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 // When the contact doesn't exist, the value "true" will trigger an insert
450 $fields['blocked'] = false;
451 $fields['pending'] = false;
454 dba::update('contact', $fields, $condition, $old_fields);
462 * @brief Switch the scheme of an url between http and https
464 * @param string $url URL
466 * @return string switched URL
468 private static function switchScheme($url)
470 $parts = parse_url($url);
472 if (!isset($parts['scheme'])) {
476 if ($parts['scheme'] == 'http') {
477 $url = str_replace('http://', 'https://', $url);
478 } elseif ($parts['scheme'] == 'https') {
479 $url = str_replace('https://', 'http://', $url);
486 * @brief Checks if a profile url should be OStatus but only provides partial information
488 * @param array $webfinger Webfinger data
489 * @param string $lrdd Path template for webfinger request
490 * @param string $type type
492 * @return array fixed webfinger data
494 private static function fixOStatus($webfinger, $lrdd, $type)
496 if (empty($webfinger['links']) || empty($webfinger['subject'])) {
503 foreach ($webfinger['links'] as $link) {
504 if ($link['rel'] == NAMESPACE_OSTATUSSUB) {
507 if ($link['rel'] == 'magic-public-key') {
512 if (!$is_ostatus || $has_key) {
516 $url = self::switchScheme($webfinger['subject']);
517 $path = str_replace('{uri}', urlencode($url), $lrdd);
518 $webfinger2 = self::webfinger($path, $type);
520 // Is the new webfinger detectable as OStatus?
521 if (self::ostatus($webfinger2, true)) {
522 $webfinger = $webfinger2;
529 * @brief Fetch information (protocol endpoints and user information) about a given uri
531 * This function is only called by the "uri" function that adds caching and rearranging of data.
533 * @param string $uri Address that should be probed
534 * @param string $network Test for this specific network
535 * @param integer $uid User ID for the probe (only used for mails)
537 * @return array uri data
539 private static function detect($uri, $network, $uid)
541 $parts = parse_url($uri);
543 if (!empty($parts["scheme"]) && !empty($parts["host"]) && !empty($parts["path"])) {
544 $host = $parts["host"];
545 if (!empty($parts["port"])) {
546 $host .= ':'.$parts["port"];
549 if ($host == 'twitter.com') {
550 return ["network" => NETWORK_TWITTER];
552 $lrdd = self::hostMeta($host);
554 if (is_bool($lrdd)) {
558 $path_parts = explode("/", trim($parts["path"], "/"));
560 while (!$lrdd && (sizeof($path_parts) > 1)) {
561 $host .= "/".array_shift($path_parts);
562 $lrdd = self::hostMeta($host);
565 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
566 return self::feed($uri);
568 $nick = array_pop($path_parts);
570 // Mastodon uses a "@" as prefix for usernames in their url format
571 $nick = ltrim($nick, '@');
573 $addr = $nick."@".$host;
574 } elseif (strstr($uri, '@')) {
575 // If the URI starts with "mailto:" then jump directly to the mail detection
576 if (strpos($uri, 'mailto:') !== false) {
577 $uri = str_replace('mailto:', '', $uri);
578 return self::mail($uri, $uid);
581 if ($network == NETWORK_MAIL) {
582 return self::mail($uri, $uid);
584 // Remove "acct:" from the URI
585 $uri = str_replace('acct:', '', $uri);
587 $host = substr($uri, strpos($uri, '@') + 1);
588 $nick = substr($uri, 0, strpos($uri, '@'));
590 if (strpos($uri, '@twitter.com')) {
591 return ["network" => NETWORK_TWITTER];
593 $lrdd = self::hostMeta($host);
595 if (is_bool($lrdd)) {
600 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
601 return self::mail($uri, $uid);
605 logger("Uri ".$uri." was not detectable", LOGGER_DEBUG);
611 /// @todo Do we need the prefix "acct:" or "acct://"?
613 foreach ($lrdd as $type => $template) {
618 // At first try it with the given uri
619 $path = str_replace('{uri}', urlencode($uri), $template);
620 $webfinger = self::webfinger($path, $type);
622 // Fix possible problems with GNU Social probing to wrong scheme
623 $webfinger = self::fixOStatus($webfinger, $template, $type);
625 // We cannot be sure that the detected address was correct, so we don't use the values
626 if ($webfinger && ($uri != $addr)) {
631 // Try webfinger with the address (user@domain.tld)
633 $path = str_replace('{uri}', urlencode($addr), $template);
634 $webfinger = self::webfinger($path, $type);
637 // Mastodon needs to have it with "acct:"
639 $path = str_replace('{uri}', urlencode("acct:".$addr), $template);
640 $webfinger = self::webfinger($path, $type);
645 return self::feed($uri);
650 logger("Probing ".$uri, LOGGER_DEBUG);
652 if (in_array($network, ["", NETWORK_DFRN])) {
653 $result = self::dfrn($webfinger);
655 if ((!$result && ($network == "")) || ($network == NETWORK_DIASPORA)) {
656 $result = self::diaspora($webfinger);
658 if ((!$result && ($network == "")) || ($network == NETWORK_OSTATUS)) {
659 $result = self::ostatus($webfinger);
661 if ((!$result && ($network == "")) || ($network == NETWORK_PUMPIO)) {
662 $result = self::pumpio($webfinger, $addr);
664 if ((!$result && ($network == "")) || ($network == NETWORK_FEED)) {
665 $result = self::feed($uri);
667 // We overwrite the detected nick with our try if the previois routines hadn't detected it.
668 // Additionally it is overwritten when the nickname doesn't make sense (contains spaces).
669 if ((empty($result["nick"]) || (strstr($result["nick"], " "))) && ($nick != "")) {
670 $result["nick"] = $nick;
673 if (empty($result["addr"]) && ($addr != "")) {
674 $result["addr"] = $addr;
678 logger($uri." is ".$result["network"], LOGGER_DEBUG);
680 if (empty($result["baseurl"])) {
681 $pos = strpos($result["url"], $host);
683 $result["baseurl"] = substr($result["url"], 0, $pos).$host;
690 * @brief Perform a webfinger request.
692 * For details see RFC 7033: <https://tools.ietf.org/html/rfc7033>
694 * @param string $url Address that should be probed
695 * @param string $type type
697 * @return array webfinger data
699 private static function webfinger($url, $type)
701 $xrd_timeout = Config::get('system', 'xrd_timeout', 20);
704 $ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => $type]);
705 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
708 $data = $ret['body'];
710 $webfinger = json_decode($data, true);
711 if (is_array($webfinger)) {
712 if (!isset($webfinger["links"])) {
713 logger("No json webfinger links for ".$url, LOGGER_DEBUG);
719 // If it is not JSON, maybe it is XML
720 $xrd = XML::parseString($data, false);
721 if (!is_object($xrd)) {
722 logger("No webfinger data retrievable for ".$url, LOGGER_DEBUG);
726 $xrd_arr = XML::elementToArray($xrd);
727 if (!isset($xrd_arr["xrd"]["link"])) {
728 logger("No XML webfinger links for ".$url, LOGGER_DEBUG);
734 if (!empty($xrd_arr["xrd"]["subject"])) {
735 $webfinger["subject"] = $xrd_arr["xrd"]["subject"];
738 if (!empty($xrd_arr["xrd"]["alias"])) {
739 $webfinger["aliases"] = $xrd_arr["xrd"]["alias"];
742 $webfinger["links"] = [];
744 foreach ($xrd_arr["xrd"]["link"] as $value => $data) {
745 if (!empty($data["@attributes"])) {
746 $attributes = $data["@attributes"];
747 } elseif ($value == "@attributes") {
753 $webfinger["links"][] = $attributes;
759 * @brief Poll the Friendica specific noscrape page.
761 * "noscrape" is a faster alternative to fetch the data from the hcard.
762 * This functionality was originally created for the directory.
764 * @param string $noscrape_url Link to the noscrape page
765 * @param array $data The already fetched data
767 * @return array noscrape data
769 private static function pollNoscrape($noscrape_url, $data)
771 $ret = Network::curl($noscrape_url);
772 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
775 $content = $ret['body'];
777 logger("Empty body for ".$noscrape_url, LOGGER_DEBUG);
781 $json = json_decode($content, true);
782 if (!is_array($json)) {
783 logger("No json data for ".$noscrape_url, LOGGER_DEBUG);
787 if (!empty($json["fn"])) {
788 $data["name"] = $json["fn"];
791 if (!empty($json["addr"])) {
792 $data["addr"] = $json["addr"];
795 if (!empty($json["nick"])) {
796 $data["nick"] = $json["nick"];
799 if (!empty($json["guid"])) {
800 $data["guid"] = $json["guid"];
803 if (!empty($json["comm"])) {
804 $data["community"] = $json["comm"];
807 if (!empty($json["tags"])) {
808 $keywords = implode(" ", $json["tags"]);
809 if ($keywords != "") {
810 $data["keywords"] = $keywords;
814 $location = Profile::formatLocation($json);
816 $data["location"] = $location;
819 if (!empty($json["about"])) {
820 $data["about"] = $json["about"];
823 if (!empty($json["key"])) {
824 $data["pubkey"] = $json["key"];
827 if (!empty($json["photo"])) {
828 $data["photo"] = $json["photo"];
831 if (!empty($json["dfrn-request"])) {
832 $data["request"] = $json["dfrn-request"];
835 if (!empty($json["dfrn-confirm"])) {
836 $data["confirm"] = $json["dfrn-confirm"];
839 if (!empty($json["dfrn-notify"])) {
840 $data["notify"] = $json["dfrn-notify"];
843 if (!empty($json["dfrn-poll"])) {
844 $data["poll"] = $json["dfrn-poll"];
851 * @brief Check for valid DFRN data
853 * @param array $data DFRN data
855 * @return int Number of errors
857 public static function validDfrn($data)
860 if (!isset($data['key'])) {
863 if (!isset($data['dfrn-request'])) {
866 if (!isset($data['dfrn-confirm'])) {
869 if (!isset($data['dfrn-notify'])) {
872 if (!isset($data['dfrn-poll'])) {
879 * @brief Fetch data from a DFRN profile page and via "noscrape"
881 * @param string $profile_link Link to the profile page
883 * @return array profile data
885 public static function profile($profile_link)
889 logger("Check profile ".$profile_link, LOGGER_DEBUG);
891 // Fetch data via noscrape - this is faster
892 $noscrape_url = str_replace(["/hcard/", "/profile/"], "/noscrape/", $profile_link);
893 $data = self::pollNoscrape($noscrape_url, $data);
895 if (!isset($data["notify"])
896 || !isset($data["confirm"])
897 || !isset($data["request"])
898 || !isset($data["poll"])
899 || !isset($data["poco"])
900 || !isset($data["name"])
901 || !isset($data["photo"])
903 $data = self::pollHcard($profile_link, $data, true);
907 $prof_data["addr"] = $data["addr"];
908 $prof_data["nick"] = $data["nick"];
909 $prof_data["dfrn-request"] = $data["request"];
910 $prof_data["dfrn-confirm"] = $data["confirm"];
911 $prof_data["dfrn-notify"] = $data["notify"];
912 $prof_data["dfrn-poll"] = $data["poll"];
913 $prof_data["dfrn-poco"] = $data["poco"];
914 $prof_data["photo"] = $data["photo"];
915 $prof_data["fn"] = $data["name"];
916 $prof_data["key"] = $data["pubkey"];
918 logger("Result for profile ".$profile_link.": ".print_r($prof_data, true), LOGGER_DEBUG);
924 * @brief Check for DFRN contact
926 * @param array $webfinger Webfinger data
928 * @return array DFRN data
930 private static function dfrn($webfinger)
934 foreach ($webfinger["links"] as $link) {
935 if (($link["rel"] == NAMESPACE_DFRN) && ($link["href"] != "")) {
936 $data["network"] = NETWORK_DFRN;
937 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
938 $data["poll"] = $link["href"];
939 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
940 $data["url"] = $link["href"];
941 } elseif (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
942 $hcard_url = $link["href"];
943 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
944 $data["poco"] = $link["href"];
945 } elseif (($link["rel"] == "http://webfinger.net/rel/avatar") && ($link["href"] != "")) {
946 $data["photo"] = $link["href"];
947 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
948 $data["baseurl"] = trim($link["href"], '/');
949 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
950 $data["guid"] = $link["href"];
951 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
952 $data["pubkey"] = base64_decode($link["href"]);
954 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
955 if (strstr($data["pubkey"], 'RSA ')) {
956 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
961 if (is_array($webfinger["aliases"])) {
962 foreach ($webfinger["aliases"] as $alias) {
963 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
964 $data["alias"] = $alias;
965 } elseif (substr($alias, 0, 5) == 'acct:') {
966 $data["addr"] = substr($alias, 5);
971 if (substr($webfinger["subject"], 0, 5) == "acct:") {
972 $data["addr"] = substr($webfinger["subject"], 5);
975 if (!isset($data["network"]) || ($hcard_url == "")) {
979 // Fetch data via noscrape - this is faster
980 $noscrape_url = str_replace("/hcard/", "/noscrape/", $hcard_url);
981 $data = self::pollNoscrape($noscrape_url, $data);
983 if (isset($data["notify"])
984 && isset($data["confirm"])
985 && isset($data["request"])
986 && isset($data["poll"])
987 && isset($data["name"])
988 && isset($data["photo"])
993 $data = self::pollHcard($hcard_url, $data, true);
999 * @brief Poll the hcard page (Diaspora and Friendica specific)
1001 * @param string $hcard_url Link to the hcard page
1002 * @param array $data The already fetched data
1003 * @param boolean $dfrn Poll DFRN specific data
1005 * @return array hcard data
1007 private static function pollHcard($hcard_url, $data, $dfrn = false)
1009 $ret = Network::curl($hcard_url);
1010 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1013 $content = $ret['body'];
1018 $doc = new DOMDocument();
1019 if (!@$doc->loadHTML($content)) {
1023 $xpath = new DomXPath($doc);
1025 $vcards = $xpath->query("//div[contains(concat(' ', @class, ' '), ' vcard ')]");
1026 if (!is_object($vcards)) {
1030 if ($vcards->length > 0) {
1031 $vcard = $vcards->item(0);
1033 // We have to discard the guid from the hcard in favour of the guid from lrdd
1034 // Reason: Hubzilla doesn't use the value "uid" in the hcard like Diaspora does.
1035 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' uid ')]", $vcard); // */
1036 if (($search->length > 0) && ($data["guid"] == "")) {
1037 $data["guid"] = $search->item(0)->nodeValue;
1040 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' nickname ')]", $vcard); // */
1041 if ($search->length > 0) {
1042 $data["nick"] = $search->item(0)->nodeValue;
1045 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' fn ')]", $vcard); // */
1046 if ($search->length > 0) {
1047 $data["name"] = $search->item(0)->nodeValue;
1050 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' searchable ')]", $vcard); // */
1051 if ($search->length > 0) {
1052 $data["searchable"] = $search->item(0)->nodeValue;
1055 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' key ')]", $vcard); // */
1056 if ($search->length > 0) {
1057 $data["pubkey"] = $search->item(0)->nodeValue;
1058 if (strstr($data["pubkey"], 'RSA ')) {
1059 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1063 $search = $xpath->query("//*[@id='pod_location']", $vcard); // */
1064 if ($search->length > 0) {
1065 $data["baseurl"] = trim($search->item(0)->nodeValue, "/");
1070 $photos = $xpath->query("//*[contains(concat(' ', @class, ' '), ' photo ') or contains(concat(' ', @class, ' '), ' avatar ')]", $vcard); // */
1071 foreach ($photos as $photo) {
1073 foreach ($photo->attributes as $attribute) {
1074 $attr[$attribute->name] = trim($attribute->value);
1077 if (isset($attr["src"]) && isset($attr["width"])) {
1078 $avatar[$attr["width"]] = $attr["src"];
1081 // We don't have a width. So we just take everything that we got.
1082 // This is a Hubzilla workaround which doesn't send a width.
1083 if ((sizeof($avatar) == 0) && !empty($attr["src"])) {
1084 $avatar[] = $attr["src"];
1088 if (sizeof($avatar)) {
1090 $data["photo"] = self::fixAvatar(array_pop($avatar), $data["baseurl"]);
1094 // Poll DFRN specific data
1095 $search = $xpath->query("//link[contains(concat(' ', @rel), ' dfrn-')]");
1096 if ($search->length > 0) {
1097 foreach ($search as $link) {
1098 //$data["request"] = $search->item(0)->nodeValue;
1100 foreach ($link->attributes as $attribute) {
1101 $attr[$attribute->name] = trim($attribute->value);
1104 $data[substr($attr["rel"], 5)] = $attr["href"];
1108 // Older Friendica versions had used the "uid" field differently than newer versions
1109 if ($data["nick"] == $data["guid"]) {
1110 unset($data["guid"]);
1119 * @brief Check for Diaspora contact
1121 * @param array $webfinger Webfinger data
1123 * @return array Diaspora data
1125 private static function diaspora($webfinger)
1129 foreach ($webfinger["links"] as $link) {
1130 if (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
1131 $hcard_url = $link["href"];
1132 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
1133 $data["baseurl"] = trim($link["href"], '/');
1134 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
1135 $data["guid"] = $link["href"];
1136 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
1137 $data["url"] = $link["href"];
1138 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1139 $data["poll"] = $link["href"];
1140 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
1141 $data["poco"] = $link["href"];
1142 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1143 $data["notify"] = $link["href"];
1144 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
1145 $data["pubkey"] = base64_decode($link["href"]);
1147 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
1148 if (strstr($data["pubkey"], 'RSA ')) {
1149 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1154 if (!isset($data["url"]) || ($hcard_url == "")) {
1158 if (is_array($webfinger["aliases"])) {
1159 foreach ($webfinger["aliases"] as $alias) {
1160 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
1161 $data["alias"] = $alias;
1162 } elseif (substr($alias, 0, 5) == 'acct:') {
1163 $data["addr"] = substr($alias, 5);
1168 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == 'acct:')) {
1169 $data["addr"] = substr($webfinger["subject"], 5);
1172 // Fetch further information from the hcard
1173 $data = self::pollHcard($hcard_url, $data);
1179 if (isset($data["url"])
1180 && isset($data["guid"])
1181 && isset($data["baseurl"])
1182 && isset($data["pubkey"])
1183 && ($hcard_url != "")
1185 $data["network"] = NETWORK_DIASPORA;
1187 // The Diaspora handle must always be lowercase
1188 $data["addr"] = strtolower($data["addr"]);
1190 // We have to overwrite the detected value for "notify" since Hubzilla doesn't send it
1191 $data["notify"] = $data["baseurl"] . "/receive/users/" . $data["guid"];
1192 $data["batch"] = $data["baseurl"] . "/receive/public";
1201 * @brief Check for OStatus contact
1203 * @param array $webfinger Webfinger data
1204 * @param bool $short Short detection mode
1206 * @return array|bool OStatus data or "false" on error or "true" on short mode
1208 private static function ostatus($webfinger, $short = false)
1212 if (is_array($webfinger["aliases"])) {
1213 foreach ($webfinger["aliases"] as $alias) {
1214 if (strstr($alias, "@") && !strstr(normalise_link($alias), "http://")) {
1215 $data["addr"] = str_replace('acct:', '', $alias);
1220 if (is_string($webfinger["subject"]) && strstr($webfinger["subject"], "@")
1221 && !strstr(normalise_link($webfinger["subject"]), "http://")
1223 $data["addr"] = str_replace('acct:', '', $webfinger["subject"]);
1227 if (is_array($webfinger["links"])) {
1228 foreach ($webfinger["links"] as $link) {
1229 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1230 && ($link["type"] == "text/html")
1231 && ($link["href"] != "")
1233 $data["url"] = $link["href"];
1234 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1235 $data["notify"] = $link["href"];
1236 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1237 $data["poll"] = $link["href"];
1238 } elseif (($link["rel"] == "magic-public-key") && ($link["href"] != "")) {
1239 $pubkey = $link["href"];
1241 if (substr($pubkey, 0, 5) === 'data:') {
1242 if (strstr($pubkey, ',')) {
1243 $pubkey = substr($pubkey, strpos($pubkey, ',') + 1);
1245 $pubkey = substr($pubkey, 5);
1247 } elseif (normalise_link($pubkey) == 'http://') {
1248 $ret = Network::curl($pubkey);
1249 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1252 $pubkey = $ret['body'];
1255 $key = explode(".", $pubkey);
1257 if (sizeof($key) >= 3) {
1258 $m = base64url_decode($key[1]);
1259 $e = base64url_decode($key[2]);
1260 $data["pubkey"] = Crypto::meToPem($m, $e);
1266 if (isset($data["notify"]) && isset($data["pubkey"])
1267 && isset($data["poll"])
1268 && isset($data["url"])
1270 $data["network"] = NETWORK_OSTATUS;
1279 // Fetch all additional data from the feed
1280 $ret = Network::curl($data["poll"]);
1281 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1284 $feed = $ret['body'];
1285 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1290 if ($feed_data["header"]["author-name"] != "") {
1291 $data["name"] = $feed_data["header"]["author-name"];
1293 if ($feed_data["header"]["author-nick"] != "") {
1294 $data["nick"] = $feed_data["header"]["author-nick"];
1296 if ($feed_data["header"]["author-avatar"] != "") {
1297 $data["photo"] = self::fixAvatar($feed_data["header"]["author-avatar"], $data["url"]);
1299 if ($feed_data["header"]["author-id"] != "") {
1300 $data["alias"] = $feed_data["header"]["author-id"];
1302 if ($feed_data["header"]["author-location"] != "") {
1303 $data["location"] = $feed_data["header"]["author-location"];
1305 if ($feed_data["header"]["author-about"] != "") {
1306 $data["about"] = $feed_data["header"]["author-about"];
1308 // OStatus has serious issues when the the url doesn't fit (ssl vs. non ssl)
1309 // So we take the value that we just fetched, although the other one worked as well
1310 if ($feed_data["header"]["author-link"] != "") {
1311 $data["url"] = $feed_data["header"]["author-link"];
1314 if (($data['poll'] == $data['url']) && ($data["alias"] != '')) {
1315 $data['url'] = $data["alias"];
1316 $data["alias"] = '';
1319 /// @todo Fetch location and "about" from the feed as well
1324 * @brief Fetch data from a pump.io profile page
1326 * @param string $profile_link Link to the profile page
1328 * @return array profile data
1330 private static function pumpioProfileData($profile_link)
1332 $doc = new DOMDocument();
1333 if (!@$doc->loadHTMLFile($profile_link)) {
1337 $xpath = new DomXPath($doc);
1341 $data["name"] = $xpath->query("//span[contains(@class, 'p-name')]")->item(0)->nodeValue;
1343 if ($data["name"] == '') {
1344 // This is ugly - but pump.io doesn't seem to know a better way for it
1345 $data["name"] = trim($xpath->query("//h1[@class='media-header']")->item(0)->nodeValue);
1346 $pos = strpos($data["name"], chr(10));
1348 $data["name"] = trim(substr($data["name"], 0, $pos));
1352 $data["location"] = $xpath->query("//p[contains(@class, 'p-locality')]")->item(0)->nodeValue;
1354 if ($data["location"] == '') {
1355 $data["location"] = $xpath->query("//p[contains(@class, 'location')]")->item(0)->nodeValue;
1358 $data["about"] = $xpath->query("//p[contains(@class, 'p-note')]")->item(0)->nodeValue;
1360 if ($data["about"] == '') {
1361 $data["about"] = $xpath->query("//p[contains(@class, 'summary')]")->item(0)->nodeValue;
1364 $avatar = $xpath->query("//img[contains(@class, 'u-photo')]")->item(0);
1366 $avatar = $xpath->query("//img[@class='img-rounded media-object']")->item(0);
1369 foreach ($avatar->attributes as $attribute) {
1370 if ($attribute->name == "src") {
1371 $data["photo"] = trim($attribute->value);
1380 * @brief Check for pump.io contact
1382 * @param array $webfinger Webfinger data
1384 * @return array pump.io data
1386 private static function pumpio($webfinger, $addr)
1389 foreach ($webfinger["links"] as $link) {
1390 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1391 && ($link["type"] == "text/html")
1392 && ($link["href"] != "")
1394 $data["url"] = $link["href"];
1395 } elseif (($link["rel"] == "activity-inbox") && ($link["href"] != "")) {
1396 $data["notify"] = $link["href"];
1397 } elseif (($link["rel"] == "activity-outbox") && ($link["href"] != "")) {
1398 $data["poll"] = $link["href"];
1399 } elseif (($link["rel"] == "dialback") && ($link["href"] != "")) {
1400 $data["dialback"] = $link["href"];
1403 if (isset($data["poll"]) && isset($data["notify"])
1404 && isset($data["dialback"])
1405 && isset($data["url"])
1407 // by now we use these fields only for the network type detection
1408 // So we unset all data that isn't used at the moment
1409 unset($data["dialback"]);
1411 $data["network"] = NETWORK_PUMPIO;
1416 $profile_data = self::pumpioProfileData($data["url"]);
1418 if (!$profile_data) {
1422 $data = array_merge($data, $profile_data);
1424 if (($addr != '') && ($data['name'] != '')) {
1425 $name = trim(str_replace($addr, '', $data['name']));
1427 $data['name'] = $name;
1435 * @brief Check page for feed link
1437 * @param string $url Page link
1439 * @return string feed link
1441 private static function getFeedLink($url)
1443 $doc = new DOMDocument();
1445 if (!@$doc->loadHTMLFile($url)) {
1449 $xpath = new DomXPath($doc);
1451 //$feeds = $xpath->query("/html/head/link[@type='application/rss+xml']");
1452 $feeds = $xpath->query("/html/head/link[@type='application/rss+xml' and @rel='alternate']");
1453 if (!is_object($feeds)) {
1457 if ($feeds->length == 0) {
1463 foreach ($feeds as $feed) {
1465 foreach ($feed->attributes as $attribute) {
1466 $attr[$attribute->name] = trim($attribute->value);
1469 if ($feed_url == "") {
1470 $feed_url = $attr["href"];
1478 * @brief Check for feed contact
1480 * @param string $url Profile link
1481 * @param boolean $probe Do a probe if the page contains a feed link
1483 * @return array feed data
1485 private static function feed($url, $probe = true)
1487 $ret = Network::curl($url);
1488 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1491 $feed = $ret['body'];
1492 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1499 $feed_url = self::getFeedLink($url);
1505 return self::feed($feed_url, false);
1508 if ($feed_data["header"]["author-name"] != "") {
1509 $data["name"] = $feed_data["header"]["author-name"];
1512 if ($feed_data["header"]["author-nick"] != "") {
1513 $data["nick"] = $feed_data["header"]["author-nick"];
1516 if ($feed_data["header"]["author-avatar"] != "") {
1517 $data["photo"] = $feed_data["header"]["author-avatar"];
1520 if ($feed_data["header"]["author-id"] != "") {
1521 $data["alias"] = $feed_data["header"]["author-id"];
1524 $data["url"] = $url;
1525 $data["poll"] = $url;
1527 if ($feed_data["header"]["author-link"] != "") {
1528 $data["baseurl"] = $feed_data["header"]["author-link"];
1530 $data["baseurl"] = $data["url"];
1533 $data["network"] = NETWORK_FEED;
1539 * @brief Check for mail contact
1541 * @param string $uri Profile link
1542 * @param integer $uid User ID
1544 * @return array mail data
1546 private static function mail($uri, $uid)
1548 if (!Network::isEmailDomainValid($uri)) {
1556 $x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1", intval($uid));
1558 $r = q("SELECT * FROM `mailacct` WHERE `uid` = %d AND `server` != '' LIMIT 1", intval($uid));
1560 if (DBM::is_result($x) && DBM::is_result($r)) {
1561 $mailbox = Email::constructMailboxName($r[0]);
1563 openssl_private_decrypt(hex2bin($r[0]['pass']), $password, $x[0]['prvkey']);
1564 $mbox = Email::connect($mailbox, $r[0]['user'], $password);
1570 $msgs = Email::poll($mbox, $uri);
1571 logger('searching '.$uri.', '.count($msgs).' messages found.', LOGGER_DEBUG);
1573 if (!count($msgs)) {
1577 $phost = substr($uri, strpos($uri, '@') + 1);
1580 $data["addr"] = $uri;
1581 $data["network"] = NETWORK_MAIL;
1582 $data["name"] = substr($uri, 0, strpos($uri, '@'));
1583 $data["nick"] = $data["name"];
1584 $data["photo"] = Network::lookupAvatarByEmail($uri);
1585 $data["url"] = 'mailto:'.$uri;
1586 $data["notify"] = 'smtp '.random_string();
1587 $data["poll"] = 'email '.random_string();
1589 $x = Email::messageMeta($mbox, $msgs[0]);
1590 if (stristr($x[0]->from, $uri)) {
1591 $adr = imap_rfc822_parse_adrlist($x[0]->from, '');
1592 } elseif (stristr($x[0]->to, $uri)) {
1593 $adr = imap_rfc822_parse_adrlist($x[0]->to, '');
1596 foreach ($adr as $feadr) {
1597 if ((strcasecmp($feadr->mailbox, $data["name"]) == 0)
1598 &&(strcasecmp($feadr->host, $phost) == 0)
1599 && (strlen($feadr->personal))
1601 $personal = imap_mime_header_decode($feadr->personal);
1603 foreach ($personal as $perspart) {
1604 if ($perspart->charset != "default") {
1605 $data["name"] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
1607 $data["name"] .= $perspart->text;
1611 $data["name"] = notags($data["name"]);
1615 if (!empty($mbox)) {
1623 * @brief Mix two paths together to possibly fix missing parts
1625 * @param string $avatar Path to the avatar
1626 * @param string $base Another path that is hopefully complete
1628 * @return string fixed avatar path
1630 public static function fixAvatar($avatar, $base)
1632 $base_parts = parse_url($base);
1634 // Remove all parts that could create a problem
1635 unset($base_parts['path']);
1636 unset($base_parts['query']);
1637 unset($base_parts['fragment']);
1639 $avatar_parts = parse_url($avatar);
1642 $parts = array_merge($base_parts, $avatar_parts);
1644 // And put them together again
1645 $scheme = isset($parts['scheme']) ? $parts['scheme'] . '://' : '';
1646 $host = isset($parts['host']) ? $parts['host'] : '';
1647 $port = isset($parts['port']) ? ':' . $parts['port'] : '';
1648 $path = isset($parts['path']) ? $parts['path'] : '';
1649 $query = isset($parts['query']) ? '?' . $parts['query'] : '';
1650 $fragment = isset($parts['fragment']) ? '#' . $parts['fragment'] : '';
1652 $fixed = $scheme.$host.$port.$path.$query.$fragment;
1654 logger('Base: '.$base.' - Avatar: '.$avatar.' - Fixed: '.$fixed, LOGGER_DATA);