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;
23 use Friendica\Util\DateTimeFormat;
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($data['nick'], ' ')));
360 if (!empty(self::$baseurl)) {
361 $data["baseurl"] = self::$baseurl;
364 if (empty($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 // When the gcontact doesn't exist, the value "true" will trigger an insert.
415 // In difference to the public contacts we want to have every contact
416 // in the world in our global contacts.
420 // These values have to be set only on insert
421 $fields['photo'] = $data['photo'];
422 $fields['created'] = DateTimeFormat::utcNow();
425 dba::update('gcontact', $fields, $condition, $old_fields);
427 $fields = ['name' => $data['name'],
428 'nick' => $data['nick'],
429 'url' => $data['url'],
430 'addr' => $data['addr'],
431 'alias' => $data['alias'],
432 'keywords' => $data['keywords'],
433 'location' => $data['location'],
434 'about' => $data['about'],
435 'batch' => $data['batch'],
436 'notify' => $data['notify'],
437 'poll' => $data['poll'],
438 'request' => $data['request'],
439 'confirm' => $data['confirm'],
440 'poco' => $data['poco'],
441 'network' => $data['network'],
442 'pubkey' => $data['pubkey'],
443 'priority' => $data['priority'],
445 'rel' => CONTACT_IS_SHARING];
449 foreach ($fields as $key => $val) {
451 unset($fields[$key]);
453 $fieldnames[] = $key;
457 $condition = ['nurl' => normalise_link($data["url"]), 'self' => false, 'uid' => 0];
459 // "$old_fields" will return a "false" when the contact doesn't exist.
460 // This won't trigger an insert. This is intended, since we only need
461 // public contacts for everyone we store items from.
462 // We don't need to store every contact on the planet.
463 $old_fields = dba::selectFirst('contact', $fieldnames, $condition);
465 $fields['name-date'] = DateTimeFormat::utcNow();
466 $fields['uri-date'] = DateTimeFormat::utcNow();
467 $fields['success_update'] = DateTimeFormat::utcNow();
469 dba::update('contact', $fields, $condition, $old_fields);
477 * @brief Switch the scheme of an url between http and https
479 * @param string $url URL
481 * @return string switched URL
483 private static function switchScheme($url)
485 $parts = parse_url($url);
487 if (!isset($parts['scheme'])) {
491 if ($parts['scheme'] == 'http') {
492 $url = str_replace('http://', 'https://', $url);
493 } elseif ($parts['scheme'] == 'https') {
494 $url = str_replace('https://', 'http://', $url);
501 * @brief Checks if a profile url should be OStatus but only provides partial information
503 * @param array $webfinger Webfinger data
504 * @param string $lrdd Path template for webfinger request
505 * @param string $type type
507 * @return array fixed webfinger data
509 private static function fixOStatus($webfinger, $lrdd, $type)
511 if (empty($webfinger['links']) || empty($webfinger['subject'])) {
518 foreach ($webfinger['links'] as $link) {
519 if ($link['rel'] == NAMESPACE_OSTATUSSUB) {
522 if ($link['rel'] == 'magic-public-key') {
527 if (!$is_ostatus || $has_key) {
531 $url = self::switchScheme($webfinger['subject']);
532 $path = str_replace('{uri}', urlencode($url), $lrdd);
533 $webfinger2 = self::webfinger($path, $type);
535 // Is the new webfinger detectable as OStatus?
536 if (self::ostatus($webfinger2, true)) {
537 $webfinger = $webfinger2;
544 * @brief Fetch information (protocol endpoints and user information) about a given uri
546 * This function is only called by the "uri" function that adds caching and rearranging of data.
548 * @param string $uri Address that should be probed
549 * @param string $network Test for this specific network
550 * @param integer $uid User ID for the probe (only used for mails)
552 * @return array uri data
554 private static function detect($uri, $network, $uid)
556 $parts = parse_url($uri);
558 if (!empty($parts["scheme"]) && !empty($parts["host"]) && !empty($parts["path"])) {
559 $host = $parts["host"];
560 if (!empty($parts["port"])) {
561 $host .= ':'.$parts["port"];
564 if ($host == 'twitter.com') {
565 return ["network" => NETWORK_TWITTER];
567 $lrdd = self::hostMeta($host);
569 if (is_bool($lrdd)) {
573 $path_parts = explode("/", trim($parts["path"], "/"));
575 while (!$lrdd && (sizeof($path_parts) > 1)) {
576 $host .= "/".array_shift($path_parts);
577 $lrdd = self::hostMeta($host);
580 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
581 return self::feed($uri);
583 $nick = array_pop($path_parts);
585 // Mastodon uses a "@" as prefix for usernames in their url format
586 $nick = ltrim($nick, '@');
588 $addr = $nick."@".$host;
589 } elseif (strstr($uri, '@')) {
590 // If the URI starts with "mailto:" then jump directly to the mail detection
591 if (strpos($uri, 'mailto:') !== false) {
592 $uri = str_replace('mailto:', '', $uri);
593 return self::mail($uri, $uid);
596 if ($network == NETWORK_MAIL) {
597 return self::mail($uri, $uid);
599 // Remove "acct:" from the URI
600 $uri = str_replace('acct:', '', $uri);
602 $host = substr($uri, strpos($uri, '@') + 1);
603 $nick = substr($uri, 0, strpos($uri, '@'));
605 if (strpos($uri, '@twitter.com')) {
606 return ["network" => NETWORK_TWITTER];
608 $lrdd = self::hostMeta($host);
610 if (is_bool($lrdd)) {
615 logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
616 return self::mail($uri, $uid);
620 logger("Uri ".$uri." was not detectable", LOGGER_DEBUG);
626 /// @todo Do we need the prefix "acct:" or "acct://"?
628 foreach ($lrdd as $type => $template) {
633 // At first try it with the given uri
634 $path = str_replace('{uri}', urlencode($uri), $template);
635 $webfinger = self::webfinger($path, $type);
637 // Fix possible problems with GNU Social probing to wrong scheme
638 $webfinger = self::fixOStatus($webfinger, $template, $type);
640 // We cannot be sure that the detected address was correct, so we don't use the values
641 if ($webfinger && ($uri != $addr)) {
646 // Try webfinger with the address (user@domain.tld)
648 $path = str_replace('{uri}', urlencode($addr), $template);
649 $webfinger = self::webfinger($path, $type);
652 // Mastodon needs to have it with "acct:"
654 $path = str_replace('{uri}', urlencode("acct:".$addr), $template);
655 $webfinger = self::webfinger($path, $type);
660 return self::feed($uri);
665 logger("Probing ".$uri, LOGGER_DEBUG);
667 if (in_array($network, ["", NETWORK_DFRN])) {
668 $result = self::dfrn($webfinger);
670 if ((!$result && ($network == "")) || ($network == NETWORK_DIASPORA)) {
671 $result = self::diaspora($webfinger);
673 if ((!$result && ($network == "")) || ($network == NETWORK_OSTATUS)) {
674 $result = self::ostatus($webfinger);
676 if ((!$result && ($network == "")) || ($network == NETWORK_PUMPIO)) {
677 $result = self::pumpio($webfinger, $addr);
679 if ((!$result && ($network == "")) || ($network == NETWORK_FEED)) {
680 $result = self::feed($uri);
682 // We overwrite the detected nick with our try if the previois routines hadn't detected it.
683 // Additionally it is overwritten when the nickname doesn't make sense (contains spaces).
684 if ((empty($result["nick"]) || (strstr($result["nick"], " "))) && ($nick != "")) {
685 $result["nick"] = $nick;
688 if (empty($result["addr"]) && ($addr != "")) {
689 $result["addr"] = $addr;
693 logger($uri." is ".$result["network"], LOGGER_DEBUG);
695 if (empty($result["baseurl"])) {
696 $pos = strpos($result["url"], $host);
698 $result["baseurl"] = substr($result["url"], 0, $pos).$host;
705 * @brief Perform a webfinger request.
707 * For details see RFC 7033: <https://tools.ietf.org/html/rfc7033>
709 * @param string $url Address that should be probed
710 * @param string $type type
712 * @return array webfinger data
714 private static function webfinger($url, $type)
716 $xrd_timeout = Config::get('system', 'xrd_timeout', 20);
719 $ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => $type]);
720 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
723 $data = $ret['body'];
725 $webfinger = json_decode($data, true);
726 if (is_array($webfinger)) {
727 if (!isset($webfinger["links"])) {
728 logger("No json webfinger links for ".$url, LOGGER_DEBUG);
734 // If it is not JSON, maybe it is XML
735 $xrd = XML::parseString($data, false);
736 if (!is_object($xrd)) {
737 logger("No webfinger data retrievable for ".$url, LOGGER_DEBUG);
741 $xrd_arr = XML::elementToArray($xrd);
742 if (!isset($xrd_arr["xrd"]["link"])) {
743 logger("No XML webfinger links for ".$url, LOGGER_DEBUG);
749 if (!empty($xrd_arr["xrd"]["subject"])) {
750 $webfinger["subject"] = $xrd_arr["xrd"]["subject"];
753 if (!empty($xrd_arr["xrd"]["alias"])) {
754 $webfinger["aliases"] = $xrd_arr["xrd"]["alias"];
757 $webfinger["links"] = [];
759 foreach ($xrd_arr["xrd"]["link"] as $value => $data) {
760 if (!empty($data["@attributes"])) {
761 $attributes = $data["@attributes"];
762 } elseif ($value == "@attributes") {
768 $webfinger["links"][] = $attributes;
774 * @brief Poll the Friendica specific noscrape page.
776 * "noscrape" is a faster alternative to fetch the data from the hcard.
777 * This functionality was originally created for the directory.
779 * @param string $noscrape_url Link to the noscrape page
780 * @param array $data The already fetched data
782 * @return array noscrape data
784 private static function pollNoscrape($noscrape_url, $data)
786 $ret = Network::curl($noscrape_url);
787 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
790 $content = $ret['body'];
792 logger("Empty body for ".$noscrape_url, LOGGER_DEBUG);
796 $json = json_decode($content, true);
797 if (!is_array($json)) {
798 logger("No json data for ".$noscrape_url, LOGGER_DEBUG);
802 if (!empty($json["fn"])) {
803 $data["name"] = $json["fn"];
806 if (!empty($json["addr"])) {
807 $data["addr"] = $json["addr"];
810 if (!empty($json["nick"])) {
811 $data["nick"] = $json["nick"];
814 if (!empty($json["guid"])) {
815 $data["guid"] = $json["guid"];
818 if (!empty($json["comm"])) {
819 $data["community"] = $json["comm"];
822 if (!empty($json["tags"])) {
823 $keywords = implode(" ", $json["tags"]);
824 if ($keywords != "") {
825 $data["keywords"] = $keywords;
829 $location = Profile::formatLocation($json);
831 $data["location"] = $location;
834 if (!empty($json["about"])) {
835 $data["about"] = $json["about"];
838 if (!empty($json["key"])) {
839 $data["pubkey"] = $json["key"];
842 if (!empty($json["photo"])) {
843 $data["photo"] = $json["photo"];
846 if (!empty($json["dfrn-request"])) {
847 $data["request"] = $json["dfrn-request"];
850 if (!empty($json["dfrn-confirm"])) {
851 $data["confirm"] = $json["dfrn-confirm"];
854 if (!empty($json["dfrn-notify"])) {
855 $data["notify"] = $json["dfrn-notify"];
858 if (!empty($json["dfrn-poll"])) {
859 $data["poll"] = $json["dfrn-poll"];
866 * @brief Check for valid DFRN data
868 * @param array $data DFRN data
870 * @return int Number of errors
872 public static function validDfrn($data)
875 if (!isset($data['key'])) {
878 if (!isset($data['dfrn-request'])) {
881 if (!isset($data['dfrn-confirm'])) {
884 if (!isset($data['dfrn-notify'])) {
887 if (!isset($data['dfrn-poll'])) {
894 * @brief Fetch data from a DFRN profile page and via "noscrape"
896 * @param string $profile_link Link to the profile page
898 * @return array profile data
900 public static function profile($profile_link)
904 logger("Check profile ".$profile_link, LOGGER_DEBUG);
906 // Fetch data via noscrape - this is faster
907 $noscrape_url = str_replace(["/hcard/", "/profile/"], "/noscrape/", $profile_link);
908 $data = self::pollNoscrape($noscrape_url, $data);
910 if (!isset($data["notify"])
911 || !isset($data["confirm"])
912 || !isset($data["request"])
913 || !isset($data["poll"])
914 || !isset($data["poco"])
915 || !isset($data["name"])
916 || !isset($data["photo"])
918 $data = self::pollHcard($profile_link, $data, true);
922 $prof_data["addr"] = $data["addr"];
923 $prof_data["nick"] = $data["nick"];
924 $prof_data["dfrn-request"] = $data["request"];
925 $prof_data["dfrn-confirm"] = $data["confirm"];
926 $prof_data["dfrn-notify"] = $data["notify"];
927 $prof_data["dfrn-poll"] = $data["poll"];
928 $prof_data["dfrn-poco"] = $data["poco"];
929 $prof_data["photo"] = $data["photo"];
930 $prof_data["fn"] = $data["name"];
931 $prof_data["key"] = $data["pubkey"];
933 logger("Result for profile ".$profile_link.": ".print_r($prof_data, true), LOGGER_DEBUG);
939 * @brief Check for DFRN contact
941 * @param array $webfinger Webfinger data
943 * @return array DFRN data
945 private static function dfrn($webfinger)
949 foreach ($webfinger["links"] as $link) {
950 if (($link["rel"] == NAMESPACE_DFRN) && ($link["href"] != "")) {
951 $data["network"] = NETWORK_DFRN;
952 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
953 $data["poll"] = $link["href"];
954 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
955 $data["url"] = $link["href"];
956 } elseif (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
957 $hcard_url = $link["href"];
958 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
959 $data["poco"] = $link["href"];
960 } elseif (($link["rel"] == "http://webfinger.net/rel/avatar") && ($link["href"] != "")) {
961 $data["photo"] = $link["href"];
962 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
963 $data["baseurl"] = trim($link["href"], '/');
964 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
965 $data["guid"] = $link["href"];
966 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
967 $data["pubkey"] = base64_decode($link["href"]);
969 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
970 if (strstr($data["pubkey"], 'RSA ')) {
971 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
976 if (is_array($webfinger["aliases"])) {
977 foreach ($webfinger["aliases"] as $alias) {
978 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
979 $data["alias"] = $alias;
980 } elseif (substr($alias, 0, 5) == 'acct:') {
981 $data["addr"] = substr($alias, 5);
986 if (substr($webfinger["subject"], 0, 5) == "acct:") {
987 $data["addr"] = substr($webfinger["subject"], 5);
990 if (!isset($data["network"]) || ($hcard_url == "")) {
994 // Fetch data via noscrape - this is faster
995 $noscrape_url = str_replace("/hcard/", "/noscrape/", $hcard_url);
996 $data = self::pollNoscrape($noscrape_url, $data);
998 if (isset($data["notify"])
999 && isset($data["confirm"])
1000 && isset($data["request"])
1001 && isset($data["poll"])
1002 && isset($data["name"])
1003 && isset($data["photo"])
1008 $data = self::pollHcard($hcard_url, $data, true);
1014 * @brief Poll the hcard page (Diaspora and Friendica specific)
1016 * @param string $hcard_url Link to the hcard page
1017 * @param array $data The already fetched data
1018 * @param boolean $dfrn Poll DFRN specific data
1020 * @return array hcard data
1022 private static function pollHcard($hcard_url, $data, $dfrn = false)
1024 $ret = Network::curl($hcard_url);
1025 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1028 $content = $ret['body'];
1033 $doc = new DOMDocument();
1034 if (!@$doc->loadHTML($content)) {
1038 $xpath = new DomXPath($doc);
1040 $vcards = $xpath->query("//div[contains(concat(' ', @class, ' '), ' vcard ')]");
1041 if (!is_object($vcards)) {
1045 if ($vcards->length > 0) {
1046 $vcard = $vcards->item(0);
1048 // We have to discard the guid from the hcard in favour of the guid from lrdd
1049 // Reason: Hubzilla doesn't use the value "uid" in the hcard like Diaspora does.
1050 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' uid ')]", $vcard); // */
1051 if (($search->length > 0) && ($data["guid"] == "")) {
1052 $data["guid"] = $search->item(0)->nodeValue;
1055 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' nickname ')]", $vcard); // */
1056 if ($search->length > 0) {
1057 $data["nick"] = $search->item(0)->nodeValue;
1060 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' fn ')]", $vcard); // */
1061 if ($search->length > 0) {
1062 $data["name"] = $search->item(0)->nodeValue;
1065 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' searchable ')]", $vcard); // */
1066 if ($search->length > 0) {
1067 $data["searchable"] = $search->item(0)->nodeValue;
1070 $search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' key ')]", $vcard); // */
1071 if ($search->length > 0) {
1072 $data["pubkey"] = $search->item(0)->nodeValue;
1073 if (strstr($data["pubkey"], 'RSA ')) {
1074 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1078 $search = $xpath->query("//*[@id='pod_location']", $vcard); // */
1079 if ($search->length > 0) {
1080 $data["baseurl"] = trim($search->item(0)->nodeValue, "/");
1085 $photos = $xpath->query("//*[contains(concat(' ', @class, ' '), ' photo ') or contains(concat(' ', @class, ' '), ' avatar ')]", $vcard); // */
1086 foreach ($photos as $photo) {
1088 foreach ($photo->attributes as $attribute) {
1089 $attr[$attribute->name] = trim($attribute->value);
1092 if (isset($attr["src"]) && isset($attr["width"])) {
1093 $avatar[$attr["width"]] = $attr["src"];
1096 // We don't have a width. So we just take everything that we got.
1097 // This is a Hubzilla workaround which doesn't send a width.
1098 if ((sizeof($avatar) == 0) && !empty($attr["src"])) {
1099 $avatar[] = $attr["src"];
1103 if (sizeof($avatar)) {
1105 $data["photo"] = self::fixAvatar(array_pop($avatar), $data["baseurl"]);
1109 // Poll DFRN specific data
1110 $search = $xpath->query("//link[contains(concat(' ', @rel), ' dfrn-')]");
1111 if ($search->length > 0) {
1112 foreach ($search as $link) {
1113 //$data["request"] = $search->item(0)->nodeValue;
1115 foreach ($link->attributes as $attribute) {
1116 $attr[$attribute->name] = trim($attribute->value);
1119 $data[substr($attr["rel"], 5)] = $attr["href"];
1123 // Older Friendica versions had used the "uid" field differently than newer versions
1124 if ($data["nick"] == $data["guid"]) {
1125 unset($data["guid"]);
1134 * @brief Check for Diaspora contact
1136 * @param array $webfinger Webfinger data
1138 * @return array Diaspora data
1140 private static function diaspora($webfinger)
1144 foreach ($webfinger["links"] as $link) {
1145 if (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
1146 $hcard_url = $link["href"];
1147 } elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
1148 $data["baseurl"] = trim($link["href"], '/');
1149 } elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
1150 $data["guid"] = $link["href"];
1151 } elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
1152 $data["url"] = $link["href"];
1153 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1154 $data["poll"] = $link["href"];
1155 } elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
1156 $data["poco"] = $link["href"];
1157 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1158 $data["notify"] = $link["href"];
1159 } elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
1160 $data["pubkey"] = base64_decode($link["href"]);
1162 //if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
1163 if (strstr($data["pubkey"], 'RSA ')) {
1164 $data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
1169 if (!isset($data["url"]) || ($hcard_url == "")) {
1173 if (is_array($webfinger["aliases"])) {
1174 foreach ($webfinger["aliases"] as $alias) {
1175 if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
1176 $data["alias"] = $alias;
1177 } elseif (substr($alias, 0, 5) == 'acct:') {
1178 $data["addr"] = substr($alias, 5);
1183 if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == 'acct:')) {
1184 $data["addr"] = substr($webfinger["subject"], 5);
1187 // Fetch further information from the hcard
1188 $data = self::pollHcard($hcard_url, $data);
1194 if (isset($data["url"])
1195 && isset($data["guid"])
1196 && isset($data["baseurl"])
1197 && isset($data["pubkey"])
1198 && ($hcard_url != "")
1200 $data["network"] = NETWORK_DIASPORA;
1202 // The Diaspora handle must always be lowercase
1203 $data["addr"] = strtolower($data["addr"]);
1205 // We have to overwrite the detected value for "notify" since Hubzilla doesn't send it
1206 $data["notify"] = $data["baseurl"] . "/receive/users/" . $data["guid"];
1207 $data["batch"] = $data["baseurl"] . "/receive/public";
1216 * @brief Check for OStatus contact
1218 * @param array $webfinger Webfinger data
1219 * @param bool $short Short detection mode
1221 * @return array|bool OStatus data or "false" on error or "true" on short mode
1223 private static function ostatus($webfinger, $short = false)
1227 if (is_array($webfinger["aliases"])) {
1228 foreach ($webfinger["aliases"] as $alias) {
1229 if (strstr($alias, "@") && !strstr(normalise_link($alias), "http://")) {
1230 $data["addr"] = str_replace('acct:', '', $alias);
1235 if (is_string($webfinger["subject"]) && strstr($webfinger["subject"], "@")
1236 && !strstr(normalise_link($webfinger["subject"]), "http://")
1238 $data["addr"] = str_replace('acct:', '', $webfinger["subject"]);
1242 if (is_array($webfinger["links"])) {
1243 foreach ($webfinger["links"] as $link) {
1244 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1245 && ($link["type"] == "text/html")
1246 && ($link["href"] != "")
1248 $data["url"] = $link["href"];
1249 } elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
1250 $data["notify"] = $link["href"];
1251 } elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
1252 $data["poll"] = $link["href"];
1253 } elseif (($link["rel"] == "magic-public-key") && ($link["href"] != "")) {
1254 $pubkey = $link["href"];
1256 if (substr($pubkey, 0, 5) === 'data:') {
1257 if (strstr($pubkey, ',')) {
1258 $pubkey = substr($pubkey, strpos($pubkey, ',') + 1);
1260 $pubkey = substr($pubkey, 5);
1262 } elseif (normalise_link($pubkey) == 'http://') {
1263 $ret = Network::curl($pubkey);
1264 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1267 $pubkey = $ret['body'];
1270 $key = explode(".", $pubkey);
1272 if (sizeof($key) >= 3) {
1273 $m = base64url_decode($key[1]);
1274 $e = base64url_decode($key[2]);
1275 $data["pubkey"] = Crypto::meToPem($m, $e);
1281 if (isset($data["notify"]) && isset($data["pubkey"])
1282 && isset($data["poll"])
1283 && isset($data["url"])
1285 $data["network"] = NETWORK_OSTATUS;
1294 // Fetch all additional data from the feed
1295 $ret = Network::curl($data["poll"]);
1296 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1299 $feed = $ret['body'];
1300 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1305 if ($feed_data["header"]["author-name"] != "") {
1306 $data["name"] = $feed_data["header"]["author-name"];
1308 if ($feed_data["header"]["author-nick"] != "") {
1309 $data["nick"] = $feed_data["header"]["author-nick"];
1311 if ($feed_data["header"]["author-avatar"] != "") {
1312 $data["photo"] = self::fixAvatar($feed_data["header"]["author-avatar"], $data["url"]);
1314 if ($feed_data["header"]["author-id"] != "") {
1315 $data["alias"] = $feed_data["header"]["author-id"];
1317 if ($feed_data["header"]["author-location"] != "") {
1318 $data["location"] = $feed_data["header"]["author-location"];
1320 if ($feed_data["header"]["author-about"] != "") {
1321 $data["about"] = $feed_data["header"]["author-about"];
1323 // OStatus has serious issues when the the url doesn't fit (ssl vs. non ssl)
1324 // So we take the value that we just fetched, although the other one worked as well
1325 if ($feed_data["header"]["author-link"] != "") {
1326 $data["url"] = $feed_data["header"]["author-link"];
1329 if (($data['poll'] == $data['url']) && ($data["alias"] != '')) {
1330 $data['url'] = $data["alias"];
1331 $data["alias"] = '';
1334 /// @todo Fetch location and "about" from the feed as well
1339 * @brief Fetch data from a pump.io profile page
1341 * @param string $profile_link Link to the profile page
1343 * @return array profile data
1345 private static function pumpioProfileData($profile_link)
1347 $doc = new DOMDocument();
1348 if (!@$doc->loadHTMLFile($profile_link)) {
1352 $xpath = new DomXPath($doc);
1356 $data["name"] = $xpath->query("//span[contains(@class, 'p-name')]")->item(0)->nodeValue;
1358 if ($data["name"] == '') {
1359 // This is ugly - but pump.io doesn't seem to know a better way for it
1360 $data["name"] = trim($xpath->query("//h1[@class='media-header']")->item(0)->nodeValue);
1361 $pos = strpos($data["name"], chr(10));
1363 $data["name"] = trim(substr($data["name"], 0, $pos));
1367 $data["location"] = $xpath->query("//p[contains(@class, 'p-locality')]")->item(0)->nodeValue;
1369 if ($data["location"] == '') {
1370 $data["location"] = $xpath->query("//p[contains(@class, 'location')]")->item(0)->nodeValue;
1373 $data["about"] = $xpath->query("//p[contains(@class, 'p-note')]")->item(0)->nodeValue;
1375 if ($data["about"] == '') {
1376 $data["about"] = $xpath->query("//p[contains(@class, 'summary')]")->item(0)->nodeValue;
1379 $avatar = $xpath->query("//img[contains(@class, 'u-photo')]")->item(0);
1381 $avatar = $xpath->query("//img[@class='img-rounded media-object']")->item(0);
1384 foreach ($avatar->attributes as $attribute) {
1385 if ($attribute->name == "src") {
1386 $data["photo"] = trim($attribute->value);
1395 * @brief Check for pump.io contact
1397 * @param array $webfinger Webfinger data
1399 * @return array pump.io data
1401 private static function pumpio($webfinger, $addr)
1404 foreach ($webfinger["links"] as $link) {
1405 if (($link["rel"] == "http://webfinger.net/rel/profile-page")
1406 && ($link["type"] == "text/html")
1407 && ($link["href"] != "")
1409 $data["url"] = $link["href"];
1410 } elseif (($link["rel"] == "activity-inbox") && ($link["href"] != "")) {
1411 $data["notify"] = $link["href"];
1412 } elseif (($link["rel"] == "activity-outbox") && ($link["href"] != "")) {
1413 $data["poll"] = $link["href"];
1414 } elseif (($link["rel"] == "dialback") && ($link["href"] != "")) {
1415 $data["dialback"] = $link["href"];
1418 if (isset($data["poll"]) && isset($data["notify"])
1419 && isset($data["dialback"])
1420 && isset($data["url"])
1422 // by now we use these fields only for the network type detection
1423 // So we unset all data that isn't used at the moment
1424 unset($data["dialback"]);
1426 $data["network"] = NETWORK_PUMPIO;
1431 $profile_data = self::pumpioProfileData($data["url"]);
1433 if (!$profile_data) {
1437 $data = array_merge($data, $profile_data);
1439 if (($addr != '') && ($data['name'] != '')) {
1440 $name = trim(str_replace($addr, '', $data['name']));
1442 $data['name'] = $name;
1450 * @brief Check page for feed link
1452 * @param string $url Page link
1454 * @return string feed link
1456 private static function getFeedLink($url)
1458 $doc = new DOMDocument();
1460 if (!@$doc->loadHTMLFile($url)) {
1464 $xpath = new DomXPath($doc);
1466 //$feeds = $xpath->query("/html/head/link[@type='application/rss+xml']");
1467 $feeds = $xpath->query("/html/head/link[@type='application/rss+xml' and @rel='alternate']");
1468 if (!is_object($feeds)) {
1472 if ($feeds->length == 0) {
1478 foreach ($feeds as $feed) {
1480 foreach ($feed->attributes as $attribute) {
1481 $attr[$attribute->name] = trim($attribute->value);
1484 if ($feed_url == "") {
1485 $feed_url = $attr["href"];
1493 * @brief Check for feed contact
1495 * @param string $url Profile link
1496 * @param boolean $probe Do a probe if the page contains a feed link
1498 * @return array feed data
1500 private static function feed($url, $probe = true)
1502 $ret = Network::curl($url);
1503 if ($ret['errno'] == CURLE_OPERATION_TIMEDOUT) {
1506 $feed = $ret['body'];
1507 $feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
1514 $feed_url = self::getFeedLink($url);
1520 return self::feed($feed_url, false);
1523 if ($feed_data["header"]["author-name"] != "") {
1524 $data["name"] = $feed_data["header"]["author-name"];
1527 if ($feed_data["header"]["author-nick"] != "") {
1528 $data["nick"] = $feed_data["header"]["author-nick"];
1531 if ($feed_data["header"]["author-avatar"] != "") {
1532 $data["photo"] = $feed_data["header"]["author-avatar"];
1535 if ($feed_data["header"]["author-id"] != "") {
1536 $data["alias"] = $feed_data["header"]["author-id"];
1539 $data["url"] = $url;
1540 $data["poll"] = $url;
1542 if ($feed_data["header"]["author-link"] != "") {
1543 $data["baseurl"] = $feed_data["header"]["author-link"];
1545 $data["baseurl"] = $data["url"];
1548 $data["network"] = NETWORK_FEED;
1554 * @brief Check for mail contact
1556 * @param string $uri Profile link
1557 * @param integer $uid User ID
1559 * @return array mail data
1561 private static function mail($uri, $uid)
1563 if (!Network::isEmailDomainValid($uri)) {
1571 $x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1", intval($uid));
1573 $r = q("SELECT * FROM `mailacct` WHERE `uid` = %d AND `server` != '' LIMIT 1", intval($uid));
1575 if (DBM::is_result($x) && DBM::is_result($r)) {
1576 $mailbox = Email::constructMailboxName($r[0]);
1578 openssl_private_decrypt(hex2bin($r[0]['pass']), $password, $x[0]['prvkey']);
1579 $mbox = Email::connect($mailbox, $r[0]['user'], $password);
1585 $msgs = Email::poll($mbox, $uri);
1586 logger('searching '.$uri.', '.count($msgs).' messages found.', LOGGER_DEBUG);
1588 if (!count($msgs)) {
1592 $phost = substr($uri, strpos($uri, '@') + 1);
1595 $data["addr"] = $uri;
1596 $data["network"] = NETWORK_MAIL;
1597 $data["name"] = substr($uri, 0, strpos($uri, '@'));
1598 $data["nick"] = $data["name"];
1599 $data["photo"] = Network::lookupAvatarByEmail($uri);
1600 $data["url"] = 'mailto:'.$uri;
1601 $data["notify"] = 'smtp '.random_string();
1602 $data["poll"] = 'email '.random_string();
1604 $x = Email::messageMeta($mbox, $msgs[0]);
1605 if (stristr($x[0]->from, $uri)) {
1606 $adr = imap_rfc822_parse_adrlist($x[0]->from, '');
1607 } elseif (stristr($x[0]->to, $uri)) {
1608 $adr = imap_rfc822_parse_adrlist($x[0]->to, '');
1611 foreach ($adr as $feadr) {
1612 if ((strcasecmp($feadr->mailbox, $data["name"]) == 0)
1613 &&(strcasecmp($feadr->host, $phost) == 0)
1614 && (strlen($feadr->personal))
1616 $personal = imap_mime_header_decode($feadr->personal);
1618 foreach ($personal as $perspart) {
1619 if ($perspart->charset != "default") {
1620 $data["name"] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
1622 $data["name"] .= $perspart->text;
1626 $data["name"] = notags($data["name"]);
1630 if (!empty($mbox)) {
1638 * @brief Mix two paths together to possibly fix missing parts
1640 * @param string $avatar Path to the avatar
1641 * @param string $base Another path that is hopefully complete
1643 * @return string fixed avatar path
1645 public static function fixAvatar($avatar, $base)
1647 $base_parts = parse_url($base);
1649 // Remove all parts that could create a problem
1650 unset($base_parts['path']);
1651 unset($base_parts['query']);
1652 unset($base_parts['fragment']);
1654 $avatar_parts = parse_url($avatar);
1657 $parts = array_merge($base_parts, $avatar_parts);
1659 // And put them together again
1660 $scheme = isset($parts['scheme']) ? $parts['scheme'] . '://' : '';
1661 $host = isset($parts['host']) ? $parts['host'] : '';
1662 $port = isset($parts['port']) ? ':' . $parts['port'] : '';
1663 $path = isset($parts['path']) ? $parts['path'] : '';
1664 $query = isset($parts['query']) ? '?' . $parts['query'] : '';
1665 $fragment = isset($parts['fragment']) ? '#' . $parts['fragment'] : '';
1667 $fixed = $scheme.$host.$port.$path.$query.$fragment;
1669 logger('Base: '.$base.' - Avatar: '.$avatar.' - Fixed: '.$fixed, LOGGER_DATA);