 247                                         if ($attribute->name == "href") {
 248                                                 $item["plink"] = $attribute->textContent;
 249                                         }
 250                                 }
 251                         }
 252
 253                         if (empty($item["plink"])) {
 254                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'link/text()', $entry);
 255                         }
 256
 257                         if (empty($item["plink"])) {
 258                                 $item["plink"] = XML::getFirstNodeValue($xpath, 'rss:link/text()', $entry);
 259                         }
 260
 261                         $item["uri"] = XML::getFirstNodeValue($xpath, 'atom:id/text()', $entry);
 262
 263                         if (empty($item["uri"])) {
 264                                 $item["uri"] = XML::getFirstNodeValue($xpath, 'guid/text()', $entry);
 265                         }
 266
 267                         if (empty($item["uri"])) {
 268                                 $item["uri"] = $item["plink"];
 269                         }
 270
 271                         $orig_plink = $item["plink"];
 272
 273                         $item["plink"] = Network::finalUrl($item["plink"]);
 274
 275                         $item["parent-uri"] = $item["uri"];
 276
 277                         if (!$dryRun) {
 278                                 $condition = ["`uid` = ? AND `uri` = ? AND `network` IN (?, ?)",
 279                                         $importer["uid"], $item["uri"], Protocol::FEED, Protocol::DFRN];
 280                                 $previous = Item::selectFirst(['id'], $condition);
 281                                 if (DBA::isResult($previous)) {
 282                                         Logger::info("Item with uri " . $item["uri"] . " for user " . $importer["uid"] . " already existed under id " . $previous["id"]);
 283                                         continue;
 284                                 }
 285                         }
 286
 287                         $item["title"] = XML::getFirstNodeValue($xpath, 'atom:title/text()', $entry);
 288
 289                         if (empty($item["title"])) {
 290                                 $item["title"] = XML::getFirstNodeValue($xpath, 'title/text()', $entry);
 291                         }
 292                         if (empty($item["title"])) {
 293                                 $item["title"] = XML::getFirstNodeValue($xpath, 'rss:title/text()', $entry);
 294                         }
 295
 296                         $item["title"] = html_entity_decode($item["title"], ENT_QUOTES, 'UTF-8');
 297
 298                         $published = XML::getFirstNodeValue($xpath, 'atom:published/text()', $entry);
 299
 300                         if (empty($published)) {
 301                                 $published = XML::getFirstNodeValue($xpath, 'pubDate/text()', $entry);
 302                         }
 303
 304                         if (empty($published)) {
 305                                 $published = XML::getFirstNodeValue($xpath, 'dc:date/text()', $entry);
 306                         }
 307
 308                         $updated = XML::getFirstNodeValue($xpath, 'atom:updated/text()', $entry);
 309
 310                         if (empty($updated) && !empty($published)) {
 311                                 $updated = $published;
 312                         }
 313
 314                         if (empty($published) && !empty($updated)) {
 315                                 $published = $updated;
 316                         }
 317
 318                         if ($published != "") {
 319                                 $item["created"] = $published;
 320                         }
 321
 322                         if ($updated != "") {
 323                                 $item["edited"] = $updated;
 324                         }
 325
 326                         $creator = XML::getFirstNodeValue($xpath, 'author/text()', $entry);
 327
 328                         if (empty($creator)) {
 329                                 $creator = XML::getFirstNodeValue($xpath, 'atom:author/atom:name/text()', $entry);
 330                         }
 331
 332                         if (empty($creator)) {
 333                                 $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 334                         }
 335
 336                         if ($creator != "") {
 337                                 $item["author-name"] = $creator;
 338                         }
 339
 340                         $creator = XML::getFirstNodeValue($xpath, 'dc:creator/text()', $entry);
 341
 342                         if ($creator != "") {
 343                                 $item["author-name"] = $creator;
 344                         }
 345
 346                         /// @TODO ?
 347                         // <category>Ausland</category>
 348                         // <media:thumbnail width="152" height="76" url="http://www.taz.de/picture/667875/192/14388767.jpg"/>
 349
 350                         $attachments = [];
 351
 352                         $enclosures = $xpath->query("enclosure|atom:link[@rel='enclosure']", $entry);
 353                         foreach ($enclosures AS $enclosure) {
 354                                 $href = "";
 355                                 $length = "";
 356                                 $type = "";
 357
 358                                 foreach ($enclosure->attributes AS $attribute) {
 359                                         if (in_array($attribute->name, ["url", "href"])) {
 360                                                 $href = $attribute->textContent;
 361                                         } elseif ($attribute->name == "length") {
 362                                                 $length = $attribute->textContent;
 363                                         } elseif ($attribute->name == "type") {
 364                                                 $type = $attribute->textContent;
 365                                         }
 366                                 }
 367
 368                                 if (!empty($item["attach"])) {
 369                                         $item["attach"] .= ',';
 370                                 } else {
 371                                         $item["attach"] = '';
 372                                 }
 373
 374                                 $attachments[] = ["link" => $href, "type" => $type, "length" => $length];
 375
 376                                 $item["attach"] .= '[attach]href="' . $href . '" length="' . $length . '" type="' . $type . '"[/attach]';
 377                         }
 378
 379                         $tags = '';
 380                         $categories = $xpath->query("category", $entry);
 381                         foreach ($categories AS $category) {
 382                                 $hashtag = $category->nodeValue;
 383                                 if ($tags != '') {
 384                                         $tags .= ', ';
 385                                 }
 386
 387                                 $taglink = "#[url=" . DI::baseUrl() . "/search?tag=" . $hashtag . "]" . $hashtag . "[/url]";
 388                                 $tags .= $taglink;
 389                         }
 390
 391                         $body = trim(XML::getFirstNodeValue($xpath, 'atom:content/text()', $entry));
 392
 393                         if (empty($body)) {
 394                                 $body = trim(XML::getFirstNodeValue($xpath, 'content:encoded/text()', $entry));
 395                         }
 396
 397                         $summary = trim(XML::getFirstNodeValue($xpath, 'atom:summary/text()', $entry));
 398
 399                         if (empty($summary)) {
 400                                 $summary = trim(XML::getFirstNodeValue($xpath, 'description/text()', $entry));
 401                         }
 402
 403                         if (empty($body)) {
 404                                 $body = $summary;
 405                                 $summary = '';
 406                         }
 407
 408                         if ($body == $summary) {
 409                                 $summary = '';
 410                         }
 411
 412                         // remove the content of the title if it is identically to the body
 413                         // This helps with auto generated titles e.g. from tumblr
 414                         if (self::titleIsBody($item["title"], $body)) {
 415                                 $item["title"] = "";
 416                         }
 417                         $item["body"] = HTML::toBBCode($body, $basepath);
 418
 419                         if (($item["body"] == '') && ($item["title"] != '')) {
 420                                 $item["body"] = $item["title"];
 421                                 $item["title"] = '';
 422                         }
 423
 424                         $preview = '';
 425                         if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] < 3)) {
 426                                 // Handle enclosures and treat them as preview picture
 427                                 foreach ($attachments AS $attachment) {
 428                                         if ($attachment["type"] == "image/jpeg") {
 429                                                 $preview = $attachment["link"];
 430                                         }
 431                                 }
 432
 433                                 // Remove a possible link to the item itself
 434                                 $item["body"] = str_replace($item["plink"], '', $item["body"]);
 435                                 $item["body"] = trim(preg_replace('/\[url\=\](\w+.*?)\[\/url\]/i', '', $item["body"]));
 436
 437                                 // Replace the content when the title is longer than the body
 438                                 $replace = (strlen($item["title"]) > strlen($item["body"]));
 439
 440                                 // Replace it, when there is an image in the body
 441                                 if (strstr($item["body"], '[/img]')) {
 442                                         $replace = true;
 443                                 }
 444
 445                                 // Replace it, when there is a link in the body
 446                                 if (strstr($item["body"], '[/url]')) {
 447                                         $replace = true;
 448                                 }
 449
 450                                 if ($replace) {
 451                                         $item["body"] = trim($item["title"]);
 452                                 }
 453
 454                                 $data = ParseUrl::getSiteinfoCached($item['plink'], true);
 455                                 if (!empty($data['text']) && !empty($data['title']) && (mb_strlen($item['body']) < mb_strlen($data['text']))) {
 456                                         // When the fetched page info text is longer than the body, we do try to enhance the body
 457                                         if (!empty($item['body']) && (strpos($data['title'], $item['body']) === false) && (strpos($data['text'], $item['body']) === false)) {
 458                                                 // The body is not part of the fetched page info title or page info text. So we add the text to the body
 459                                                 $item['body'] .= "\n\n" . $data['text'];
 460                                         } else {
 461                                                 // Else we replace the body with the page info text
 462                                                 $item['body'] = $data['text'];
 463                                         }
 464                                 }
 465
 466                                 // We always strip the title since it will be added in the page information
 467                                 $item["title"] = "";
 468                                 $item["body"] = $item["body"] . add_page_info($item["plink"], false, $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 469                                 $item["tag"] = add_page_keywords($item["plink"], $preview, ($contact["fetch_further_information"] == 2), $contact["ffi_keyword_blacklist"]);
 470                                 $item["object-type"] = Activity\ObjectType::BOOKMARK;
 471                                 unset($item["attach"]);
 472                         } else {
 473                                 if (!empty($summary)) {
 474                                         $item["body"] = '[abstract]' . HTML::toBBCode($summary, $basepath) . "[/abstract]\n" . $item["body"];
 475                                 }
 476
 477                                 if (!empty($contact["fetch_further_information"]) && ($contact["fetch_further_information"] == 3)) {
 478                                         if (!empty($tags)) {
 479                                                 $item["tag"] = $tags;
 480                                         } else {
 481                                                 // @todo $preview is never set in this case, is it intended? - @MrPetovan 2018-02-13
 482                                                 $item["tag"] = add_page_keywords($item["plink"], $preview, true, $contact["ffi_keyword_blacklist"]);
 483                                         }
 484                                         $item["body"] .= "\n" . $item['tag'];
 485                                 }
 486
 487                                 // Add the link to the original feed entry if not present in feed
 488                                 if (($item['plink'] != '') && !strstr($item["body"], $item['plink'])) {
 489                                         $item["body"] .= "[hr][url]" . $item['plink'] . "[/url]";
 490                                 }
 491                         }
 492
 493                         if ($dryRun) {
 494                                 $items[] = $item;
 495                                 break;
 496                         } else {
 497                                 Logger::info("Stored feed: " . print_r($item, true));
 498
 499                                 $notify = Item::isRemoteSelf($contact, $item);
 500
 501                                 // Distributed items should have a well formatted URI.
 502                                 // Additionally we have to avoid conflicts with identical URI between imported feeds and these items.
 503                                 if ($notify) {
 504                                         $item['guid'] = Item::guidFromUri($orig_plink, DI::baseUrl()->getHostname());
 505                                         unset($item['uri']);
 506                                         unset($item['parent-uri']);
 507
 508                                         // Set the delivery priority for "remote self" to "medium"
 509                                         $notify = PRIORITY_MEDIUM;
 510                                 }
 511
 512                                 $id = Item::insert($item, false, $notify);
 513
 514                                 Logger::info("Feed for contact " . $contact["url"] . " stored under id " . $id);
 515                         }
 516                 }
 517
 518                 return ["header" => $author, "items" => $items];
 519         }
 520
 521         private static function titleIsBody($title, $body)
 522         {
 523                 $title = strip_tags($title);
 524                 $title = trim($title);
 525                 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 526                 $title = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $title);
 527
 528                 $body = strip_tags($body);
 529                 $body = trim($body);
 530                 $body = html_entity_decode($body, ENT_QUOTES, 'UTF-8');
 531                 $body = str_replace(["\n", "\r", "\t", " "], ["", "", "", ""], $body);
 532
 533                 if (strlen($title) < strlen($body)) {
 534                         $body = substr($body, 0, strlen($title));
 535                 }
 536
 537                 if (($title != $body) && (substr($title, -3) == "...")) {
 538                         $pos = strrpos($title, "...");
 539                         if ($pos > 0) {
 540                                 $title = substr($title, 0, $pos);
 541                                 $body = substr($body, 0, $pos);
 542                         }
 543                 }
 544                 return ($title == $body);
 545         }
 546 }