3 * @file src/Network/Probe.php
5 namespace Friendica\Network;
8 * @file src/Network/Probe.php
9 * @brief Functions for probing URL
13 use Friendica\Core\System;
14 use Friendica\Core\Cache;
15 use Friendica\Core\Config;
16 use Friendica\Database\DBM;
17 use Friendica\Model\Profile;
18 use Friendica\Protocol\Email;
19 use Friendica\Protocol\Feed;
20 use Friendica\Util\Crypto;
21 use Friendica\Util\Network;
22 use Friendica\Util\XML;
28 require_once 'include/dba.php';
31 * @brief This class contain functions for probing URL
36 private static $baseurl;
39 * @brief Rearrange the array so that it always has the same order
41 * @param array $data Unordered data
43 * @return array Ordered data
45 private static function rearrangeData($data)
47 $fields = ["name", "nick", "guid", "url", "addr", "alias",
48 "photo", "community", "keywords", "location", "about",
49 "batch", "notify", "poll", "request", "confirm", "poco",
50 "priority", "network", "pubkey", "baseurl"];
53 foreach ($fields as $field) {
54 if (isset($data[$field])) {
55 $newdata[$field] = $data[$field];
57 $newdata[$field] = "";
61 // We don't use the "priority" field anymore and replace it with a dummy.
62 $newdata["priority"] = 0;
68 * @brief Check if the hostname belongs to the own server
70 * @param string $host The hostname that is to be checked
72 * @return bool Does the testes hostname belongs to the own server?
74 private static function ownHost($host)
76 $own_host = get_app()->get_hostname();
78 $parts = parse_url($host);
80 if (!isset($parts['scheme'])) {
81 $parts = parse_url('http://'.$host);
84 if (!isset($parts['host'])) {
87 return $parts['host'] == $own_host;
91 * @brief Probes for webfinger path via "host-meta"
93 * We have to check if the servers in the future still will offer this.
94 * It seems as if it was dropped from the standard.
96 * @param string $host The host part of an url
98 * @return array with template and type of the webfinger template for JSON or XML
100 private static function hostMeta($host)
102 // Reset the static variable
105 $ssl_url = "https://".$host."/.well-known/host-meta";
106 $url = "http://".$host."/.well-known/host-meta";
108 $xrd_timeout = Config::get('system', 'xrd_timeout', 20);
111 logger("Probing for ".$host, LOGGER_DEBUG);
113 $ret = Network::curl($ssl_url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
114 if ($ret['success']) {
116 $xrd = XML::parseString($xml, false);
117 $host_url = 'https://'.$host;
120 if (!is_object($xrd)) {
121 $ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
122 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
123 logger("Probing timeout for ".$url, LOGGER_DEBUG);
127 $xrd = XML::parseString($xml, false);
128 $host_url = 'http://'.$host;
130 if (!is_object($xrd)) {
131 logger("No xrd object found for ".$host, LOGGER_DEBUG);
135 $links = XML::elementToArray($xrd);
136 if (!isset($links["xrd"]["link"])) {
137 logger("No xrd data found for ".$host, LOGGER_DEBUG);
142 // The following webfinger path is defined in RFC 7033 https://tools.ietf.org/html/rfc7033
143 // Problem is that Hubzilla currently doesn't provide all data in the JSON webfinger
144 // compared to the XML webfinger. So this is commented out by now.
145 // $lrdd = array("application/jrd+json" => $host_url.'/.well-known/webfinger?resource={uri}');
147 foreach ($links["xrd"]["link"] as $value => $link) {
148 if (!empty($link["@attributes"])) {
149 $attributes = $link["@attributes"];
150 } elseif ($value == "@attributes") {
156 if (($attributes["rel"] == "lrdd") && !empty($attributes["template"])) {
157 $type = (empty($attributes["type"]) ? '' : $attributes["type"]);
159 $lrdd[$type] = $attributes["template"];
163 self::$baseurl = "http://".$host;
165 logger("Probing successful for ".$host, LOGGER_DEBUG);
171 * @brief Perform Webfinger lookup and return DFRN data
173 * Given an email style address, perform webfinger lookup and
174 * return the resulting DFRN profile URL, or if no DFRN profile URL
175 * is located, returns an OStatus subscription template (prefixed
176 * with the string 'stat:' to identify it as on OStatus template).
177 * If this isn't an email style address just return $webbie.
178 * Return an empty string if email-style addresses but webfinger fails,
179 * or if the resultant personal XRD doesn't contain a supported
180 * subscription/friend-request attribute.
182 * amended 7/9/2011 to return an hcard which could save potentially loading
183 * a lengthy content page to scrape dfrn attributes
185 * @param string $webbie Address that should be probed
186 * @param string $hcard_url Link to the hcard - is returned by reference
188 * @return string profile link
190 public static function webfingerDfrn($webbie, &$hcard_url)
194 $links = self::lrdd($webbie);
195 logger('webfingerDfrn: '.$webbie.':'.print_r($links, true), LOGGER_DATA);
197 foreach ($links as $link) {
198 if ($link['@attributes']['rel'] === NAMESPACE_DFRN) {
199 $profile_link = $link['@attributes']['href'];
201 if (($link['@attributes']['rel'] === NAMESPACE_OSTATUSSUB) && ($profile_link == "")) {
202 $profile_link = 'stat:'.$link['@attributes']['template'];
204 if ($link['@attributes']['rel'] === 'http://microformats.org/profile/hcard') {
205 $hcard_url = $link['@attributes']['href'];
209 return $profile_link;
213 * @brief Check an URI for LRDD data
215 * this is a replacement for the "lrdd" function.
216 * It isn't used in this class and has some redundancies in the code.
217 * When time comes we can check the existing calls for "lrdd" if we can rework them.
219 * @param string $uri Address that should be probed
221 * @return array uri data
223 public static function lrdd($uri)
225 $lrdd = self::hostMeta($uri);
228 if (is_bool($lrdd)) {
233 $parts = @parse_url($uri);
238 $host = $parts["host"];
239 if (!empty($parts["port"])) {
240 $host .= ':'.$parts["port"];
243 $path_parts = explode("/", trim($parts["path"], "/"));
245 $nick = array_pop($path_parts);
248 $lrdd = self::hostMeta($host);
249 $host .= "/".array_shift($path_parts);
250 } while (!$lrdd && (sizeof($path_parts) > 0));
254 logger("No lrdd data found for ".$uri, LOGGER_DEBUG);
258 foreach ($lrdd as $type => $template) {
263 $path = str_replace('{uri}', urlencode($uri), $template);
264 $webfinger = self::webfinger($path, $type);
266 if (!$webfinger && (strstr($uri, "@"))) {
267 $path = str_replace('{uri}', urlencode("acct:".$uri), $template);
268 $webfinger = self::webfinger($path, $type);
271 // Special treatment for Mastodon
272 // Problem is that Mastodon uses an URL format like http://domain.tld/@nick
273 // But the webfinger for this format fails.
274 if (!$webfinger && !empty($nick)) {
275 // Mastodon uses a "@" as prefix for usernames in their url format
276 $nick = ltrim($nick, '@');
278 $addr = $nick."@".$host;
280 $path = str_replace('{uri}', urlencode("acct:".$addr), $template);
281 $webfinger = self::webfinger($path, $type);
285 if (!is_array($webfinger["links"])) {
286 logger("No webfinger links found for ".$uri, LOGGER_DEBUG);
292 foreach ($webfinger["links"] as $link) {
293 $data[] = ["@attributes" => $link];
296 if (is_array($webfinger["aliases"])) {
297 foreach ($webfinger["aliases"] as $alias) {
298 $data[] = ["@attributes" =>
308 * @brief Fetch information (protocol endpoints and user information) about a given uri
310 * @param string $uri Address that should be probed
311 * @param string $network Test for this specific network
312 * @param integer $uid User ID for the probe (only used for mails)
313 * @param boolean $cache Use cached values?
315 * @return array uri data
317 public static function uri($uri, $network = "", $uid = -1, $cache = true)
320 $result = Cache::get("Probe::uri:".$network.":".$uri);
321 if (!is_null($result)) {
330 $data = self::detect($uri, $network, $uid);
332 if (!isset($data["url"])) {
336 if (x($data, "photo")) {
337 $data["baseurl"] = Network::getUrlMatch(normalise_link($data["baseurl"]), normalise_link($data["photo"]));
339 $data["photo"] = System::baseUrl().'/images/person-175.jpg';
342 if (empty($data["name"])) {
343 if (!empty($data["nick"])) {
344 $data["name"] = $data["nick"];
347 if (!x($data, "name")) {
348 $data["name"] = $data["url"];
352 if (empty($data["nick"])) {
353 $data["nick"] = strtolower($data["name"]);
355 if (strpos($data['nick'], ' ')) {
356 $data['nick'] = trim(substr($data['nick'], 0, strpos($data['nick'], ' ')));
360 if (self::$baseurl != "") {
361 $data["baseurl"] = self::$baseurl;
364 if (!isset($data["network"])) {
365 $data["network"] = NETWORK_PHANTOM;
368 $data = self::rearrangeData($data);
370 // Only store into the cache if the value seems to be valid
371 if (!in_array($data['network'], [NETWORK_PHANTOM, NETWORK_MAIL])) {
372 Cache::set("Probe::uri:".$network.":".$uri, $data, CACHE_DAY);
374 /// @todo temporary fix - we need a real contact update function that updates only changing fields
375 /// The biggest problem is the avatar picture that could have a reduced image size.
376 /// It should only be updated if the existing picture isn't existing anymore.
377 /// We only update the contact when it is no probing for a specific network.
378 if (($data['network'] != NETWORK_FEED)
386 $fields = ['name' => $data['name'],
387 'nick' => $data['nick'],
388 'url' => $data['url'],
389 'addr' => $data['addr'],
390 'photo' => $data['photo'],
391 'keywords' => $data['keywords'],
392 'location' => $data['location'],
393 'about' => $data['about'],
394 'notify' => $data['notify'],
395 'network' => $data['network'],
396 'server_url' => $data['baseurl']];
400 foreach ($fields as $key => $val) {
402 unset($fields[$key]);
404 $fieldnames[] = $key;
408 $fields['updated'] = DBM::date();
410 $condition = ['nurl' => normalise_link($data["url"])];
412 $old_fields = dba::selectFirst('gcontact', $fieldnames, $condition);
414 dba::update('gcontact', $fields, $condition, $old_fields);
416 $fields = ['name' => $data['name'],
417 'nick' => $data['nick'],
418 'url' => $data['url'],
419 'addr' => $data['addr'],
420 'alias' => $data['alias'],
421 'keywords' => $data['keywords'],
422 'location' => $data['location'],
423 'about' => $data['about'],
424 'batch' => $data['batch'],
425 'notify' => $data['notify'],
426 'poll' => $data['poll'],
427 'request' => $data['request'],
428 'confirm' => $data['confirm'],
429 'poco' => $data['poco'],
430 'network' => $data['network'],
431 'success_update' => DBM::date()];
435 foreach ($fields as $key => $val) {
437 unset($fields[$key]);
439 $fieldnames[] = $key;
443 $condition = ['nurl' => normalise_link($data["url"]), 'self' => false, 'uid' => 0];
445 $old_fields = dba::selectFirst('contact', $fieldnames, $condition);
447 dba::update('contact', $fields, $condition, $old_fields);
455 * @brief Switch the scheme of an url between http and https
457 * @param string $url URL
459 * @return string switched URL
461 private static function switchScheme($url)
463 $parts = parse_url($url);
465 if (!isset($parts['scheme'])) {
469 if ($parts['scheme'] == 'http') {
470 $url = str_replace('http://', 'https://', $url);
471 } elseif ($parts['scheme'] == 'https') {
472 $url = str_replace('https://', 'http://', $url);
479 * @brief Checks if a profile url should be OStatus but only provides partial information
481 * @param array $webfinger Webfinger data
482 * @param string $lrdd Path template for webfinger request
483 * @param string $type type
485 * @return array fixed webfinger data
487 private static function fixOStatus($webfinger, $lrdd, $type)
489 if (empty($webfinger['links']) || empty($webfinger['subject'])) {
496 foreach ($webfinger['links'] as $link) {
497 if ($link['rel'] == NAMESPACE_OSTATUSSUB) {
500 if ($link['rel'] == 'magic-public-key') {
505 if (!$is_ostatus || $has_key) {
509 $url = self::switchScheme($webfinger['subject']);
510 $path = str_replace('{uri}', urlencode($url), $lrdd);
511 $webfinger2 = self::webfinger($path, $type);
513 // Is the new webfinger detectable as OStatus?
514 if (self::ostatus($webfinger2, true)) {
515 $webfinger = $webfinger2;
522 * @brief Fetch information (protocol endpoints and user information) about a given uri
524 * This function is only called by the "uri" function that adds caching and rearranging of data.
526 * @param string $uri Address that should be probed
527 * @param string $network Test for this specific network
528 * @param integer $uid User ID for the probe (only used for mails)
530 * @return array uri data
532 private static function detect($uri, $network, $uid)
534 $parts = parse_url($uri);
536 if (!empty($parts["scheme"]) && !empty($parts["host"]) && !empty($parts["path"])) {
537 $host = $parts["host"];
538 if (!empty($parts["port"])) {
539 $host .= ':'.$parts["port"];
542 if ($host == 'twitter.com') {
543 return ["network" => NETWORK_TWITTER];
545 $lrdd = self::hostMeta($host);
547 if (is_bool($lrdd)) {
551 $path_parts = explode("/", trim($parts["path"], "/"));
553 while (!$lrdd && (sizeof($path_parts) > 1)) {
554 $host .= "/".array_shift($path_parts);
555 $lrdd = self::hostMeta($host);
558 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
559 return self::feed($uri);
561 $nick = array_pop($path_parts);
563 // Mastodon uses a "@" as prefix for usernames in their url format
564 $nick = ltrim($nick, '@');
566 $addr = $nick."@".$host;
567 } elseif (strstr($uri, '@')) {
568 // If the URI starts with "mailto:" then jump directly to the mail detection
569 if (strpos($uri, 'mailto:') !== false) {
570 $uri = str_replace('mailto:', '', $uri);
571 return self::mail($uri, $uid);
574 if ($network == NETWORK_MAIL) {
575 return self::mail($uri, $uid);
577 // Remove "acct:" from the URI
578 $uri = str_replace('acct:', '', $uri);
580 $host = substr($uri, strpos($uri, '@') + 1);
581 $nick = substr($uri, 0, strpos($uri, '@'));
583 if (strpos($uri, '@twitter.com')) {
584 return ["network" => NETWORK_TWITTER];
586 $lrdd = self::hostMeta($host);
588 if (is_bool($lrdd)) {
593 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
594 return self::mail($uri, $uid);
598 logger("Uri ".$uri." was not detectable", LOGGER_DEBUG);
604 /// @todo Do we need the prefix "acct:" or "acct://"?
606 foreach ($lrdd as $type => $template) {
611 // At first try it with the given uri
612 $path = str_replace('{uri}', urlencode($uri), $template);
613 $webfinger = self::webfinger($path, $type);
615 // Fix possible problems with GNU Social probing to wrong scheme
616 $webfinger = self::fixOStatus($webfinger, $template, $type);
618 // We cannot be sure that the detected address was correct, so we don't use the values
619 if ($webfinger && ($uri != $addr)) {
624 // Try webfinger with the address (user@domain.tld)
626 $path = str_replace('{uri}', urlencode($addr), $template);
627 $webfinger = self::webfinger($path, $type);
630 // Mastodon needs to have it with "acct:"
632 $path = str_replace('{uri}', urlencode("acct:".$addr), $template);
633 $webfinger = self::webfinger($path, $type);
638 return self::feed($uri);
643 logger("Probing ".$uri, LOGGER_DEBUG);
645 if (in_array($network, ["", NETWORK_DFRN])) {
646 $result = self::dfrn($webfinger);
648 if ((!$result && ($network == "")) || ($network == NETWORK_DIASPORA)) {
649 $result = self::diaspora($webfinger);
651 if ((!$result && ($network == "")) || ($network == NETWORK_OSTATUS)) {
652 $result = self::ostatus($webfinger);
654 if ((!$result && ($network == "")) || ($network == NETWORK_PUMPIO)) {
655 $result = self::pumpio($webfinger);
657 if ((!$result && ($network == "")) || ($network == NETWORK_FEED)) {
658 $result = self::feed($uri);
660 // We overwrite the detected nick with our try if the previois routines hadn't detected it.
661 // Additionally it is overwritten when the nickname doesn't make sense (contains spaces).
662 if ((empty($result["nick"]) || (strstr($result["nick"], " "))) && ($nick != "")) {
663 $result["nick"] = $nick;
666 if (empty($result["addr"]) && ($addr != "")) {
667 $result["addr"] = $addr;
671 logger($uri." is ".$result["network"], LOGGER_DEBUG);
673 if (empty($result["baseurl"])) {
674 $pos = strpos($result["url"], $host);
676 $result["baseurl"] = substr($result["url"], 0, $pos).$host;
684 * @brief Perform a webfinger request.
686 * For details see RFC 7033: <https://tools.ietf.org/html/rfc7033>
688 * @param string $url Address that should be probed
689 * @param string $type type
691 * @return array webfinger data
693 private static function webfinger($url, $type)
695 $xrd_timeout = Config::get('system', 'xrd_timeout', 20);
698 $ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => $type]);
699 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
702 $data = $ret['body'];
704 $webfinger = json_decode($data, true);
705 if (is_array($webfinger)) {
706 if (!isset($webfinger["links"])) {
707 logger("No json webfinger links for ".$url, LOGGER_DEBUG);
713 // If it is not JSON, maybe it is XML
714 $xrd = XML::parseString($data, false);
715 if (!is_object($xrd)) {
716 logger("No webfinger data retrievable for ".$url, LOGGER_DEBUG);
720 $xrd_arr = XML::elementToArray($xrd);
721 if (!isset($xrd_arr["xrd"]["link"])) {
722 logger("No XML webfinger links for ".$url, LOGGER_DEBUG);
728 if (!empty($xrd_arr["xrd"]["subject"])) {
729 $webfinger["subject"] = $xrd_arr["xrd"]["subject"];
732 if (!empty($xrd_arr["xrd"]["alias"])) {
733 $webfinger["aliases"] = $xrd_arr["xrd"]["alias"];
736 $webfinger["links"] = [];
738 foreach ($xrd_arr["xrd"]["link"] as $value => $data) {
739 if (!empty($data["@attributes"])) {
740 $attributes = $data["@attributes"];
741 } elseif ($value == "@attributes") {
747 $webfinger["links"][] = $attributes;
753 * @brief Poll the Friendica specific noscrape page.
755 * "noscrape" is a faster alternative to fetch the data from the hcard.
756 * This functionality was originally created for the directory.
758 * @param string $noscrape_url Link to the noscrape page
759 * @param array $data The already fetched data
761 * @return array noscrape data
763 private static function pollNoscrape($noscrape_url, $data)
765 $ret = Network::curl($noscrape_url);
766 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
769 $content = $ret['body'];
771 logger("Empty body for ".$noscrape_url, LOGGER_DEBUG);
775 $json = json_decode($content, true);
776 if (!is_array($json)) {
777 logger("No json data for ".$noscrape_url, LOGGER_DEBUG);
781 if (!empty($json["fn"])) {
782 $data["name"] = $json["fn"];
785 if (!empty($json["addr"])) {
786 $data["addr"] = $json["addr"];
789 if (!empty($json["nick"])) {
790 $data["nick"] = $json["nick"];
793 if (!empty($json["guid"])) {
794 $data["guid"] = $json["guid"];
797 if (!empty($json["comm"])) {
798 $data["community"] = $json["comm"];
801 if (!empty($json["tags"])) {
802 $keywords = implode(" ", $json["tags"]);
803 if ($keywords != "") {
804 $data["keywords"] = $keywords;
808 $location = Profile::formatLocation($json);
810 $data["location"] = $location;
813 if (!empty($json["about"])) {
814 $data["about"] = $json["about"];
817 if (!empty($json["key"])) {
818 $data["pubkey"] = $json["key"];
821 if (!empty($json["photo"])) {
822 $data["photo"] = $json["photo"];
825 if (!empty($json["dfrn-request"])) {
826 $data["request"] = $json["dfrn-request"];
829 if (!empty($json["dfrn-confirm"])) {
830 $data["confirm"] = $json["dfrn-confirm"];
833 if (!empty($json["dfrn-notify"])) {
834 $data["notify"] = $json["dfrn-notify"];
837 if (!empty($json["dfrn-poll"])) {
838 $data["poll"] = $json["dfrn-poll"];
845 * @brief Check for valid DFRN data
847 * @param array $data DFRN data
849 * @return int Number of errors
851 public static function validDfrn($data)
854 if (!isset($data['key'])) {
857 if (!isset($data['dfrn-request'])) {
860 if (!isset($data['dfrn-confirm'])) {
863 if (!isset($data['dfrn-notify'])) {
866 if (!isset($data['dfrn-poll'])) {
873 * @brief Fetch data from a DFRN profile page and via "noscrape"
875 * @param string $profile_link Link to the profile page
877 * @return array profile data
879 public static function profile($profile_link)
883 logger("Check profile ".$profile_link, LOGGER_DEBUG);
885 // Fetch data via noscrape - this is faster
886 $noscrape_url = str_replace(["/hcard/", "/profile/"], "/noscrape/", $profile_link);
887 $data = self::pollNoscrape($noscrape_url, $data);
889 if (!isset($data["notify"])
890 || !isset($data["confirm"])
891 || !isset($data["request"])
892 || !isset($data["poll"])
893 || !isset($data["poco"])
894 || !isset($data["name"])
895 || !isset($data["photo"])
897 $data = self::pollHcard($profile_link, $data, true);
901 $prof_data["addr"] = $data["addr"];
902 $prof_data["nick"] = $data["nick"];
903 $prof_data["dfrn-request"] = $data["request"];
904 $prof_data["dfrn-confirm"] = $data["confirm"];
905 $prof_data["dfrn-notify"] = $data["notify"];
906 $prof_data["dfrn-poll"] = $data["poll"];
907 $prof_data["dfrn-poco"] = $data["poco"];
908 $prof_data["photo"] = $data["photo"];
909 $prof_data["fn"] = $data["name"];
910 $prof_data["key"] = $data["pubkey"];
912 logger("Result for profile ".$profile_link.": ".print_r($prof_data, true), LOGGER_DEBUG);
918 * @brief Check for DFRN contact
920 * @param array $webfinger Webfinger data
922 * @return array DFRN data
924 private static function dfrn($webfinger)
928 foreach ($webfinger["links"] as $link) {
929 if (($link["rel"] == NAMESPACE_DFRN) && ($link["href"] != "")) {
930 $data["network"] = NETWORK_DFRN;
931 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
932 $data["poll"] = $link["href"];
933 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
934 $data["url"] = $link["href"];
935 } elseif (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
936 $hcard_url = $link["href"];
937 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
938 $data["poco"] = $link["href"];
939 } elseif (($link["rel"] == "http://webfinger.net/rel/avatar") && ($link["href"] != "")) {
940 $data["photo"] = $link["href"];
941 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
942 $data["baseurl"] = trim($link["href"], '/');
943 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
944 $data["guid"] = $link["href"];
945 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
946 $data["pubkey"] = base64_decode($link["href"]);
948 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
949 if (strstr($data["pubkey"], 'RSA ')) {
950 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
955 if (is_array($webfinger["aliases"])) {
956 foreach ($webfinger["aliases"] as $alias) {
957 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
958 $data["alias"] = $alias;
959 } elseif (substr($alias, 0, 5) == 'acct:') {
960 $data["addr"] = substr($alias, 5);
965 if (substr($webfinger["subject"], 0, 5) == "acct:") {
966 $data["addr"] = substr($webfinger["subject"], 5);
969 if (!isset($data["network"]) || ($hcard_url == "")) {
973 // Fetch data via noscrape - this is faster
974 $noscrape_url = str_replace("/hcard/", "/noscrape/", $hcard_url);
975 $data = self::pollNoscrape($noscrape_url, $data);
977 if (isset($data["notify"])
978 && isset($data["confirm"])
979 && isset($data["request"])
980 && isset($data["poll"])
981 && isset($data["name"])
982 && isset($data["photo"])
987 $data = self::pollHcard($hcard_url, $data, true);
993 * @brief Poll the hcard page (Diaspora and Friendica specific)
995 * @param string $hcard_url Link to the hcard page
996 * @param array $data The already fetched data
997 * @param boolean $dfrn Poll DFRN specific data
999 * @return array hcard data
1001 private static function pollHcard($hcard_url, $data, $dfrn = false)
1003 $ret = Network::curl($hcard_url);
1004 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1007 $content = $ret['body'];
1012 $doc = new DOMDocument();
1013 if (!@$doc->loadHTML($content)) {
1017 $xpath = new DomXPath($doc);
1019 $vcards = $xpath->query("//div[contains(concat(' ', @class, ' '), ' vcard ')]");
1020 if (!is_object($vcards)) {
1024 if ($vcards->length > 0) {
1025 $vcard = $vcards->item(0);
1027 // We have to discard the guid from the hcard in favour of the guid from lrdd
1028 // Reason: Hubzilla doesn't use the value "uid" in the hcard like Diaspora does.
1029 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' uid ')]", $vcard); // */
1030 if (($search->length > 0) && ($data["guid"] == "")) {
1031 $data["guid"] = $search->item(0)->nodeValue;
1034 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' nickname ')]", $vcard); // */
1035 if ($search->length > 0) {
1036 $data["nick"] = $search->item(0)->nodeValue;
1039 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' fn ')]", $vcard); // */
1040 if ($search->length > 0) {
1041 $data["name"] = $search->item(0)->nodeValue;
1044 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' searchable ')]", $vcard); // */
1045 if ($search->length > 0) {
1046 $data["searchable"] = $search->item(0)->nodeValue;
1049 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' key ')]", $vcard); // */
1050 if ($search->length > 0) {
1051 $data["pubkey"] = $search->item(0)->nodeValue;
1052 if (strstr($data["pubkey"], 'RSA ')) {
1053 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1057 $search = $xpath->query("//*[@id='pod_location']", $vcard); // */
1058 if ($search->length > 0) {
1059 $data["baseurl"] = trim($search->item(0)->nodeValue, "/");
1064 $photos = $xpath->query("//*[contains(concat(' ', @class, ' '), ' photo ') or contains(concat(' ', @class, ' '), ' avatar ')]", $vcard); // */
1065 foreach ($photos as $photo) {
1067 foreach ($photo->attributes as $attribute) {
1068 $attr[$attribute->name] = trim($attribute->value);
1071 if (isset($attr["src"]) && isset($attr["width"])) {
1072 $avatar[$attr["width"]] = $attr["src"];
1075 // We don't have a width. So we just take everything that we got.
1076 // This is a Hubzilla workaround which doesn't send a width.
1077 if ((sizeof($avatar) == 0) && !empty($attr["src"])) {
1078 $avatar[] = $attr["src"];
1082 if (sizeof($avatar)) {
1084 $data["photo"] = self::fixAvatar(array_pop($avatar), $data["baseurl"]);
1088 // Poll DFRN specific data
1089 $search = $xpath->query("//link[contains(concat(' ', @rel), ' dfrn-')]");
1090 if ($search->length > 0) {
1091 foreach ($search as $link) {
1092 //$data["request"] = $search->item(0)->nodeValue;
1094 foreach ($link->attributes as $attribute) {
1095 $attr[$attribute->name] = trim($attribute->value);
1098 $data[substr($attr["rel"], 5)] = $attr["href"];
1102 // Older Friendica versions had used the "uid" field differently than newer versions
1103 if ($data["nick"] == $data["guid"]) {
1104 unset($data["guid"]);
1113 * @brief Check for Diaspora contact
1115 * @param array $webfinger Webfinger data
1117 * @return array Diaspora data
1119 private static function diaspora($webfinger)
1123 foreach ($webfinger["links"] as $link) {
1124 if (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
1125 $hcard_url = $link["href"];
1126 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
1127 $data["baseurl"] = trim($link["href"], '/');
1128 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
1129 $data["guid"] = $link["href"];
1130 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
1131 $data["url"] = $link["href"];
1132 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1133 $data["poll"] = $link["href"];
1134 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
1135 $data["poco"] = $link["href"];
1136 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1137 $data["notify"] = $link["href"];
1138 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
1139 $data["pubkey"] = base64_decode($link["href"]);
1141 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
1142 if (strstr($data["pubkey"], 'RSA ')) {
1143 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1148 if (!isset($data["url"]) || ($hcard_url == "")) {
1152 if (is_array($webfinger["aliases"])) {
1153 foreach ($webfinger["aliases"] as $alias) {
1154 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
1155 $data["alias"] = $alias;
1156 } elseif (substr($alias, 0, 5) == 'acct:') {
1157 $data["addr"] = substr($alias, 5);
1162 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == 'acct:')) {
1163 $data["addr"] = substr($webfinger["subject"], 5);
1166 // Fetch further information from the hcard
1167 $data = self::pollHcard($hcard_url, $data);
1173 if (isset($data["url"])
1174 && isset($data["guid"])
1175 && isset($data["baseurl"])
1176 && isset($data["pubkey"])
1177 && ($hcard_url != "")
1179 $data["network"] = NETWORK_DIASPORA;
1181 // The Diaspora handle must always be lowercase
1182 $data["addr"] = strtolower($data["addr"]);
1184 // We have to overwrite the detected value for "notify" since Hubzilla doesn't send it
1185 $data["notify"] = $data["baseurl"] . "/receive/users/" . $data["guid"];
1186 $data["batch"] = $data["baseurl"] . "/receive/public";
1195 * @brief Check for OStatus contact
1197 * @param array $webfinger Webfinger data
1198 * @param bool $short Short detection mode
1200 * @return array|bool OStatus data or "false" on error or "true" on short mode
1202 private static function ostatus($webfinger, $short = false)
1206 if (is_array($webfinger["aliases"])) {
1207 foreach ($webfinger["aliases"] as $alias) {
1208 if (strstr($alias, "@") && !strstr(normalise_link($alias), "http://")) {
1209 $data["addr"] = str_replace('acct:', '', $alias);
1214 if (is_string($webfinger["subject"]) && strstr($webfinger["subject"], "@")
1215 && !strstr(normalise_link($webfinger["subject"]), "http://")
1217 $data["addr"] = str_replace('acct:', '', $webfinger["subject"]);
1221 if (is_array($webfinger["links"])) {
1222 foreach ($webfinger["links"] as $link) {
1223 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1224 && ($link["type"] == "text/html")
1225 && ($link["href"] != "")
1227 $data["url"] = $link["href"];
1228 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1229 $data["notify"] = $link["href"];
1230 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1231 $data["poll"] = $link["href"];
1232 } elseif (($link["rel"] == "magic-public-key") && ($link["href"] != "")) {
1233 $pubkey = $link["href"];
1235 if (substr($pubkey, 0, 5) === 'data:') {
1236 if (strstr($pubkey, ',')) {
1237 $pubkey = substr($pubkey, strpos($pubkey, ',') + 1);
1239 $pubkey = substr($pubkey, 5);
1241 } elseif (normalise_link($pubkey) == 'http://') {
1242 $ret = Network::curl($pubkey);
1243 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1246 $pubkey = $ret['body'];
1249 $key = explode(".", $pubkey);
1251 if (sizeof($key) >= 3) {
1252 $m = base64url_decode($key[1]);
1253 $e = base64url_decode($key[2]);
1254 $data["pubkey"] = Crypto::meToPem($m, $e);
1260 if (isset($data["notify"]) && isset($data["pubkey"])
1261 && isset($data["poll"])
1262 && isset($data["url"])
1264 $data["network"] = NETWORK_OSTATUS;
1273 // Fetch all additional data from the feed
1274 $ret = Network::curl($data["poll"]);
1275 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1278 $feed = $ret['body'];
1279 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1284 if ($feed_data["header"]["author-name"] != "") {
1285 $data["name"] = $feed_data["header"]["author-name"];
1287 if ($feed_data["header"]["author-nick"] != "") {
1288 $data["nick"] = $feed_data["header"]["author-nick"];
1290 if ($feed_data["header"]["author-avatar"] != "") {
1291 $data["photo"] = self::fixAvatar($feed_data["header"]["author-avatar"], $data["url"]);
1293 if ($feed_data["header"]["author-id"] != "") {
1294 $data["alias"] = $feed_data["header"]["author-id"];
1296 if ($feed_data["header"]["author-location"] != "") {
1297 $data["location"] = $feed_data["header"]["author-location"];
1299 if ($feed_data["header"]["author-about"] != "") {
1300 $data["about"] = $feed_data["header"]["author-about"];
1302 // OStatus has serious issues when the the url doesn't fit (ssl vs. non ssl)
1303 // So we take the value that we just fetched, although the other one worked as well
1304 if ($feed_data["header"]["author-link"] != "") {
1305 $data["url"] = $feed_data["header"]["author-link"];
1308 if (($data['poll'] == $data['url']) && ($data["alias"] != '')) {
1309 $data['url'] = $data["alias"];
1310 $data["alias"] = '';
1313 /// @todo Fetch location and "about" from the feed as well
1318 * @brief Fetch data from a pump.io profile page
1320 * @param string $profile_link Link to the profile page
1322 * @return array profile data
1324 private static function pumpioProfileData($profile_link)
1326 $doc = new DOMDocument();
1327 if (!@$doc->loadHTMLFile($profile_link)) {
1331 $xpath = new DomXPath($doc);
1335 // This is ugly - but pump.io doesn't seem to know a better way for it
1336 $data["name"] = trim($xpath->query("//h1[@class='media-header']")->item(0)->nodeValue);
1337 $pos = strpos($data["name"], chr(10));
1339 $data["name"] = trim(substr($data["name"], 0, $pos));
1342 $avatar = $xpath->query("//img[@class='img-rounded media-object']")->item(0);
1344 foreach ($avatar->attributes as $attribute) {
1345 if ($attribute->name == "src") {
1346 $data["photo"] = trim($attribute->value);
1351 $data["location"] = $xpath->query("//p[@class='location']")->item(0)->nodeValue;
1352 $data["about"] = $xpath->query("//p[@class='summary']")->item(0)->nodeValue;
1358 * @brief Check for pump.io contact
1360 * @param array $webfinger Webfinger data
1362 * @return array pump.io data
1364 private static function pumpio($webfinger)
1367 foreach ($webfinger["links"] as $link) {
1368 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1369 && ($link["type"] == "text/html")
1370 && ($link["href"] != "")
1372 $data["url"] = $link["href"];
1373 } elseif (($link["rel"] == "activity-inbox") && ($link["href"] != "")) {
1374 $data["notify"] = $link["href"];
1375 } elseif (($link["rel"] == "activity-outbox") && ($link["href"] != "")) {
1376 $data["poll"] = $link["href"];
1377 } elseif (($link["rel"] == "dialback") && ($link["href"] != "")) {
1378 $data["dialback"] = $link["href"];
1381 if (isset($data["poll"]) && isset($data["notify"])
1382 && isset($data["dialback"])
1383 && isset($data["url"])
1385 // by now we use these fields only for the network type detection
1386 // So we unset all data that isn't used at the moment
1387 unset($data["dialback"]);
1389 $data["network"] = NETWORK_PUMPIO;
1394 $profile_data = self::pumpioProfileData($data["url"]);
1396 if (!$profile_data) {
1400 $data = array_merge($data, $profile_data);
1406 * @brief Check page for feed link
1408 * @param string $url Page link
1410 * @return string feed link
1412 private static function getFeedLink($url)
1414 $doc = new DOMDocument();
1416 if (!@$doc->loadHTMLFile($url)) {
1420 $xpath = new DomXPath($doc);
1422 //$feeds = $xpath->query("/html/head/link[@type='application/rss+xml']");
1423 $feeds = $xpath->query("/html/head/link[@type='application/rss+xml' and @rel='alternate']");
1424 if (!is_object($feeds)) {
1428 if ($feeds->length == 0) {
1434 foreach ($feeds as $feed) {
1436 foreach ($feed->attributes as $attribute) {
1437 $attr[$attribute->name] = trim($attribute->value);
1440 if ($feed_url == "") {
1441 $feed_url = $attr["href"];
1449 * @brief Check for feed contact
1451 * @param string $url Profile link
1452 * @param boolean $probe Do a probe if the page contains a feed link
1454 * @return array feed data
1456 private static function feed($url, $probe = true)
1458 $ret = Network::curl($url);
1459 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1462 $feed = $ret['body'];
1463 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1470 $feed_url = self::getFeedLink($url);
1476 return self::feed($feed_url, false);
1479 if ($feed_data["header"]["author-name"] != "") {
1480 $data["name"] = $feed_data["header"]["author-name"];
1483 if ($feed_data["header"]["author-nick"] != "") {
1484 $data["nick"] = $feed_data["header"]["author-nick"];
1487 if ($feed_data["header"]["author-avatar"] != "") {
1488 $data["photo"] = $feed_data["header"]["author-avatar"];
1491 if ($feed_data["header"]["author-id"] != "") {
1492 $data["alias"] = $feed_data["header"]["author-id"];
1495 $data["url"] = $url;
1496 $data["poll"] = $url;
1498 if ($feed_data["header"]["author-link"] != "") {
1499 $data["baseurl"] = $feed_data["header"]["author-link"];
1501 $data["baseurl"] = $data["url"];
1504 $data["network"] = NETWORK_FEED;
1510 * @brief Check for mail contact
1512 * @param string $uri Profile link
1513 * @param integer $uid User ID
1515 * @return array mail data
1517 private static function mail($uri, $uid)
1519 if (!Network::isEmailDomainValid($uri)) {
1524 $x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1", intval($uid));
1526 $r = q("SELECT * FROM `mailacct` WHERE `uid` = %d AND `server` != '' LIMIT 1", intval($uid));
1528 if (DBM::is_result($x) && DBM::is_result($r)) {
1529 $mailbox = Email::constructMailboxName($r[0]);
1531 openssl_private_decrypt(hex2bin($r[0]['pass']), $password, $x[0]['prvkey']);
1532 $mbox = Email::connect($mailbox, $r[0]['user'], $password);
1538 $msgs = Email::poll($mbox, $uri);
1539 logger('searching '.$uri.', '.count($msgs).' messages found.', LOGGER_DEBUG);
1541 if (!count($msgs)) {
1546 $phost = substr($uri, strpos($uri, '@') + 1);
1549 $data["addr"] = $uri;
1550 $data["network"] = NETWORK_MAIL;
1551 $data["name"] = substr($uri, 0, strpos($uri, '@'));
1552 $data["nick"] = $data["name"];
1553 $data["photo"] = Network::lookupAvatarByEmail($uri);
1554 $data["url"] = 'mailto:'.$uri;
1555 $data["notify"] = 'smtp '.random_string();
1556 $data["poll"] = 'email '.random_string();
1558 $x = Email::messageMeta($mbox, $msgs[0]);
1559 if (stristr($x[0]->from, $uri)) {
1560 $adr = imap_rfc822_parse_adrlist($x[0]->from, '');
1561 } elseif (stristr($x[0]->to, $uri)) {
1562 $adr = imap_rfc822_parse_adrlist($x[0]->to, '');
1565 foreach ($adr as $feadr) {
1566 if ((strcasecmp($feadr->mailbox, $data["name"]) == 0)
1567 &&(strcasecmp($feadr->host, $phost) == 0)
1568 && (strlen($feadr->personal))
1570 $personal = imap_mime_header_decode($feadr->personal);
1572 foreach ($personal as $perspart) {
1573 if ($perspart->charset != "default") {
1574 $data["name"] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
1576 $data["name"] .= $perspart->text;
1580 $data["name"] = notags($data["name"]);
1584 if (!empty($mbox)) {
1592 * @brief Mix two paths together to possibly fix missing parts
1594 * @param string $avatar Path to the avatar
1595 * @param string $base Another path that is hopefully complete
1597 * @return string fixed avatar path
1599 public static function fixAvatar($avatar, $base)
1601 $base_parts = parse_url($base);
1603 // Remove all parts that could create a problem
1604 unset($base_parts['path']);
1605 unset($base_parts['query']);
1606 unset($base_parts['fragment']);
1608 $avatar_parts = parse_url($avatar);
1611 $parts = array_merge($base_parts, $avatar_parts);
1613 // And put them together again
1614 $scheme = isset($parts['scheme']) ? $parts['scheme'] . '://' : '';
1615 $host = isset($parts['host']) ? $parts['host'] : '';
1616 $port = isset($parts['port']) ? ':' . $parts['port'] : '';
1617 $path = isset($parts['path']) ? $parts['path'] : '';
1618 $query = isset($parts['query']) ? '?' . $parts['query'] : '';
1619 $fragment = isset($parts['fragment']) ? '#' . $parts['fragment'] : '';
1621 $fixed = $scheme.$host.$port.$path.$query.$fragment;
1623 logger('Base: '.$base.' - Avatar: '.$avatar.' - Fixed: '.$fixed, LOGGER_DATA);