3 * @copyright Copyright (C) 2020, Friendica
5 * @license GNU AGPL version 3 or any later version
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.
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.
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/>.
22 namespace Friendica\Network;
26 use Friendica\Core\Cache\Duration;
27 use Friendica\Core\Hook;
28 use Friendica\Core\Logger;
29 use Friendica\Core\Protocol;
30 use Friendica\Core\System;
31 use Friendica\Database\DBA;
33 use Friendica\Model\Contact;
34 use Friendica\Model\GServer;
35 use Friendica\Model\Profile;
36 use Friendica\Protocol\ActivityNamespace;
37 use Friendica\Protocol\ActivityPub;
38 use Friendica\Protocol\Email;
39 use Friendica\Protocol\Feed;
40 use Friendica\Util\Crypto;
41 use Friendica\Util\Network;
42 use Friendica\Util\Strings;
43 use Friendica\Util\XML;
46 * This class contain functions for probing URL
50 private static $baseurl;
51 private static $istimeout;
54 * Remove stuff from an URI that doesn't belong there
57 * @return string Cleaned URI
59 public static function cleanURI(string $URI)
61 // At first remove leading and trailing junk
62 $URI = trim($URI, "@#?:/ \t\n\r\0\x0B");
64 $parts = parse_url($URI);
66 if (empty($parts['scheme'])) {
70 // Remove the URL fragment, since these shouldn't be part of any profile URL
71 unset($parts['fragment']);
73 $URI = Network::unparseURL($parts);
79 * Rearrange the array so that it always has the same order
81 * @param array $data Unordered data
83 * @return array Ordered data
85 private static function rearrangeData($data)
87 $fields = ["name", "nick", "guid", "url", "addr", "alias", "photo", "account-type",
88 "community", "keywords", "location", "about", "hide",
89 "batch", "notify", "poll", "request", "confirm", "subscribe", "poco",
90 "following", "followers", "inbox", "outbox", "sharedinbox",
91 "priority", "network", "pubkey", "baseurl", "gsid"];
94 foreach ($fields as $field) {
95 if (isset($data[$field])) {
96 if (in_array($field, ["gsid", "hide", "account-type"])) {
97 $newdata[$field] = (int)$data[$field];
99 $newdata[$field] = $data[$field];
101 } elseif ($field != "gsid") {
102 $newdata[$field] = "";
104 $newdata[$field] = null;
108 // We don't use the "priority" field anymore and replace it with a dummy.
109 $newdata["priority"] = 0;
115 * Check if the hostname belongs to the own server
117 * @param string $host The hostname that is to be checked
119 * @return bool Does the testes hostname belongs to the own server?
121 private static function ownHost($host)
123 $own_host = DI::baseUrl()->getHostname();
125 $parts = parse_url($host);
127 if (!isset($parts['scheme'])) {
128 $parts = parse_url('http://'.$host);
131 if (!isset($parts['host'])) {
134 return $parts['host'] == $own_host;
138 * Probes for webfinger path via "host-meta"
140 * We have to check if the servers in the future still will offer this.
141 * It seems as if it was dropped from the standard.
143 * @param string $host The host part of an url
145 * @return array with template and type of the webfinger template for JSON or XML
146 * @throws HTTPException\InternalServerErrorException
148 private static function hostMeta($host)
150 // Reset the static variable
153 // Handles the case when the hostname contains the scheme
154 if (!parse_url($host, PHP_URL_SCHEME)) {
155 $ssl_url = "https://" . $host . "/.well-known/host-meta";
156 $url = "http://" . $host . "/.well-known/host-meta";
158 $ssl_url = $host . "/.well-known/host-meta";
162 $xrd_timeout = DI::config()->get('system', 'xrd_timeout', 20);
164 Logger::info('Probing', ['host' => $host, 'ssl_url' => $ssl_url, 'url' => $url, 'callstack' => System::callstack(20)]);
167 $curlResult = Network::curl($ssl_url, false, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
168 $ssl_connection_error = ($curlResult->getErrorNumber() == CURLE_COULDNT_CONNECT) || ($curlResult->getReturnCode() == 0);
169 if ($curlResult->isSuccess()) {
170 $xml = $curlResult->getBody();
171 $xrd = XML::parseString($xml, true);
173 $host_url = 'https://' . $host;
177 } elseif ($curlResult->isTimeout()) {
178 Logger::info('Probing timeout', ['url' => $ssl_url], Logger::DEBUG);
179 self::$istimeout = true;
183 if (!is_object($xrd) && !empty($url)) {
184 $curlResult = Network::curl($url, false, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
185 $connection_error = ($curlResult->getErrorNumber() == CURLE_COULDNT_CONNECT) || ($curlResult->getReturnCode() == 0);
186 if ($curlResult->isTimeout()) {
187 Logger::info('Probing timeout', ['url' => $url], Logger::DEBUG);
188 self::$istimeout = true;
190 } elseif ($connection_error && $ssl_connection_error) {
191 self::$istimeout = true;
195 $xml = $curlResult->getBody();
196 $xrd = XML::parseString($xml, true);
197 $host_url = 'http://'.$host;
199 if (!is_object($xrd)) {
200 Logger::log("No xrd object found for ".$host, Logger::DEBUG);
204 $links = XML::elementToArray($xrd);
205 if (!isset($links["xrd"]["link"])) {
206 Logger::log("No xrd data found for ".$host, Logger::DEBUG);
210 $lrdd = ['application/jrd+json' => $host_url . '/.well-known/webfinger?resource={uri}'];
212 foreach ($links["xrd"]["link"] as $value => $link) {
213 if (!empty($link["@attributes"])) {
214 $attributes = $link["@attributes"];
215 } elseif ($value == "@attributes") {
221 if (!empty($attributes["rel"]) && $attributes["rel"] == "lrdd" && !empty($attributes["template"])) {
222 $type = (empty($attributes["type"]) ? '' : $attributes["type"]);
224 $lrdd[$type] = $attributes["template"];
228 self::$baseurl = $host_url;
230 Logger::log("Probing successful for ".$host, Logger::DEBUG);
236 * Perform Webfinger lookup and return DFRN data
238 * Given an email style address, perform webfinger lookup and
239 * return the resulting DFRN profile URL, or if no DFRN profile URL
240 * is located, returns an OStatus subscription template (prefixed
241 * with the string 'stat:' to identify it as on OStatus template).
242 * If this isn't an email style address just return $webbie.
243 * Return an empty string if email-style addresses but webfinger fails,
244 * or if the resultant personal XRD doesn't contain a supported
245 * subscription/friend-request attribute.
247 * amended 7/9/2011 to return an hcard which could save potentially loading
248 * a lengthy content page to scrape dfrn attributes
250 * @param string $webbie Address that should be probed
251 * @param string $hcard_url Link to the hcard - is returned by reference
253 * @return string profile link
254 * @throws HTTPException\InternalServerErrorException
256 public static function webfingerDfrn($webbie, &$hcard_url)
260 $links = self::lrdd($webbie);
261 Logger::log('webfingerDfrn: '.$webbie.':'.print_r($links, true), Logger::DATA);
262 if (!empty($links) && is_array($links)) {
263 foreach ($links as $link) {
264 if ($link['@attributes']['rel'] === ActivityNamespace::DFRN) {
265 $profile_link = $link['@attributes']['href'];
267 if (($link['@attributes']['rel'] === ActivityNamespace::OSTATUSSUB) && ($profile_link == "")) {
268 $profile_link = 'stat:'.$link['@attributes']['template'];
270 if ($link['@attributes']['rel'] === 'http://microformats.org/profile/hcard') {
271 $hcard_url = $link['@attributes']['href'];
275 return $profile_link;
279 * Check an URI for LRDD data
281 * @param string $uri Address that should be probed
283 * @return array uri data
284 * @throws HTTPException\InternalServerErrorException
286 public static function lrdd(string $uri)
288 $lrdd = self::hostMeta($uri);
291 if (is_bool($lrdd)) {
296 $parts = @parse_url($uri);
297 if (!$parts || empty($parts["host"]) || empty($parts["path"])) {
301 $host = $parts['scheme'] . '://' . $parts["host"];
302 if (!empty($parts["port"])) {
303 $host .= ':'.$parts["port"];
306 $path_parts = explode("/", trim($parts["path"], "/"));
308 $nick = array_pop($path_parts);
311 $lrdd = self::hostMeta($host);
312 $host .= "/".array_shift($path_parts);
313 } while (!$lrdd && (sizeof($path_parts) > 0));
317 Logger::log("No lrdd data found for ".$uri, Logger::DEBUG);
321 foreach ($lrdd as $type => $template) {
326 $path = str_replace('{uri}', urlencode($uri), $template);
327 $webfinger = self::webfinger($path, $type);
329 if (!$webfinger && (strstr($uri, "@"))) {
330 $path = str_replace('{uri}', urlencode("acct:".$uri), $template);
331 $webfinger = self::webfinger($path, $type);
334 // Special treatment for Mastodon
335 // Problem is that Mastodon uses an URL format like http://domain.tld/@nick
336 // But the webfinger for this format fails.
337 if (!$webfinger && !empty($nick)) {
338 // Mastodon uses a "@" as prefix for usernames in their url format
339 $nick = ltrim($nick, '@');
341 $addr = $nick."@".$host;
343 $path = str_replace('{uri}', urlencode("acct:".$addr), $template);
344 $webfinger = self::webfinger($path, $type);
348 if (empty($webfinger["links"])) {
349 Logger::log("No webfinger links found for ".$uri, Logger::DEBUG);
355 foreach ($webfinger["links"] as $link) {
356 $data[] = ["@attributes" => $link];
359 if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
360 foreach ($webfinger["aliases"] as $alias) {
361 $data[] = ["@attributes" =>
371 * Fetch information (protocol endpoints and user information) about a given uri
373 * @param string $uri Address that should be probed
374 * @param string $network Test for this specific network
375 * @param integer $uid User ID for the probe (only used for mails)
376 * @param boolean $cache Use cached values?
378 * @return array uri data
379 * @throws HTTPException\InternalServerErrorException
380 * @throws \ImagickException
382 public static function uri($uri, $network = '', $uid = -1, $cache = true)
385 $result = DI::cache()->get('Probe::uri:' . $network . ':' . $uri);
386 if (!is_null($result)) {
395 self::$istimeout = false;
397 if ($network != Protocol::ACTIVITYPUB) {
398 $data = self::detect($uri, $network, $uid);
403 // When the previous detection process had got a time out
404 // we could falsely detect a Friendica profile as AP profile.
405 if (!self::$istimeout) {
406 $ap_profile = ActivityPub::probeProfile($uri, !$cache);
408 if (empty($data) || (!empty($ap_profile) && empty($network) && (($data['network'] ?? '') != Protocol::DFRN))) {
410 } elseif (!empty($ap_profile)) {
411 $ap_profile['batch'] = '';
412 $data = array_merge($ap_profile, $data);
415 Logger::notice('Time out detected. AP will not be probed.', ['uri' => $uri]);
418 if (!isset($data['url'])) {
422 if (empty($data['photo'])) {
423 $data['photo'] = DI::baseUrl() . '/images/person-300.jpg';
426 if (empty($data['name'])) {
427 if (!empty($data['nick'])) {
428 $data['name'] = $data['nick'];
431 if (empty($data['name'])) {
432 $data['name'] = $data['url'];
436 if (empty($data['nick'])) {
437 $data['nick'] = strtolower($data['name']);
439 if (strpos($data['nick'], ' ')) {
440 $data['nick'] = trim(substr($data['nick'], 0, strpos($data['nick'], ' ')));
444 if (empty($data['baseurl']) && !empty(self::$baseurl)) {
445 $data['baseurl'] = self::$baseurl;
448 if (!empty($data['baseurl']) && empty($data['gsid'])) {
449 $data['gsid'] = GServer::getID($data['baseurl']);
452 if (empty($data['network'])) {
453 $data['network'] = Protocol::PHANTOM;
456 // Ensure that local connections always are DFRN
457 if (($network == '') && ($data['network'] != Protocol::PHANTOM) && (self::ownHost($data['baseurl'] ?? '') || self::ownHost($data['url']))) {
458 $data['network'] = Protocol::DFRN;
461 if (!isset($data['hide']) && in_array($data['network'], Protocol::FEDERATED)) {
462 $data['hide'] = self::getHideStatus($data['url']);
465 $data = self::rearrangeData($data);
467 // Only store into the cache if the value seems to be valid
468 if (!in_array($data['network'], [Protocol::PHANTOM, Protocol::MAIL])) {
469 DI::cache()->set('Probe::uri:' . $network . ':' . $uri, $data, Duration::DAY);
477 * Fetches the "hide" status from the profile
479 * @param string $url URL of the profile
481 * @return boolean "hide" status
483 private static function getHideStatus($url)
485 $curlResult = Network::curl($url);
486 if (!$curlResult->isSuccess()) {
490 // If the file is too large then exit
491 if (($curlResult->getInfo()['download_content_length'] ?? 0) > 1000000) {
495 // If it isn't a HTML file then exit
496 if (($curlResult->getContentType() != '') && !strstr(strtolower($curlResult->getContentType()), 'html')) {
500 $body = $curlResult->getBody();
502 $doc = new DOMDocument();
503 @$doc->loadHTML($body);
505 $xpath = new DOMXPath($doc);
507 $list = $xpath->query('//meta[@name]');
508 foreach ($list as $node) {
510 if ($node->attributes->length) {
511 foreach ($node->attributes as $attribute) {
512 $meta_tag[$attribute->name] = $attribute->value;
516 if (empty($meta_tag['content'])) {
520 $content = strtolower(trim($meta_tag['content']));
522 switch (strtolower(trim($meta_tag['name']))) {
523 case 'dfrn-global-visibility':
524 if ($content == 'false') {
529 if (strpos($content, 'noindex') !== false) {
540 * Fetch the "subscribe" and add it to the result
542 * @param array $result
543 * @param array $webfinger
544 * @return array result
546 private static function getSubscribeLink(array $result, array $webfinger)
548 if (empty($webfinger['links'])) {
552 foreach ($webfinger['links'] as $link) {
553 if (!empty($link['template']) && ($link['rel'] === ActivityNamespace::OSTATUSSUB)) {
554 $result['subscribe'] = $link['template'];
562 * Fetch information (protocol endpoints and user information) about a given uri
564 * This function is only called by the "uri" function that adds caching and rearranging of data.
566 * @param string $uri Address that should be probed
567 * @param string $network Test for this specific network
568 * @param integer $uid User ID for the probe (only used for mails)
570 * @return array uri data
571 * @throws HTTPException\InternalServerErrorException
573 private static function detect($uri, $network, $uid)
575 $parts = parse_url($uri);
579 'network' => $network,
584 Hook::callAll('probe_detect', $hookData);
586 if ($hookData['result']) {
587 return $hookData['result'];
590 if (!empty($parts["scheme"]) && !empty($parts["host"])) {
591 $host = $parts["host"];
592 if (!empty($parts["port"])) {
593 $host .= ':'.$parts["port"];
596 if ($host == 'twitter.com') {
597 return self::twitter($uri);
599 $lrdd = self::hostMeta($host);
601 if (is_bool($lrdd)) {
605 $path_parts = explode("/", trim($parts['path'] ?? '', "/"));
607 while (!$lrdd && (sizeof($path_parts) > 1)) {
608 $host .= "/".array_shift($path_parts);
609 $lrdd = self::hostMeta($host);
612 Logger::log('No XRD data was found for '.$uri, Logger::DEBUG);
613 return self::feed($uri);
615 $nick = array_pop($path_parts);
617 // Mastodon uses a "@" as prefix for usernames in their url format
618 $nick = ltrim($nick, '@');
620 $addr = $nick."@".$host;
621 } elseif (strstr($uri, '@')) {
622 // If the URI starts with "mailto:" then jump directly to the mail detection
623 if (strpos($uri, 'mailto:') !== false) {
624 $uri = str_replace('mailto:', '', $uri);
625 return self::mail($uri, $uid);
628 if ($network == Protocol::MAIL) {
629 return self::mail($uri, $uid);
631 // Remove "acct:" from the URI
632 $uri = str_replace('acct:', '', $uri);
634 $host = substr($uri, strpos($uri, '@') + 1);
635 $nick = substr($uri, 0, strpos($uri, '@'));
637 if (strpos($uri, '@twitter.com')) {
638 return self::twitter($uri);
640 $lrdd = self::hostMeta($host);
642 if (is_bool($lrdd)) {
647 Logger::log('No XRD data was found for '.$uri, Logger::DEBUG);
648 return self::mail($uri, $uid);
652 Logger::log("Uri ".$uri." was not detectable", Logger::DEBUG);
658 /// @todo Do we need the prefix "acct:" or "acct://"?
660 foreach ($lrdd as $type => $template) {
667 $path = str_replace('{uri}', urlencode($uri), $template);
668 $webfinger = self::webfinger($path, $type);
671 // Then try the address
673 $path = str_replace('{uri}', urlencode("acct:" . $addr), $template);
674 $webfinger = self::webfinger($path, $type);
677 // Finally try without the "acct"
679 $path = str_replace('{uri}', urlencode($addr), $template);
680 $webfinger = self::webfinger($path, $type);
683 // We cannot be sure that the detected address was correct, so we don't use the values
684 if ($webfinger && ($uri != $addr)) {
691 return self::feed($uri);
696 Logger::info("Probing", ['uri' => $uri]);
698 if (in_array($network, ["", Protocol::DFRN])) {
699 $result = self::dfrn($webfinger);
701 if ((!$result && ($network == "")) || ($network == Protocol::DIASPORA)) {
702 $result = self::diaspora($webfinger);
704 if ((!$result && ($network == "")) || ($network == Protocol::OSTATUS)) {
705 $result = self::ostatus($webfinger);
707 if (in_array($network, ['', Protocol::ZOT])) {
708 $result = self::zot($webfinger, $result);
710 if ((!$result && ($network == "")) || ($network == Protocol::PUMPIO)) {
711 $result = self::pumpio($webfinger, $addr);
713 if ((!$result && ($network == "")) || ($network == Protocol::FEED)) {
714 $result = self::feed($uri);
716 // We overwrite the detected nick with our try if the previois routines hadn't detected it.
717 // Additionally it is overwritten when the nickname doesn't make sense (contains spaces).
718 if ((empty($result["nick"]) || (strstr($result["nick"], " "))) && ($nick != "")) {
719 $result["nick"] = $nick;
722 if (empty($result["addr"]) && ($addr != "")) {
723 $result["addr"] = $addr;
727 $result = self::getSubscribeLink($result, $webfinger);
729 if (empty($result["network"])) {
730 $result["network"] = Protocol::PHANTOM;
733 if (empty($result["url"])) {
734 $result["url"] = $uri;
737 Logger::log($uri." is ".$result["network"], Logger::DEBUG);
743 * Check for Zot contact
745 * @param array $webfinger Webfinger data
746 * @param array $data previously probed data
748 * @return array Zot data
749 * @throws HTTPException\InternalServerErrorException
751 private static function zot($webfinger, $data)
753 if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
754 foreach ($webfinger["aliases"] as $alias) {
755 if (substr($alias, 0, 5) == 'acct:') {
756 $data["addr"] = substr($alias, 5);
761 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == "acct:")) {
762 $data["addr"] = substr($webfinger["subject"], 5);
766 foreach ($webfinger['links'] as $link) {
767 if (($link['rel'] == 'http://purl.org/zot/protocol') && !empty($link['href'])) {
768 $zot_url = $link['href'];
772 if (empty($zot_url) && !empty($data['addr']) && !empty(self::$baseurl)) {
773 $condition = ['nurl' => Strings::normaliseLink(self::$baseurl), 'platform' => ['hubzilla']];
774 if (!DBA::exists('gserver', $condition)) {
777 $zot_url = self::$baseurl . '/.well-known/zot-info?address=' . $data['addr'];
780 if (empty($zot_url)) {
784 $data = self::pollZot($zot_url, $data);
786 if (!empty($data['url']) && !empty($webfinger['aliases']) && is_array($webfinger['aliases'])) {
787 foreach ($webfinger['aliases'] as $alias) {
788 if (!strstr($alias, '@') && Strings::normaliseLink($alias) != Strings::normaliseLink($data['url'])) {
789 $data['alias'] = $alias;
797 public static function pollZot($url, $data)
799 $curlResult = Network::curl($url);
800 if ($curlResult->isTimeout()) {
803 $content = $curlResult->getBody();
808 $json = json_decode($content, true);
809 if (!is_array($json)) {
813 if (empty($data['network'])) {
814 if (!empty($json['protocols']) && in_array('zot', $json['protocols'])) {
815 $data['network'] = Protocol::ZOT;
816 } elseif (!isset($json['protocols'])) {
817 $data['network'] = Protocol::ZOT;
821 if (!empty($json['guid']) && empty($data['guid'])) {
822 $data['guid'] = $json['guid'];
824 if (!empty($json['key']) && empty($data['pubkey'])) {
825 $data['pubkey'] = $json['key'];
827 if (!empty($json['name'])) {
828 $data['name'] = $json['name'];
830 if (!empty($json['photo'])) {
831 $data['photo'] = $json['photo'];
832 if (!empty($json['photo_updated'])) {
833 $data['photo'] .= '?rev=' . urlencode($json['photo_updated']);
836 if (!empty($json['address'])) {
837 $data['addr'] = $json['address'];
839 if (!empty($json['url'])) {
840 $data['url'] = $json['url'];
842 if (!empty($json['connections_url'])) {
843 $data['poco'] = $json['connections_url'];
845 if (isset($json['searchable'])) {
846 $data['hide'] = !$json['searchable'];
848 if (!empty($json['public_forum'])) {
849 $data['community'] = $json['public_forum'];
850 $data['account-type'] = Contact::PAGE_COMMUNITY;
853 if (!empty($json['profile'])) {
854 $profile = $json['profile'];
855 if (!empty($profile['description'])) {
856 $data['about'] = $profile['description'];
858 if (!empty($profile['keywords'])) {
859 $keywords = implode(', ', $profile['keywords']);
860 if (!empty($keywords)) {
861 $data['keywords'] = $keywords;
866 if (!empty($profile['region'])) {
867 $loc['region'] = $profile['region'];
869 if (!empty($profile['country'])) {
870 $loc['country-name'] = $profile['country'];
872 $location = Profile::formatLocation($loc);
873 if (!empty($location)) {
874 $data['location'] = $location;
882 * Perform a webfinger request.
884 * For details see RFC 7033: <https://tools.ietf.org/html/rfc7033>
886 * @param string $url Address that should be probed
887 * @param string $type type
889 * @return array webfinger data
890 * @throws HTTPException\InternalServerErrorException
892 public static function webfinger($url, $type)
894 $xrd_timeout = DI::config()->get('system', 'xrd_timeout', 20);
896 $curlResult = Network::curl($url, false, ['timeout' => $xrd_timeout, 'accept_content' => $type]);
897 if ($curlResult->isTimeout()) {
898 self::$istimeout = true;
901 $data = $curlResult->getBody();
903 $webfinger = json_decode($data, true);
904 if (!empty($webfinger)) {
905 if (!isset($webfinger["links"])) {
906 Logger::log("No json webfinger links for ".$url, Logger::DEBUG);
912 // If it is not JSON, maybe it is XML
913 $xrd = XML::parseString($data, true);
914 if (!is_object($xrd)) {
915 Logger::log("No webfinger data retrievable for ".$url, Logger::DEBUG);
919 $xrd_arr = XML::elementToArray($xrd);
920 if (!isset($xrd_arr["xrd"]["link"])) {
921 Logger::log("No XML webfinger links for ".$url, Logger::DEBUG);
927 if (!empty($xrd_arr["xrd"]["subject"])) {
928 $webfinger["subject"] = $xrd_arr["xrd"]["subject"];
931 if (!empty($xrd_arr["xrd"]["alias"])) {
932 $webfinger["aliases"] = $xrd_arr["xrd"]["alias"];
935 $webfinger["links"] = [];
937 foreach ($xrd_arr["xrd"]["link"] as $value => $data) {
938 if (!empty($data["@attributes"])) {
939 $attributes = $data["@attributes"];
940 } elseif ($value == "@attributes") {
946 $webfinger["links"][] = $attributes;
952 * Poll the Friendica specific noscrape page.
954 * "noscrape" is a faster alternative to fetch the data from the hcard.
955 * This functionality was originally created for the directory.
957 * @param string $noscrape_url Link to the noscrape page
958 * @param array $data The already fetched data
960 * @return array noscrape data
961 * @throws HTTPException\InternalServerErrorException
963 private static function pollNoscrape($noscrape_url, $data)
965 $curlResult = Network::curl($noscrape_url);
966 if ($curlResult->isTimeout()) {
967 self::$istimeout = true;
970 $content = $curlResult->getBody();
972 Logger::log("Empty body for ".$noscrape_url, Logger::DEBUG);
976 $json = json_decode($content, true);
977 if (!is_array($json)) {
978 Logger::log("No json data for ".$noscrape_url, Logger::DEBUG);
982 if (!empty($json["fn"])) {
983 $data["name"] = $json["fn"];
986 if (!empty($json["addr"])) {
987 $data["addr"] = $json["addr"];
990 if (!empty($json["nick"])) {
991 $data["nick"] = $json["nick"];
994 if (!empty($json["guid"])) {
995 $data["guid"] = $json["guid"];
998 if (!empty($json["comm"])) {
999 $data["community"] = $json["comm"];
1002 if (!empty($json["tags"])) {
1003 $keywords = implode(", ", $json["tags"]);
1004 if ($keywords != "") {
1005 $data["keywords"] = $keywords;
1009 $location = Profile::formatLocation($json);
1011 $data["location"] = $location;
1014 if (!empty($json["about"])) {
1015 $data["about"] = $json["about"];
1018 if (!empty($json["key"])) {
1019 $data["pubkey"] = $json["key"];
1022 if (!empty($json["photo"])) {
1023 $data["photo"] = $json["photo"];
1026 if (!empty($json["dfrn-request"])) {
1027 $data["request"] = $json["dfrn-request"];
1030 if (!empty($json["dfrn-confirm"])) {
1031 $data["confirm"] = $json["dfrn-confirm"];
1034 if (!empty($json["dfrn-notify"])) {
1035 $data["notify"] = $json["dfrn-notify"];
1038 if (!empty($json["dfrn-poll"])) {
1039 $data["poll"] = $json["dfrn-poll"];
1042 if (isset($json["hide"])) {
1043 $data["hide"] = (bool)$json["hide"];
1045 $data["hide"] = false;
1052 * Check for valid DFRN data
1054 * @param array $data DFRN data
1056 * @return int Number of errors
1058 public static function validDfrn($data)
1061 if (!isset($data['key'])) {
1064 if (!isset($data['dfrn-request'])) {
1067 if (!isset($data['dfrn-confirm'])) {
1070 if (!isset($data['dfrn-notify'])) {
1073 if (!isset($data['dfrn-poll'])) {
1080 * Fetch data from a DFRN profile page and via "noscrape"
1082 * @param string $profile_link Link to the profile page
1084 * @return array profile data
1085 * @throws HTTPException\InternalServerErrorException
1086 * @throws \ImagickException
1088 public static function profile($profile_link)
1092 Logger::log("Check profile ".$profile_link, Logger::DEBUG);
1094 // Fetch data via noscrape - this is faster
1095 $noscrape_url = str_replace(["/hcard/", "/profile/"], "/noscrape/", $profile_link);
1096 $data = self::pollNoscrape($noscrape_url, $data);
1098 if (!isset($data["notify"])
1099 || !isset($data["confirm"])
1100 || !isset($data["request"])
1101 || !isset($data["poll"])
1102 || !isset($data["name"])
1103 || !isset($data["photo"])
1105 $data = self::pollHcard($profile_link, $data, true);
1110 if (empty($data["addr"]) || empty($data["nick"])) {
1111 $probe_data = self::uri($profile_link);
1112 $data["addr"] = ($data["addr"] ?? '') ?: $probe_data["addr"];
1113 $data["nick"] = ($data["nick"] ?? '') ?: $probe_data["nick"];
1116 $prof_data["addr"] = $data["addr"];
1117 $prof_data["nick"] = $data["nick"];
1118 $prof_data["dfrn-request"] = $data['request'] ?? null;
1119 $prof_data["dfrn-confirm"] = $data['confirm'] ?? null;
1120 $prof_data["dfrn-notify"] = $data['notify'] ?? null;
1121 $prof_data["dfrn-poll"] = $data['poll'] ?? null;
1122 $prof_data["photo"] = $data['photo'] ?? null;
1123 $prof_data["fn"] = $data['name'] ?? null;
1124 $prof_data["key"] = $data['pubkey'] ?? null;
1126 Logger::log("Result for profile ".$profile_link.": ".print_r($prof_data, true), Logger::DEBUG);
1132 * Check for DFRN contact
1134 * @param array $webfinger Webfinger data
1136 * @return array DFRN data
1137 * @throws HTTPException\InternalServerErrorException
1139 private static function dfrn($webfinger)
1143 // The array is reversed to take into account the order of preference for same-rel links
1144 // See: https://tools.ietf.org/html/rfc7033#section-4.4.4
1145 foreach (array_reverse($webfinger["links"]) as $link) {
1146 if (($link["rel"] == ActivityNamespace::DFRN) && !empty($link["href"])) {
1147 $data["network"] = Protocol::DFRN;
1148 } elseif (($link["rel"] == ActivityNamespace::FEED) && !empty($link["href"])) {
1149 $data["poll"] = $link["href"];
1150 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && (($link["type"] ?? "") == "text/html") && !empty($link["href"])) {
1151 $data["url"] = $link["href"];
1152 } elseif (($link["rel"] == "http://microformats.org/profile/hcard") && !empty($link["href"])) {
1153 $hcard_url = $link["href"];
1154 } elseif (($link["rel"] == ActivityNamespace::POCO) && !empty($link["href"])) {
1155 $data["poco"] = $link["href"];
1156 } elseif (($link["rel"] == "http://webfinger.net/rel/avatar") && !empty($link["href"])) {
1157 $data["photo"] = $link["href"];
1158 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && !empty($link["href"])) {
1159 $data["baseurl"] = trim($link["href"], '/');
1160 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && !empty($link["href"])) {
1161 $data["guid"] = $link["href"];
1162 } elseif (($link["rel"] == "diaspora-public-key") && !empty($link["href"])) {
1163 $data["pubkey"] = base64_decode($link["href"]);
1165 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
1166 if (strstr($data["pubkey"], 'RSA ')) {
1167 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1172 if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
1173 foreach ($webfinger["aliases"] as $alias) {
1174 if (empty($data["url"]) && !strstr($alias, "@")) {
1175 $data["url"] = $alias;
1176 } elseif (!strstr($alias, "@") && Strings::normaliseLink($alias) != Strings::normaliseLink($data["url"])) {
1177 $data["alias"] = $alias;
1178 } elseif (substr($alias, 0, 5) == 'acct:') {
1179 $data["addr"] = substr($alias, 5);
1184 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == "acct:")) {
1185 $data["addr"] = substr($webfinger["subject"], 5);
1188 if (!isset($data["network"]) || ($hcard_url == "")) {
1192 // Fetch data via noscrape - this is faster
1193 $noscrape_url = str_replace("/hcard/", "/noscrape/", $hcard_url);
1194 $data = self::pollNoscrape($noscrape_url, $data);
1196 if (isset($data["notify"])
1197 && isset($data["confirm"])
1198 && isset($data["request"])
1199 && isset($data["poll"])
1200 && isset($data["name"])
1201 && isset($data["photo"])
1206 $data = self::pollHcard($hcard_url, $data, true);
1212 * Poll the hcard page (Diaspora and Friendica specific)
1214 * @param string $hcard_url Link to the hcard page
1215 * @param array $data The already fetched data
1216 * @param boolean $dfrn Poll DFRN specific data
1218 * @return array hcard data
1219 * @throws HTTPException\InternalServerErrorException
1221 private static function pollHcard($hcard_url, $data, $dfrn = false)
1223 $curlResult = Network::curl($hcard_url);
1224 if ($curlResult->isTimeout()) {
1225 self::$istimeout = true;
1228 $content = $curlResult->getBody();
1233 $doc = new DOMDocument();
1234 if (!@$doc->loadHTML($content)) {
1238 $xpath = new DomXPath($doc);
1240 $vcards = $xpath->query("//div[contains(concat(' ', @class, ' '), ' vcard ')]");
1241 if (!is_object($vcards)) {
1245 if (!isset($data["baseurl"])) {
1246 $data["baseurl"] = "";
1249 if ($vcards->length > 0) {
1250 $vcard = $vcards->item(0);
1252 // We have to discard the guid from the hcard in favour of the guid from lrdd
1253 // Reason: Hubzilla doesn't use the value "uid" in the hcard like Diaspora does.
1254 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' uid ')]", $vcard); // */
1255 if (($search->length > 0) && empty($data["guid"])) {
1256 $data["guid"] = $search->item(0)->nodeValue;
1259 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' nickname ')]", $vcard); // */
1260 if ($search->length > 0) {
1261 $data["nick"] = $search->item(0)->nodeValue;
1264 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' fn ')]", $vcard); // */
1265 if ($search->length > 0) {
1266 $data["name"] = $search->item(0)->nodeValue;
1269 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' searchable ')]", $vcard); // */
1270 if ($search->length > 0) {
1271 $data["searchable"] = $search->item(0)->nodeValue;
1274 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' key ')]", $vcard); // */
1275 if ($search->length > 0) {
1276 $data["pubkey"] = $search->item(0)->nodeValue;
1277 if (strstr($data["pubkey"], 'RSA ')) {
1278 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1282 $search = $xpath->query("//*[@id='pod_location']", $vcard); // */
1283 if ($search->length > 0) {
1284 $data["baseurl"] = trim($search->item(0)->nodeValue, "/");
1289 if (!empty($vcard)) {
1290 $photos = $xpath->query("//*[contains(concat(' ', @class, ' '), ' photo ') or contains(concat(' ', @class, ' '), ' avatar ')]", $vcard); // */
1291 foreach ($photos as $photo) {
1293 foreach ($photo->attributes as $attribute) {
1294 $attr[$attribute->name] = trim($attribute->value);
1297 if (isset($attr["src"]) && isset($attr["width"])) {
1298 $avatar[$attr["width"]] = $attr["src"];
1301 // We don't have a width. So we just take everything that we got.
1302 // This is a Hubzilla workaround which doesn't send a width.
1303 if ((sizeof($avatar) == 0) && !empty($attr["src"])) {
1304 $avatar[] = $attr["src"];
1309 if (sizeof($avatar)) {
1311 $data["photo"] = self::fixAvatar(array_pop($avatar), $data["baseurl"]);
1315 // Poll DFRN specific data
1316 $search = $xpath->query("//link[contains(concat(' ', @rel), ' dfrn-')]");
1317 if ($search->length > 0) {
1318 foreach ($search as $link) {
1319 //$data["request"] = $search->item(0)->nodeValue;
1321 foreach ($link->attributes as $attribute) {
1322 $attr[$attribute->name] = trim($attribute->value);
1325 $data[substr($attr["rel"], 5)] = $attr["href"];
1329 // Older Friendica versions had used the "uid" field differently than newer versions
1330 if (!empty($data["nick"]) && !empty($data["guid"]) && ($data["nick"] == $data["guid"])) {
1331 unset($data["guid"]);
1340 * Check for Diaspora contact
1342 * @param array $webfinger Webfinger data
1344 * @return array Diaspora data
1345 * @throws HTTPException\InternalServerErrorException
1347 private static function diaspora($webfinger)
1352 // The array is reversed to take into account the order of preference for same-rel links
1353 // See: https://tools.ietf.org/html/rfc7033#section-4.4.4
1354 foreach (array_reverse($webfinger["links"]) as $link) {
1355 if (($link["rel"] == "http://microformats.org/profile/hcard") && !empty($link["href"])) {
1356 $hcard_url = $link["href"];
1357 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && !empty($link["href"])) {
1358 $data["baseurl"] = trim($link["href"], '/');
1359 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && !empty($link["href"])) {
1360 $data["guid"] = $link["href"];
1361 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && (($link["type"] ?? "") == "text/html") && !empty($link["href"])) {
1362 $data["url"] = $link["href"];
1363 } elseif (($link["rel"] == ActivityNamespace::FEED) && !empty($link["href"])) {
1364 $data["poll"] = $link["href"];
1365 } elseif (($link["rel"] == ActivityNamespace::POCO) && !empty($link["href"])) {
1366 $data["poco"] = $link["href"];
1367 } elseif (($link["rel"] == "salmon") && !empty($link["href"])) {
1368 $data["notify"] = $link["href"];
1369 } elseif (($link["rel"] == "diaspora-public-key") && !empty($link["href"])) {
1370 $data["pubkey"] = base64_decode($link["href"]);
1372 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
1373 if (strstr($data["pubkey"], 'RSA ')) {
1374 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1379 if (empty($data["url"]) || empty($hcard_url)) {
1383 if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
1384 foreach ($webfinger["aliases"] as $alias) {
1385 if (Strings::normaliseLink($alias) != Strings::normaliseLink($data["url"]) && ! strstr($alias, "@")) {
1386 $data["alias"] = $alias;
1387 } elseif (substr($alias, 0, 5) == 'acct:') {
1388 $data["addr"] = substr($alias, 5);
1393 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == 'acct:')) {
1394 $data["addr"] = substr($webfinger["subject"], 5);
1397 // Fetch further information from the hcard
1398 $data = self::pollHcard($hcard_url, $data);
1404 if (!empty($data["url"])
1405 && !empty($data["guid"])
1406 && !empty($data["baseurl"])
1407 && !empty($data["pubkey"])
1408 && !empty($hcard_url)
1410 $data["network"] = Protocol::DIASPORA;
1412 // The Diaspora handle must always be lowercase
1413 if (!empty($data["addr"])) {
1414 $data["addr"] = strtolower($data["addr"]);
1417 // We have to overwrite the detected value for "notify" since Hubzilla doesn't send it
1418 $data["notify"] = $data["baseurl"] . "/receive/users/" . $data["guid"];
1419 $data["batch"] = $data["baseurl"] . "/receive/public";
1428 * Check for OStatus contact
1430 * @param array $webfinger Webfinger data
1431 * @param bool $short Short detection mode
1433 * @return array|bool OStatus data or "false" on error or "true" on short mode
1434 * @throws HTTPException\InternalServerErrorException
1436 private static function ostatus($webfinger, $short = false)
1440 if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
1441 foreach ($webfinger["aliases"] as $alias) {
1442 if (strstr($alias, "@") && !strstr(Strings::normaliseLink($alias), "http://")) {
1443 $data["addr"] = str_replace('acct:', '', $alias);
1448 if (!empty($webfinger["subject"]) && strstr($webfinger["subject"], "@")
1449 && !strstr(Strings::normaliseLink($webfinger["subject"]), "http://")
1451 $data["addr"] = str_replace('acct:', '', $webfinger["subject"]);
1454 if (!empty($webfinger["links"])) {
1455 // The array is reversed to take into account the order of preference for same-rel links
1456 // See: https://tools.ietf.org/html/rfc7033#section-4.4.4
1457 foreach (array_reverse($webfinger["links"]) as $link) {
1458 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1459 && (($link["type"] ?? "") == "text/html")
1460 && ($link["href"] != "")
1462 $data["url"] = $data["alias"] = $link["href"];
1463 } elseif (($link["rel"] == "salmon") && !empty($link["href"])) {
1464 $data["notify"] = $link["href"];
1465 } elseif (($link["rel"] == ActivityNamespace::FEED) && !empty($link["href"])) {
1466 $data["poll"] = $link["href"];
1467 } elseif (($link["rel"] == "magic-public-key") && !empty($link["href"])) {
1468 $pubkey = $link["href"];
1470 if (substr($pubkey, 0, 5) === 'data:') {
1471 if (strstr($pubkey, ',')) {
1472 $pubkey = substr($pubkey, strpos($pubkey, ',') + 1);
1474 $pubkey = substr($pubkey, 5);
1476 } elseif (Strings::normaliseLink($pubkey) == 'http://') {
1477 $curlResult = Network::curl($pubkey);
1478 if ($curlResult->isTimeout()) {
1479 self::$istimeout = true;
1480 return $short ? false : [];
1482 $pubkey = $curlResult->getBody();
1485 $key = explode(".", $pubkey);
1487 if (sizeof($key) >= 3) {
1488 $m = Strings::base64UrlDecode($key[1]);
1489 $e = Strings::base64UrlDecode($key[2]);
1490 $data["pubkey"] = Crypto::meToPem($m, $e);
1496 if (isset($data["notify"]) && isset($data["pubkey"])
1497 && isset($data["poll"])
1498 && isset($data["url"])
1500 $data["network"] = Protocol::OSTATUS;
1502 return $short ? false : [];
1509 // Fetch all additional data from the feed
1510 $curlResult = Network::curl($data["poll"]);
1511 if ($curlResult->isTimeout()) {
1512 self::$istimeout = true;
1515 $feed = $curlResult->getBody();
1516 $feed_data = Feed::import($feed);
1521 if (!empty($feed_data["header"]["author-name"])) {
1522 $data["name"] = $feed_data["header"]["author-name"];
1524 if (!empty($feed_data["header"]["author-nick"])) {
1525 $data["nick"] = $feed_data["header"]["author-nick"];
1527 if (!empty($feed_data["header"]["author-avatar"])) {
1528 $data["photo"] = self::fixAvatar($feed_data["header"]["author-avatar"], $data["url"]);
1530 if (!empty($feed_data["header"]["author-id"])) {
1531 $data["alias"] = $feed_data["header"]["author-id"];
1533 if (!empty($feed_data["header"]["author-location"])) {
1534 $data["location"] = $feed_data["header"]["author-location"];
1536 if (!empty($feed_data["header"]["author-about"])) {
1537 $data["about"] = $feed_data["header"]["author-about"];
1539 // OStatus has serious issues when the the url doesn't fit (ssl vs. non ssl)
1540 // So we take the value that we just fetched, although the other one worked as well
1541 if (!empty($feed_data["header"]["author-link"])) {
1542 $data["url"] = $feed_data["header"]["author-link"];
1545 if ($data["url"] == $data["alias"]) {
1546 $data["alias"] = '';
1549 /// @todo Fetch location and "about" from the feed as well
1554 * Fetch data from a pump.io profile page
1556 * @param string $profile_link Link to the profile page
1558 * @return array profile data
1560 private static function pumpioProfileData($profile_link)
1562 $curlResult = Network::curl($profile_link);
1563 if (!$curlResult->isSuccess()) {
1567 $doc = new DOMDocument();
1568 if (!@$doc->loadHTML($curlResult->getBody())) {
1572 $xpath = new DomXPath($doc);
1576 $data["name"] = $xpath->query("//span[contains(@class, 'p-name')]")->item(0)->nodeValue;
1578 if ($data["name"] == '') {
1579 // This is ugly - but pump.io doesn't seem to know a better way for it
1580 $data["name"] = trim($xpath->query("//h1[@class='media-header']")->item(0)->nodeValue);
1581 $pos = strpos($data["name"], chr(10));
1583 $data["name"] = trim(substr($data["name"], 0, $pos));
1587 $data["location"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'p-locality')]");
1589 if ($data["location"] == '') {
1590 $data["location"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'location')]");
1593 $data["about"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'p-note')]");
1595 if ($data["about"] == '') {
1596 $data["about"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'summary')]");
1599 $avatar = $xpath->query("//img[contains(@class, 'u-photo')]")->item(0);
1601 $avatar = $xpath->query("//img[@class='img-rounded media-object']")->item(0);
1604 foreach ($avatar->attributes as $attribute) {
1605 if ($attribute->name == "src") {
1606 $data["photo"] = trim($attribute->value);
1615 * Check for pump.io contact
1617 * @param array $webfinger Webfinger data
1618 * @param string $addr
1619 * @return array pump.io data
1621 private static function pumpio($webfinger, $addr)
1624 // The array is reversed to take into account the order of preference for same-rel links
1625 // See: https://tools.ietf.org/html/rfc7033#section-4.4.4
1626 foreach (array_reverse($webfinger["links"]) as $link) {
1627 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1628 && (($link["type"] ?? "") == "text/html")
1629 && ($link["href"] != "")
1631 $data["url"] = $link["href"];
1632 } elseif (($link["rel"] == "activity-inbox") && ($link["href"] != "")) {
1633 $data["notify"] = $link["href"];
1634 } elseif (($link["rel"] == "activity-outbox") && ($link["href"] != "")) {
1635 $data["poll"] = $link["href"];
1636 } elseif (($link["rel"] == "dialback") && ($link["href"] != "")) {
1637 $data["dialback"] = $link["href"];
1640 if (isset($data["poll"]) && isset($data["notify"])
1641 && isset($data["dialback"])
1642 && isset($data["url"])
1644 // by now we use these fields only for the network type detection
1645 // So we unset all data that isn't used at the moment
1646 unset($data["dialback"]);
1648 $data["network"] = Protocol::PUMPIO;
1653 $profile_data = self::pumpioProfileData($data["url"]);
1655 if (!$profile_data) {
1659 $data = array_merge($data, $profile_data);
1661 if (($addr != '') && ($data['name'] != '')) {
1662 $name = trim(str_replace($addr, '', $data['name']));
1664 $data['name'] = $name;
1672 * Check for twitter contact
1674 * @param string $uri
1676 * @return array twitter data
1678 private static function twitter($uri)
1680 if (preg_match('=(.*)@twitter.com=i', $uri, $matches)) {
1681 $nick = $matches[1];
1682 } elseif (preg_match('=https?://twitter.com/(.*)=i', $uri, $matches)) {
1683 $nick = $matches[1];
1689 $data['url'] = 'https://twitter.com/' . $nick;
1690 $data['addr'] = $nick . '@twitter.com';
1691 $data['nick'] = $data['name'] = $nick;
1692 $data['network'] = Protocol::TWITTER;
1693 $data['baseurl'] = 'https://twitter.com';
1699 * Checks HTML page for RSS feed link
1701 * @param string $url Page link
1702 * @param string $body Page body string
1703 * @return string|false Feed link or false if body was invalid HTML document
1705 public static function getFeedLink(string $url, string $body)
1707 $doc = new DOMDocument();
1708 if (!@$doc->loadHTML($body)) {
1712 $xpath = new DOMXPath($doc);
1714 $feedUrl = $xpath->evaluate('string(/html/head/link[@type="application/rss+xml" and @rel="alternate"]/@href)');
1716 $feedUrl = $feedUrl ? self::ensureAbsoluteLinkFromHTMLDoc($feedUrl, $url, $xpath) : '';
1722 * Return an absolute URL in the context of a HTML document retrieved from the provided URL.
1724 * Loosely based on RFC 1808
1726 * @see https://tools.ietf.org/html/rfc1808
1728 * @param string $href The potential relative href found in the HTML document
1729 * @param string $base The HTML document URL
1730 * @param DOMXPath $xpath The HTML document XPath
1733 private static function ensureAbsoluteLinkFromHTMLDoc(string $href, string $base, DOMXPath $xpath)
1735 if (filter_var($href, FILTER_VALIDATE_URL)) {
1739 $base = $xpath->evaluate('string(/html/head/base/@href)') ?: $base;
1741 $baseParts = parse_url($base);
1743 // Naked domain case (scheme://basehost)
1744 $path = $baseParts['path'] ?? '/';
1746 // Remove the filename part of the path if it exists (/base/path/file)
1747 $path = implode('/', array_slice(explode('/', $path), 0, -1));
1749 $hrefParts = parse_url($href);
1751 // Root path case (/path) including relative scheme case (//host/path)
1752 if ($hrefParts['path'] && $hrefParts['path'][0] == '/') {
1753 $path = $hrefParts['path'];
1755 $path = $path . '/' . $hrefParts['path'];
1757 // Resolve arbitrary relative path
1758 // Lifted from https://www.php.net/manual/en/function.realpath.php#84012
1759 $parts = array_filter(explode('/', $path), 'strlen');
1760 $absolutes = array();
1761 foreach ($parts as $part) {
1762 if ('.' == $part) continue;
1763 if ('..' == $part) {
1764 array_pop($absolutes);
1766 $absolutes[] = $part;
1770 $path = '/' . implode('/', $absolutes);
1773 // Relative scheme case (//host/path)
1774 $baseParts['host'] = $hrefParts['host'] ?? $baseParts['host'];
1775 $baseParts['path'] = $path;
1776 unset($baseParts['query']);
1777 unset($baseParts['fragment']);
1779 return Network::unparseURL($baseParts);
1783 * Check for feed contact
1785 * @param string $url Profile link
1786 * @param boolean $probe Do a probe if the page contains a feed link
1788 * @return array feed data
1789 * @throws HTTPException\InternalServerErrorException
1791 private static function feed($url, $probe = true)
1793 $curlResult = Network::curl($url);
1794 if ($curlResult->isTimeout()) {
1795 self::$istimeout = true;
1798 $feed = $curlResult->getBody();
1799 $feed_data = Feed::import($feed);
1806 $feed_url = self::getFeedLink($url, $feed);
1812 return self::feed($feed_url, false);
1815 if (!empty($feed_data["header"]["author-name"])) {
1816 $data["name"] = $feed_data["header"]["author-name"];
1819 if (!empty($feed_data["header"]["author-nick"])) {
1820 $data["nick"] = $feed_data["header"]["author-nick"];
1823 if (!empty($feed_data["header"]["author-avatar"])) {
1824 $data["photo"] = $feed_data["header"]["author-avatar"];
1827 if (!empty($feed_data["header"]["author-id"])) {
1828 $data["alias"] = $feed_data["header"]["author-id"];
1831 $data["url"] = $url;
1832 $data["poll"] = $url;
1834 if (!empty($feed_data["header"]["author-link"])) {
1835 $data["baseurl"] = $feed_data["header"]["author-link"];
1837 $data["baseurl"] = $data["url"];
1840 $data["network"] = Protocol::FEED;
1846 * Check for mail contact
1848 * @param string $uri Profile link
1849 * @param integer $uid User ID
1851 * @return array mail data
1852 * @throws \Exception
1854 private static function mail($uri, $uid)
1856 if (!Network::isEmailDomainValid($uri)) {
1864 $user = DBA::selectFirst('user', ['prvkey'], ['uid' => $uid]);
1866 $condition = ["`uid` = ? AND `server` != ''", $uid];
1867 $fields = ['pass', 'user', 'server', 'port', 'ssltype', 'mailbox'];
1868 $mailacct = DBA::selectFirst('mailacct', $fields, $condition);
1870 if (!DBA::isResult($user) || !DBA::isResult($mailacct)) {
1874 $mailbox = Email::constructMailboxName($mailacct);
1876 openssl_private_decrypt(hex2bin($mailacct['pass']), $password, $user['prvkey']);
1877 $mbox = Email::connect($mailbox, $mailacct['user'], $password);
1882 $msgs = Email::poll($mbox, $uri);
1883 Logger::log('searching '.$uri.', '.count($msgs).' messages found.', Logger::DEBUG);
1885 if (!count($msgs)) {
1889 $phost = substr($uri, strpos($uri, '@') + 1);
1892 $data["addr"] = $uri;
1893 $data["network"] = Protocol::MAIL;
1894 $data["name"] = substr($uri, 0, strpos($uri, '@'));
1895 $data["nick"] = $data["name"];
1896 $data["photo"] = Network::lookupAvatarByEmail($uri);
1897 $data["url"] = 'mailto:'.$uri;
1898 $data["notify"] = 'smtp ' . Strings::getRandomHex();
1899 $data["poll"] = 'email ' . Strings::getRandomHex();
1901 $x = Email::messageMeta($mbox, $msgs[0]);
1902 if (stristr($x[0]->from, $uri)) {
1903 $adr = imap_rfc822_parse_adrlist($x[0]->from, '');
1904 } elseif (stristr($x[0]->to, $uri)) {
1905 $adr = imap_rfc822_parse_adrlist($x[0]->to, '');
1908 foreach ($adr as $feadr) {
1909 if ((strcasecmp($feadr->mailbox, $data["name"]) == 0)
1910 &&(strcasecmp($feadr->host, $phost) == 0)
1911 && (strlen($feadr->personal))
1913 $personal = imap_mime_header_decode($feadr->personal);
1915 foreach ($personal as $perspart) {
1916 if ($perspart->charset != "default") {
1917 $data["name"] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
1919 $data["name"] .= $perspart->text;
1923 $data["name"] = Strings::escapeTags($data["name"]);
1927 if (!empty($mbox)) {
1934 * Mix two paths together to possibly fix missing parts
1936 * @param string $avatar Path to the avatar
1937 * @param string $base Another path that is hopefully complete
1939 * @return string fixed avatar path
1940 * @throws \Exception
1942 public static function fixAvatar($avatar, $base)
1944 $base_parts = parse_url($base);
1946 // Remove all parts that could create a problem
1947 unset($base_parts['path']);
1948 unset($base_parts['query']);
1949 unset($base_parts['fragment']);
1951 $avatar_parts = parse_url($avatar);
1954 $parts = array_merge($base_parts, $avatar_parts);
1956 // And put them together again
1957 $scheme = isset($parts['scheme']) ? $parts['scheme'] . '://' : '';
1958 $host = isset($parts['host']) ? $parts['host'] : '';
1959 $port = isset($parts['port']) ? ':' . $parts['port'] : '';
1960 $path = isset($parts['path']) ? $parts['path'] : '';
1961 $query = isset($parts['query']) ? '?' . $parts['query'] : '';
1962 $fragment = isset($parts['fragment']) ? '#' . $parts['fragment'] : '';
1964 $fixed = $scheme.$host.$port.$path.$query.$fragment;
1966 Logger::log('Base: '.$base.' - Avatar: '.$avatar.' - Fixed: '.$fixed, Logger::DATA);