friendica/src/Network/Probe.php

1700 lines
46 KiB
PHP
Raw Normal View History

2016-07-03 16:27:16 -04:00
<?php
/**
* @file src/Network/Probe.php
*/
2017-05-07 14:44:30 -04:00
namespace Friendica\Network;
2016-07-03 16:27:16 -04:00
/**
2017-05-07 14:44:30 -04:00
* @file src/Network/Probe.php
2016-07-08 16:55:39 -04:00
* @brief Functions for probing URL
2016-07-03 16:27:16 -04:00
*/
use DOMDocument;
use Friendica\Core\Cache;
2017-04-30 00:01:26 -04:00
use Friendica\Core\Config;
use Friendica\Core\System;
use Friendica\Database\DBA;
use Friendica\Model\Contact;
2017-12-07 08:57:35 -05:00
use Friendica\Model\Profile;
use Friendica\Protocol\Email;
2017-12-13 02:02:52 -05:00
use Friendica\Protocol\Feed;
use Friendica\Util\Crypto;
use Friendica\Util\DateTimeFormat;
use Friendica\Util\Network;
use Friendica\Util\XML;
2018-07-20 01:10:16 -04:00
use DomXPath;
2017-05-11 11:53:04 -04:00
require_once 'include/dba.php';
2016-07-03 16:27:16 -04:00
2016-07-08 16:55:39 -04:00
/**
* @brief This class contain functions for probing URL
*
*/
class Probe
{
private static $baseurl;
2016-07-07 17:04:30 -04:00
/**
* @brief Rearrange the array so that it always has the same order
*
* @param array $data Unordered data
*
* @return array Ordered data
*/
private static function rearrangeData($data)
{
$fields = ["name", "nick", "guid", "url", "addr", "alias",
2016-07-03 17:11:21 -04:00
"photo", "community", "keywords", "location", "about",
"batch", "notify", "poll", "request", "confirm", "poco",
"priority", "network", "pubkey", "baseurl"];
2016-07-03 16:27:16 -04:00
$newdata = [];
2017-05-08 11:06:57 -04:00
foreach ($fields as $field) {
if (isset($data[$field])) {
2016-07-03 16:27:16 -04:00
$newdata[$field] = $data[$field];
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
$newdata[$field] = "";
2017-05-08 11:06:57 -04:00
}
}
2016-07-03 16:27:16 -04:00
// We don't use the "priority" field anymore and replace it with a dummy.
$newdata["priority"] = 0;
return $newdata;
}
/**
* @brief Check if the hostname belongs to the own server
*
* @param string $host The hostname that is to be checked
*
* @return bool Does the testes hostname belongs to the own server?
*/
private static function ownHost($host)
{
$own_host = get_app()->get_hostname();
$parts = parse_url($host);
if (!isset($parts['scheme'])) {
$parts = parse_url('http://'.$host);
}
if (!isset($parts['host'])) {
return false;
}
return $parts['host'] == $own_host;
}
2016-07-03 16:27:16 -04:00
/**
* @brief Probes for webfinger path via "host-meta"
2017-05-08 11:06:57 -04:00
*
2018-02-02 08:53:40 -05:00
* We have to check if the servers in the future still will offer this.
* It seems as if it was dropped from the standard.
*
* @param string $host The host part of an url
2017-05-08 11:06:57 -04:00
*
* @return array with template and type of the webfinger template for JSON or XML
2016-07-03 16:27:16 -04:00
*/
private static function hostMeta($host)
{
2017-03-25 06:07:45 -04:00
// Reset the static variable
self::$baseurl = '';
2016-07-03 16:27:16 -04:00
$ssl_url = "https://".$host."/.well-known/host-meta";
$url = "http://".$host."/.well-known/host-meta";
2017-05-08 11:06:57 -04:00
$xrd_timeout = Config::get('system', 'xrd_timeout', 20);
2016-07-03 16:27:16 -04:00
$redirects = 0;
2017-05-23 16:19:39 -04:00
logger("Probing for ".$host, LOGGER_DEBUG);
$xrd = null;
2017-05-23 16:19:39 -04:00
$ret = Network::curl($ssl_url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
if ($ret['success']) {
$xml = $ret['body'];
$xrd = XML::parseString($xml, false);
$host_url = 'https://'.$host;
2017-02-01 16:35:01 -05:00
}
2016-07-03 16:27:16 -04:00
if (!is_object($xrd)) {
$ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => 'application/xrd+xml']);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-05-28 01:38:12 -04:00
logger("Probing timeout for ".$url, LOGGER_DEBUG);
return false;
}
$xml = $ret['body'];
$xrd = XML::parseString($xml, false);
$host_url = 'http://'.$host;
2016-07-03 16:27:16 -04:00
}
2017-05-08 11:06:57 -04:00
if (!is_object($xrd)) {
2017-05-28 01:38:12 -04:00
logger("No xrd object found for ".$host, LOGGER_DEBUG);
return [];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$links = XML::elementToArray($xrd);
2017-05-08 11:06:57 -04:00
if (!isset($links["xrd"]["link"])) {
2017-05-23 16:19:39 -04:00
logger("No xrd data found for ".$host, LOGGER_DEBUG);
return [];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$lrdd = [];
// The following webfinger path is defined in RFC 7033 https://tools.ietf.org/html/rfc7033
// Problem is that Hubzilla currently doesn't provide all data in the JSON webfinger
// compared to the XML webfinger. So this is commented out by now.
// $lrdd = array("application/jrd+json" => $host_url.'/.well-known/webfinger?resource={uri}');
2016-07-09 14:09:09 -04:00
2017-05-08 11:06:57 -04:00
foreach ($links["xrd"]["link"] as $value => $link) {
if (!empty($link["@attributes"])) {
2016-07-03 16:27:16 -04:00
$attributes = $link["@attributes"];
2017-05-08 11:06:57 -04:00
} elseif ($value == "@attributes") {
2016-07-03 16:27:16 -04:00
$attributes = $link;
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
continue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (($attributes["rel"] == "lrdd") && !empty($attributes["template"])) {
$type = (empty($attributes["type"]) ? '' : $attributes["type"]);
$lrdd[$type] = $attributes["template"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
self::$baseurl = "http://".$host;
2017-05-28 01:38:12 -04:00
logger("Probing successful for ".$host, LOGGER_DEBUG);
return $lrdd;
2016-07-03 16:27:16 -04:00
}
2016-07-09 14:09:09 -04:00
/**
* @brief Perform Webfinger lookup and return DFRN data
*
* Given an email style address, perform webfinger lookup and
* return the resulting DFRN profile URL, or if no DFRN profile URL
* is located, returns an OStatus subscription template (prefixed
* with the string 'stat:' to identify it as on OStatus template).
* If this isn't an email style address just return $webbie.
* Return an empty string if email-style addresses but webfinger fails,
* or if the resultant personal XRD doesn't contain a supported
* subscription/friend-request attribute.
*
* amended 7/9/2011 to return an hcard which could save potentially loading
* a lengthy content page to scrape dfrn attributes
*
* @param string $webbie Address that should be probed
* @param string $hcard_url Link to the hcard - is returned by reference
2016-07-09 14:09:09 -04:00
*
* @return string profile link
*/
public static function webfingerDfrn($webbie, &$hcard_url)
{
2016-07-09 14:09:09 -04:00
$profile_link = '';
2016-10-09 06:47:47 -04:00
$links = self::lrdd($webbie);
2017-05-08 11:06:57 -04:00
logger('webfingerDfrn: '.$webbie.':'.print_r($links, true), LOGGER_DATA);
2016-07-09 14:09:09 -04:00
if (count($links)) {
foreach ($links as $link) {
2017-05-08 11:06:57 -04:00
if ($link['@attributes']['rel'] === NAMESPACE_DFRN) {
2016-07-09 14:09:09 -04:00
$profile_link = $link['@attributes']['href'];
2017-05-08 11:06:57 -04:00
}
if (($link['@attributes']['rel'] === NAMESPACE_OSTATUSSUB) && ($profile_link == "")) {
2016-07-09 14:09:09 -04:00
$profile_link = 'stat:'.$link['@attributes']['template'];
2017-05-08 11:06:57 -04:00
}
if ($link['@attributes']['rel'] === 'http://microformats.org/profile/hcard') {
$hcard_url = $link['@attributes']['href'];
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
}
}
return $profile_link;
}
/**
* @brief Check an URI for LRDD data
*
* this is a replacement for the "lrdd" function.
2016-07-09 14:09:09 -04:00
* It isn't used in this class and has some redundancies in the code.
* When time comes we can check the existing calls for "lrdd" if we can rework them.
*
* @param string $uri Address that should be probed
*
* @return array uri data
*/
public static function lrdd($uri)
{
$lrdd = self::hostMeta($uri);
$webfinger = null;
2016-07-09 14:09:09 -04:00
if (is_bool($lrdd)) {
return [];
}
2016-07-09 14:09:09 -04:00
if (!$lrdd) {
$parts = @parse_url($uri);
2017-05-08 11:06:57 -04:00
if (!$parts) {
return [];
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
$host = $parts["host"];
if (!empty($parts["port"])) {
2017-05-07 09:11:11 -04:00
$host .= ':'.$parts["port"];
}
2016-07-09 14:09:09 -04:00
$path_parts = explode("/", trim($parts["path"], "/"));
$nick = array_pop($path_parts);
2016-07-09 14:09:09 -04:00
do {
$lrdd = self::hostMeta($host);
2016-07-09 14:09:09 -04:00
$host .= "/".array_shift($path_parts);
} while (!$lrdd && (sizeof($path_parts) > 0));
2016-07-09 14:09:09 -04:00
}
2017-05-08 11:06:57 -04:00
if (!$lrdd) {
2017-05-23 16:19:39 -04:00
logger("No lrdd data found for ".$uri, LOGGER_DEBUG);
return [];
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
foreach ($lrdd as $type => $template) {
2017-05-08 11:06:57 -04:00
if ($webfinger) {
2016-07-09 14:09:09 -04:00
continue;
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
$path = str_replace('{uri}', urlencode($uri), $template);
$webfinger = self::webfinger($path, $type);
if (!$webfinger && (strstr($uri, "@"))) {
$path = str_replace('{uri}', urlencode("acct:".$uri), $template);
$webfinger = self::webfinger($path, $type);
}
// Special treatment for Mastodon
// Problem is that Mastodon uses an URL format like http://domain.tld/@nick
// But the webfinger for this format fails.
if (!$webfinger && !empty($nick)) {
// Mastodon uses a "@" as prefix for usernames in their url format
$nick = ltrim($nick, '@');
$addr = $nick."@".$host;
$path = str_replace('{uri}', urlencode("acct:".$addr), $template);
$webfinger = self::webfinger($path, $type);
}
2016-07-09 14:09:09 -04:00
}
2017-05-08 11:06:57 -04:00
if (!is_array($webfinger["links"])) {
2017-05-23 16:19:39 -04:00
logger("No webfinger links found for ".$uri, LOGGER_DEBUG);
2016-07-09 14:09:09 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
$data = [];
2016-07-09 14:09:09 -04:00
2017-05-08 11:06:57 -04:00
foreach ($webfinger["links"] as $link) {
$data[] = ["@attributes" => $link];
2017-05-08 11:06:57 -04:00
}
2016-07-09 14:09:09 -04:00
2017-05-08 11:06:57 -04:00
if (is_array($webfinger["aliases"])) {
foreach ($webfinger["aliases"] as $alias) {
$data[] = ["@attributes" =>
["rel" => "alias",
"href" => $alias]];
2017-05-08 11:06:57 -04:00
}
}
2016-07-09 14:09:09 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
2016-07-10 12:44:48 -04:00
* @brief Fetch information (protocol endpoints and user information) about a given uri
2016-07-07 17:04:30 -04:00
*
* @param string $uri Address that should be probed
* @param string $network Test for this specific network
* @param integer $uid User ID for the probe (only used for mails)
* @param boolean $cache Use cached values?
2016-07-07 17:04:30 -04:00
*
* @return array uri data
*/
public static function uri($uri, $network = "", $uid = -1, $cache = true)
{
if ($cache) {
$result = Cache::get("Probe::uri:".$network.":".$uri);
if (!is_null($result)) {
return $result;
}
2016-07-03 18:14:08 -04:00
}
2017-09-23 18:53:44 -04:00
if ($uid == -1) {
$uid = local_user();
2017-05-08 11:06:57 -04:00
}
$data = self::detect($uri, $network, $uid);
2016-07-03 16:27:16 -04:00
2017-05-08 11:06:57 -04:00
if (!isset($data["url"])) {
2016-07-03 16:27:16 -04:00
$data["url"] = $uri;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (x($data, "photo")) {
$data["baseurl"] = Network::getUrlMatch(normalise_link(defaults($data, "baseurl", "")), normalise_link($data["photo"]));
2017-05-08 11:06:57 -04:00
} else {
$data["photo"] = System::baseUrl().'/images/person-175.jpg';
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (empty($data["name"])) {
if (!empty($data["nick"])) {
2016-07-10 03:27:40 -04:00
$data["name"] = $data["nick"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!x($data, "name")) {
2016-07-10 03:27:40 -04:00
$data["name"] = $data["url"];
2017-05-08 11:06:57 -04:00
}
2016-07-10 03:27:40 -04:00
}
if (empty($data["nick"])) {
2016-07-03 16:27:16 -04:00
$data["nick"] = strtolower($data["name"]);
2017-05-08 11:06:57 -04:00
if (strpos($data['nick'], ' ')) {
$data['nick'] = trim(substr($data['nick'], 0, strpos($data['nick'], ' ')));
2017-05-08 11:06:57 -04:00
}
}
if (!empty(self::$baseurl)) {
$data["baseurl"] = self::$baseurl;
}
if (empty($data["network"])) {
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_PHANTOM;
}
2016-07-03 16:27:16 -04:00
$data = self::rearrangeData($data);
2016-07-03 16:27:16 -04:00
2016-07-03 18:14:08 -04:00
// Only store into the cache if the value seems to be valid
if (!in_array($data['network'], [NETWORK_PHANTOM, NETWORK_MAIL])) {
Cache::set("Probe::uri:".$network.":".$uri, $data, CACHE_DAY);
2016-07-03 18:14:08 -04:00
/// @todo temporary fix - we need a real contact update function that updates only changing fields
/// The biggest problem is the avatar picture that could have a reduced image size.
/// It should only be updated if the existing picture isn't existing anymore.
2017-05-08 16:01:18 -04:00
/// We only update the contact when it is no probing for a specific network.
2017-05-08 11:36:04 -04:00
if (($data['network'] != NETWORK_FEED)
&& ($network == "")
&& $data["name"]
&& $data["nick"]
&& $data["url"]
&& $data["addr"]
&& $data["poll"]
2017-05-08 11:36:04 -04:00
) {
$fields = ['name' => $data['name'],
2017-08-30 14:05:41 -04:00
'nick' => $data['nick'],
'url' => $data['url'],
'addr' => $data['addr'],
'photo' => $data['photo'],
'keywords' => $data['keywords'],
'location' => $data['location'],
'about' => $data['about'],
'notify' => $data['notify'],
'network' => $data['network'],
'server_url' => $data['baseurl']];
2017-08-30 15:33:16 -04:00
2018-05-11 04:04:01 -04:00
// This doesn't cover the case when a community isn't a community anymore
if (!empty($data['community']) && $data['community']) {
$fields['community'] = $data['community'];
$fields['contact-type'] = Contact::ACCOUNT_TYPE_COMMUNITY;
2018-05-11 04:04:01 -04:00
}
$fieldnames = [];
2017-08-30 14:05:41 -04:00
foreach ($fields as $key => $val) {
2017-08-30 14:05:41 -04:00
if (empty($val)) {
unset($fields[$key]);
2017-08-30 15:33:16 -04:00
} else {
$fieldnames[] = $key;
2017-08-30 14:05:41 -04:00
}
}
$fields['updated'] = DateTimeFormat::utcNow();
2017-08-30 15:33:16 -04:00
$condition = ['nurl' => normalise_link($data["url"])];
2017-08-30 15:33:16 -04:00
$old_fields = DBA::selectFirst('gcontact', $fieldnames, $condition);
2017-08-30 15:33:16 -04:00
// When the gcontact doesn't exist, the value "true" will trigger an insert.
// In difference to the public contacts we want to have every contact
// in the world in our global contacts.
if (!$old_fields) {
$old_fields = true;
// These values have to be set only on insert
$fields['photo'] = $data['photo'];
$fields['created'] = DateTimeFormat::utcNow();
}
DBA::update('gcontact', $fields, $condition, $old_fields);
2017-08-30 14:05:41 -04:00
$fields = ['name' => $data['name'],
2017-08-30 14:05:41 -04:00
'nick' => $data['nick'],
'url' => $data['url'],
'addr' => $data['addr'],
'alias' => $data['alias'],
'keywords' => $data['keywords'],
'location' => $data['location'],
'about' => $data['about'],
'batch' => $data['batch'],
'notify' => $data['notify'],
'poll' => $data['poll'],
'request' => $data['request'],
'confirm' => $data['confirm'],
'poco' => $data['poco'],
'network' => $data['network'],
'pubkey' => $data['pubkey'],
'priority' => $data['priority'],
'writable' => true,
'rel' => Contact::SHARING];
2017-08-30 14:05:41 -04:00
$fieldnames = [];
2017-08-30 15:33:16 -04:00
foreach ($fields as $key => $val) {
2017-08-30 14:05:41 -04:00
if (empty($val)) {
unset($fields[$key]);
2017-08-30 15:33:16 -04:00
} else {
$fieldnames[] = $key;
2017-08-30 14:05:41 -04:00
}
}
$condition = ['nurl' => normalise_link($data["url"]), 'self' => false, 'uid' => 0];
2017-08-30 15:33:16 -04:00
// "$old_fields" will return a "false" when the contact doesn't exist.
// This won't trigger an insert. This is intended, since we only need
// public contacts for everyone we store items from.
// We don't need to store every contact on the planet.
$old_fields = DBA::selectFirst('contact', $fieldnames, $condition);
2017-08-30 15:33:16 -04:00
$fields['name-date'] = DateTimeFormat::utcNow();
$fields['uri-date'] = DateTimeFormat::utcNow();
$fields['success_update'] = DateTimeFormat::utcNow();
DBA::update('contact', $fields, $condition, $old_fields);
2017-05-08 11:06:57 -04:00
}
2016-07-03 18:14:08 -04:00
}
2016-07-03 16:27:16 -04:00
return $data;
}
/**
* @brief Switch the scheme of an url between http and https
*
* @param string $url URL
*
* @return string switched URL
*/
private static function switchScheme($url)
{
$parts = parse_url($url);
if (!isset($parts['scheme'])) {
return $url;
}
if ($parts['scheme'] == 'http') {
$url = str_replace('http://', 'https://', $url);
} elseif ($parts['scheme'] == 'https') {
$url = str_replace('https://', 'http://', $url);
}
return $url;
}
/**
* @brief Checks if a profile url should be OStatus but only provides partial information
*
* @param array $webfinger Webfinger data
* @param string $lrdd Path template for webfinger request
* @param string $type type
*
* @return array fixed webfinger data
*/
private static function fixOStatus($webfinger, $lrdd, $type)
{
if (empty($webfinger['links']) || empty($webfinger['subject'])) {
return $webfinger;
}
$is_ostatus = false;
$has_key = false;
foreach ($webfinger['links'] as $link) {
if ($link['rel'] == NAMESPACE_OSTATUSSUB) {
$is_ostatus = true;
}
if ($link['rel'] == 'magic-public-key') {
$has_key = true;
}
}
if (!$is_ostatus || $has_key) {
return $webfinger;
}
$url = self::switchScheme($webfinger['subject']);
$path = str_replace('{uri}', urlencode($url), $lrdd);
$webfinger2 = self::webfinger($path, $type);
// Is the new webfinger detectable as OStatus?
if (self::ostatus($webfinger2, true)) {
$webfinger = $webfinger2;
}
return $webfinger;
}
2016-07-07 17:04:30 -04:00
/**
2016-07-10 12:44:48 -04:00
* @brief Fetch information (protocol endpoints and user information) about a given uri
*
* This function is only called by the "uri" function that adds caching and rearranging of data.
2016-07-07 17:04:30 -04:00
*
* @param string $uri Address that should be probed
* @param string $network Test for this specific network
* @param integer $uid User ID for the probe (only used for mails)
2016-07-07 17:04:30 -04:00
*
* @return array uri data
*/
private static function detect($uri, $network, $uid)
{
$parts = parse_url($uri);
if (!empty($parts["scheme"]) && !empty($parts["host"]) && !empty($parts["path"])) {
$host = $parts["host"];
if (!empty($parts["port"])) {
2017-05-07 09:11:11 -04:00
$host .= ':'.$parts["port"];
}
2017-03-27 17:02:56 -04:00
if ($host == 'twitter.com') {
return ["network" => NETWORK_TWITTER];
2017-03-27 17:02:56 -04:00
}
$lrdd = self::hostMeta($host);
if (is_bool($lrdd)) {
return [];
}
$path_parts = explode("/", trim($parts["path"], "/"));
while (!$lrdd && (sizeof($path_parts) > 1)) {
$host .= "/".array_shift($path_parts);
$lrdd = self::hostMeta($host);
}
2017-03-27 17:02:56 -04:00
if (!$lrdd) {
2017-05-28 01:38:12 -04:00
logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
return self::feed($uri);
2017-03-27 17:02:56 -04:00
}
$nick = array_pop($path_parts);
// Mastodon uses a "@" as prefix for usernames in their url format
$nick = ltrim($nick, '@');
$addr = $nick."@".$host;
} elseif (strstr($uri, '@')) {
2016-07-10 03:29:43 -04:00
// If the URI starts with "mailto:" then jump directly to the mail detection
if (strpos($uri, 'mailto:') !== false) {
$uri = str_replace('mailto:', '', $uri);
return self::mail($uri, $uid);
2016-07-03 16:27:16 -04:00
}
if ($network == NETWORK_MAIL) {
return self::mail($uri, $uid);
}
2016-07-03 16:27:16 -04:00
// Remove "acct:" from the URI
$uri = str_replace('acct:', '', $uri);
2017-05-08 11:06:57 -04:00
$host = substr($uri, strpos($uri, '@') + 1);
$nick = substr($uri, 0, strpos($uri, '@'));
2016-07-03 16:27:16 -04:00
2017-03-27 17:02:56 -04:00
if (strpos($uri, '@twitter.com')) {
return ["network" => NETWORK_TWITTER];
2017-03-27 17:02:56 -04:00
}
$lrdd = self::hostMeta($host);
if (is_bool($lrdd)) {
return [];
}
2017-03-27 17:02:56 -04:00
if (!$lrdd) {
2017-05-28 01:38:12 -04:00
logger('No XRD data was found for '.$uri, LOGGER_DEBUG);
return self::mail($uri, $uid);
2017-03-27 17:02:56 -04:00
}
2016-07-03 16:27:16 -04:00
$addr = $uri;
} else {
2017-05-23 16:19:39 -04:00
logger("Uri ".$uri." was not detectable", LOGGER_DEBUG);
return false;
2016-07-03 16:27:16 -04:00
}
2016-07-03 16:27:16 -04:00
$webfinger = false;
/// @todo Do we need the prefix "acct:" or "acct://"?
foreach ($lrdd as $type => $template) {
if ($webfinger) {
2016-07-03 16:27:16 -04:00
continue;
}
// At first try it with the given uri
$path = str_replace('{uri}', urlencode($uri), $template);
$webfinger = self::webfinger($path, $type);
// Fix possible problems with GNU Social probing to wrong scheme
$webfinger = self::fixOStatus($webfinger, $template, $type);
// We cannot be sure that the detected address was correct, so we don't use the values
if ($webfinger && ($uri != $addr)) {
$nick = "";
$addr = "";
}
// Try webfinger with the address (user@domain.tld)
if (!$webfinger) {
$path = str_replace('{uri}', urlencode($addr), $template);
$webfinger = self::webfinger($path, $type);
}
// Mastodon needs to have it with "acct:"
if (!$webfinger) {
$path = str_replace('{uri}', urlencode("acct:".$addr), $template);
$webfinger = self::webfinger($path, $type);
}
2016-07-03 16:27:16 -04:00
}
if (!$webfinger) {
2016-07-03 16:27:16 -04:00
return self::feed($uri);
}
2016-07-03 16:27:16 -04:00
$result = false;
2016-07-03 18:49:38 -04:00
logger("Probing ".$uri, LOGGER_DEBUG);
if (in_array($network, ["", NETWORK_DFRN])) {
2016-07-03 16:27:16 -04:00
$result = self::dfrn($webfinger);
2017-05-08 11:06:57 -04:00
}
if ((!$result && ($network == "")) || ($network == NETWORK_DIASPORA)) {
2016-07-03 16:27:16 -04:00
$result = self::diaspora($webfinger);
2017-05-08 11:06:57 -04:00
}
if ((!$result && ($network == "")) || ($network == NETWORK_OSTATUS)) {
2016-07-03 16:27:16 -04:00
$result = self::ostatus($webfinger);
2017-05-08 11:06:57 -04:00
}
if ((!$result && ($network == "")) || ($network == NETWORK_PUMPIO)) {
2018-02-13 02:10:05 -05:00
$result = self::pumpio($webfinger, $addr);
2017-05-08 11:06:57 -04:00
}
if ((!$result && ($network == "")) || ($network == NETWORK_FEED)) {
2016-07-03 16:27:16 -04:00
$result = self::feed($uri);
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
// We overwrite the detected nick with our try if the previois routines hadn't detected it.
// Additionally it is overwritten when the nickname doesn't make sense (contains spaces).
if ((empty($result["nick"]) || (strstr($result["nick"], " "))) && ($nick != "")) {
2016-07-03 16:27:16 -04:00
$result["nick"] = $nick;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (empty($result["addr"]) && ($addr != "")) {
2016-07-03 16:27:16 -04:00
$result["addr"] = $addr;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
if (empty($result["network"])) {
$result["network"] = NETWORK_PHANTOM;
}
if (empty($result["url"])) {
$result["url"] = $uri;
}
2016-07-03 18:49:38 -04:00
logger($uri." is ".$result["network"], LOGGER_DEBUG);
if (empty($result["baseurl"])) {
2016-07-03 16:27:16 -04:00
$pos = strpos($result["url"], $host);
2017-05-08 11:06:57 -04:00
if ($pos) {
2016-07-03 16:27:16 -04:00
$result["baseurl"] = substr($result["url"], 0, $pos).$host;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
return $result;
}
2016-07-07 17:04:30 -04:00
/**
2016-07-11 13:48:37 -04:00
* @brief Perform a webfinger request.
*
* For details see RFC 7033: <https://tools.ietf.org/html/rfc7033>
2016-07-07 17:04:30 -04:00
*
* @param string $url Address that should be probed
* @param string $type type
2016-07-07 17:04:30 -04:00
*
* @return array webfinger data
*/
private static function webfinger($url, $type)
{
2017-05-08 11:06:57 -04:00
$xrd_timeout = Config::get('system', 'xrd_timeout', 20);
2016-07-03 16:27:16 -04:00
$redirects = 0;
$ret = Network::curl($url, false, $redirects, ['timeout' => $xrd_timeout, 'accept_content' => $type]);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-02-01 16:35:01 -05:00
return false;
}
$data = $ret['body'];
$webfinger = json_decode($data, true);
if (is_array($webfinger)) {
2017-05-08 11:06:57 -04:00
if (!isset($webfinger["links"])) {
2017-05-23 16:19:39 -04:00
logger("No json webfinger links for ".$url, LOGGER_DEBUG);
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
return $webfinger;
}
// If it is not JSON, maybe it is XML
$xrd = XML::parseString($data, false);
if (!is_object($xrd)) {
logger("No webfinger data retrievable for ".$url, LOGGER_DEBUG);
return false;
}
$xrd_arr = XML::elementToArray($xrd);
2017-05-08 11:06:57 -04:00
if (!isset($xrd_arr["xrd"]["link"])) {
2017-05-23 16:19:39 -04:00
logger("No XML webfinger links for ".$url, LOGGER_DEBUG);
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$webfinger = [];
2016-07-03 16:27:16 -04:00
if (!empty($xrd_arr["xrd"]["subject"])) {
2016-07-03 16:27:16 -04:00
$webfinger["subject"] = $xrd_arr["xrd"]["subject"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!empty($xrd_arr["xrd"]["alias"])) {
2016-07-03 16:27:16 -04:00
$webfinger["aliases"] = $xrd_arr["xrd"]["alias"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$webfinger["links"] = [];
2016-07-03 16:27:16 -04:00
2017-05-08 11:06:57 -04:00
foreach ($xrd_arr["xrd"]["link"] as $value => $data) {
if (!empty($data["@attributes"])) {
2016-07-03 16:27:16 -04:00
$attributes = $data["@attributes"];
2017-05-08 11:06:57 -04:00
} elseif ($value == "@attributes") {
2016-07-03 16:27:16 -04:00
$attributes = $data;
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
continue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$webfinger["links"][] = $attributes;
}
return $webfinger;
}
2016-07-07 17:04:30 -04:00
/**
2016-07-11 13:48:37 -04:00
* @brief Poll the Friendica specific noscrape page.
2016-07-07 17:04:30 -04:00
*
2016-07-11 13:48:37 -04:00
* "noscrape" is a faster alternative to fetch the data from the hcard.
* This functionality was originally created for the directory.
2016-07-10 12:44:48 -04:00
*
* @param string $noscrape_url Link to the noscrape page
* @param array $data The already fetched data
2016-07-07 17:04:30 -04:00
*
* @return array noscrape data
*/
private static function pollNoscrape($noscrape_url, $data)
{
$ret = Network::curl($noscrape_url);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-02-01 16:35:01 -05:00
return false;
}
$content = $ret['body'];
if (!$content) {
2017-05-23 16:19:39 -04:00
logger("Empty body for ".$noscrape_url, LOGGER_DEBUG);
2016-07-03 18:14:08 -04:00
return false;
2017-02-01 16:35:01 -05:00
}
2016-07-03 17:11:21 -04:00
$json = json_decode($content, true);
2017-05-08 11:06:57 -04:00
if (!is_array($json)) {
2017-05-23 16:19:39 -04:00
logger("No json data for ".$noscrape_url, LOGGER_DEBUG);
2016-07-03 18:14:08 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["fn"])) {
2016-07-03 17:11:21 -04:00
$data["name"] = $json["fn"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["addr"])) {
2016-07-03 17:11:21 -04:00
$data["addr"] = $json["addr"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["nick"])) {
2016-07-03 17:11:21 -04:00
$data["nick"] = $json["nick"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
2017-10-28 07:48:29 -04:00
if (!empty($json["guid"])) {
$data["guid"] = $json["guid"];
}
if (!empty($json["comm"])) {
2016-07-03 17:11:21 -04:00
$data["community"] = $json["comm"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["tags"])) {
2016-07-03 17:11:21 -04:00
$keywords = implode(" ", $json["tags"]);
2017-05-08 11:06:57 -04:00
if ($keywords != "") {
2016-07-03 17:11:21 -04:00
$data["keywords"] = $keywords;
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
}
$location = Profile::formatLocation($json);
2017-05-08 11:06:57 -04:00
if ($location) {
2016-07-03 17:11:21 -04:00
$data["location"] = $location;
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["about"])) {
2016-07-03 17:11:21 -04:00
$data["about"] = $json["about"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["key"])) {
2016-07-03 17:11:21 -04:00
$data["pubkey"] = $json["key"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["photo"])) {
2016-07-03 17:11:21 -04:00
$data["photo"] = $json["photo"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["dfrn-request"])) {
2016-07-03 17:11:21 -04:00
$data["request"] = $json["dfrn-request"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["dfrn-confirm"])) {
2016-07-03 17:11:21 -04:00
$data["confirm"] = $json["dfrn-confirm"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["dfrn-notify"])) {
2016-07-03 17:11:21 -04:00
$data["notify"] = $json["dfrn-notify"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
if (!empty($json["dfrn-poll"])) {
2016-07-03 17:11:21 -04:00
$data["poll"] = $json["dfrn-poll"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for valid DFRN data
*
* @param array $data DFRN data
*
* @return int Number of errors
*/
public static function validDfrn($data)
{
$errors = 0;
2017-05-08 11:06:57 -04:00
if (!isset($data['key'])) {
$errors ++;
2017-05-08 11:06:57 -04:00
}
if (!isset($data['dfrn-request'])) {
$errors ++;
2017-05-08 11:06:57 -04:00
}
if (!isset($data['dfrn-confirm'])) {
$errors ++;
2017-05-08 11:06:57 -04:00
}
if (!isset($data['dfrn-notify'])) {
$errors ++;
2017-05-08 11:06:57 -04:00
}
if (!isset($data['dfrn-poll'])) {
$errors ++;
2017-05-08 11:06:57 -04:00
}
return $errors;
}
2016-07-07 17:04:30 -04:00
/**
2016-07-10 12:44:48 -04:00
* @brief Fetch data from a DFRN profile page and via "noscrape"
2016-07-07 17:04:30 -04:00
*
* @param string $profile_link Link to the profile page
2016-07-07 17:04:30 -04:00
*
* @return array profile data
*/
public static function profile($profile_link)
{
$data = [];
logger("Check profile ".$profile_link, LOGGER_DEBUG);
// Fetch data via noscrape - this is faster
$noscrape_url = str_replace(["/hcard/", "/profile/"], "/noscrape/", $profile_link);
$data = self::pollNoscrape($noscrape_url, $data);
2017-05-08 11:36:04 -04:00
if (!isset($data["notify"])
|| !isset($data["confirm"])
|| !isset($data["request"])
|| !isset($data["poll"])
|| !isset($data["name"])
|| !isset($data["photo"])
2017-05-08 11:36:04 -04:00
) {
$data = self::pollHcard($profile_link, $data, true);
2017-05-08 11:06:57 -04:00
}
$prof_data = [];
2017-05-08 12:28:30 -04:00
$prof_data["addr"] = $data["addr"];
$prof_data["nick"] = $data["nick"];
$prof_data["dfrn-request"] = $data["request"];
$prof_data["dfrn-confirm"] = $data["confirm"];
2017-05-08 12:28:30 -04:00
$prof_data["dfrn-notify"] = $data["notify"];
$prof_data["dfrn-poll"] = $data["poll"];
$prof_data["photo"] = $data["photo"];
$prof_data["fn"] = $data["name"];
$prof_data["key"] = $data["pubkey"];
logger("Result for profile ".$profile_link.": ".print_r($prof_data, true), LOGGER_DEBUG);
return $prof_data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for DFRN contact
*
* @param array $webfinger Webfinger data
*
* @return array DFRN data
*/
private static function dfrn($webfinger)
{
$hcard_url = "";
$data = [];
2017-05-08 11:06:57 -04:00
foreach ($webfinger["links"] as $link) {
if (($link["rel"] == NAMESPACE_DFRN) && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_DFRN;
} elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["poll"] = $link["href"];
} elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["url"] = $link["href"];
} elseif (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
$hcard_url = $link["href"];
} elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["poco"] = $link["href"];
} elseif (($link["rel"] == "http://webfinger.net/rel/avatar") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["photo"] = $link["href"];
} elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["baseurl"] = trim($link["href"], '/');
} elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["guid"] = $link["href"];
} elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["pubkey"] = base64_decode($link["href"]);
//if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
2017-05-08 11:06:57 -04:00
if (strstr($data["pubkey"], 'RSA ')) {
$data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
}
if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
foreach ($webfinger["aliases"] as $alias) {
2017-10-28 07:48:29 -04:00
if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
$data["alias"] = $alias;
} elseif (substr($alias, 0, 5) == 'acct:') {
$data["addr"] = substr($alias, 5);
}
2017-07-10 18:35:32 -04:00
}
}
if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == "acct:")) {
$data["addr"] = substr($webfinger["subject"], 5);
}
if (!isset($data["network"]) || ($hcard_url == "")) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
2016-07-03 17:11:21 -04:00
// Fetch data via noscrape - this is faster
$noscrape_url = str_replace("/hcard/", "/noscrape/", $hcard_url);
$data = self::pollNoscrape($noscrape_url, $data);
2016-07-03 17:11:21 -04:00
2017-05-08 11:36:04 -04:00
if (isset($data["notify"])
&& isset($data["confirm"])
&& isset($data["request"])
&& isset($data["poll"])
&& isset($data["name"])
&& isset($data["photo"])
2017-05-08 11:36:04 -04:00
) {
2016-07-03 17:11:21 -04:00
return $data;
2017-05-08 11:06:57 -04:00
}
2016-07-03 17:11:21 -04:00
$data = self::pollHcard($hcard_url, $data, true);
2016-07-03 16:27:16 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Poll the hcard page (Diaspora and Friendica specific)
*
* @param string $hcard_url Link to the hcard page
* @param array $data The already fetched data
* @param boolean $dfrn Poll DFRN specific data
2016-07-07 17:04:30 -04:00
*
* @return array hcard data
*/
private static function pollHcard($hcard_url, $data, $dfrn = false)
{
$ret = Network::curl($hcard_url);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-02-01 16:35:01 -05:00
return false;
}
$content = $ret['body'];
if (!$content) {
return false;
2017-02-01 16:35:01 -05:00
}
2017-05-11 11:53:04 -04:00
$doc = new DOMDocument();
2017-05-08 11:06:57 -04:00
if (!@$doc->loadHTML($content)) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
2017-05-11 11:53:04 -04:00
$xpath = new DomXPath($doc);
2016-07-03 16:27:16 -04:00
$vcards = $xpath->query("//div[contains(concat(' ', @class, ' '), ' vcard ')]");
2017-05-08 11:06:57 -04:00
if (!is_object($vcards)) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!isset($data["baseurl"])) {
$data["baseurl"] = "";
}
if ($vcards->length > 0) {
$vcard = $vcards->item(0);
2016-07-03 16:27:16 -04:00
// We have to discard the guid from the hcard in favour of the guid from lrdd
// Reason: Hubzilla doesn't use the value "uid" in the hcard like Diaspora does.
$search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' uid ')]", $vcard); // */
if (($search->length > 0) && empty($data["guid"])) {
$data["guid"] = $search->item(0)->nodeValue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' nickname ')]", $vcard); // */
2017-05-08 11:06:57 -04:00
if ($search->length > 0) {
$data["nick"] = $search->item(0)->nodeValue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' fn ')]", $vcard); // */
2017-05-08 11:06:57 -04:00
if ($search->length > 0) {
$data["name"] = $search->item(0)->nodeValue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' searchable ')]", $vcard); // */
2017-05-08 11:06:57 -04:00
if ($search->length > 0) {
$data["searchable"] = $search->item(0)->nodeValue;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$search = $xpath->query("//*[contains(concat(' ', @class, ' '), ' key ')]", $vcard); // */
if ($search->length > 0) {
$data["pubkey"] = $search->item(0)->nodeValue;
2017-05-08 11:06:57 -04:00
if (strstr($data["pubkey"], 'RSA ')) {
$data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
2017-05-08 11:06:57 -04:00
}
}
2016-07-03 16:27:16 -04:00
$search = $xpath->query("//*[@id='pod_location']", $vcard); // */
2017-05-08 11:06:57 -04:00
if ($search->length > 0) {
$data["baseurl"] = trim($search->item(0)->nodeValue, "/");
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
$avatar = [];
if (!empty($vcard)) {
$photos = $xpath->query("//*[contains(concat(' ', @class, ' '), ' photo ') or contains(concat(' ', @class, ' '), ' avatar ')]", $vcard); // */
foreach ($photos as $photo) {
$attr = [];
foreach ($photo->attributes as $attribute) {
$attr[$attribute->name] = trim($attribute->value);
}
2016-07-03 16:27:16 -04:00
if (isset($attr["src"]) && isset($attr["width"])) {
$avatar[$attr["width"]] = $attr["src"];
}
// We don't have a width. So we just take everything that we got.
// This is a Hubzilla workaround which doesn't send a width.
if ((sizeof($avatar) == 0) && !empty($attr["src"])) {
$avatar[] = $attr["src"];
}
}
2016-07-03 16:27:16 -04:00
}
if (sizeof($avatar)) {
ksort($avatar);
$data["photo"] = self::fixAvatar(array_pop($avatar), $data["baseurl"]);
2016-07-03 16:27:16 -04:00
}
if ($dfrn) {
// Poll DFRN specific data
$search = $xpath->query("//link[contains(concat(' ', @rel), ' dfrn-')]");
if ($search->length > 0) {
2017-05-08 11:06:57 -04:00
foreach ($search as $link) {
2016-07-03 16:27:16 -04:00
//$data["request"] = $search->item(0)->nodeValue;
$attr = [];
2017-05-08 11:06:57 -04:00
foreach ($link->attributes as $attribute) {
2016-07-03 16:27:16 -04:00
$attr[$attribute->name] = trim($attribute->value);
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$data[substr($attr["rel"], 5)] = $attr["href"];
}
}
// Older Friendica versions had used the "uid" field differently than newer versions
if (!empty($data["nick"]) && !empty($data["guid"]) && ($data["nick"] == $data["guid"])) {
2016-07-03 16:27:16 -04:00
unset($data["guid"]);
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for Diaspora contact
*
* @param array $webfinger Webfinger data
*
* @return array Diaspora data
*/
private static function diaspora($webfinger)
{
$hcard_url = "";
$data = [];
2017-05-08 11:06:57 -04:00
foreach ($webfinger["links"] as $link) {
if (($link["rel"] == "http://microformats.org/profile/hcard") && ($link["href"] != "")) {
$hcard_url = $link["href"];
} elseif (($link["rel"] == "http://joindiaspora.com/seed_location") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["baseurl"] = trim($link["href"], '/');
} elseif (($link["rel"] == "http://joindiaspora.com/guid") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["guid"] = $link["href"];
} elseif (($link["rel"] == "http://webfinger.net/rel/profile-page") && ($link["type"] == "text/html") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["url"] = $link["href"];
} elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["poll"] = $link["href"];
} elseif (($link["rel"] == NAMESPACE_POCO) && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["poco"] = $link["href"];
} elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["notify"] = $link["href"];
} elseif (($link["rel"] == "diaspora-public-key") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["pubkey"] = base64_decode($link["href"]);
//if (strstr($data["pubkey"], 'RSA ') || ($link["type"] == "RSA"))
2017-05-08 11:06:57 -04:00
if (strstr($data["pubkey"], 'RSA ')) {
$data["pubkey"] = Crypto::rsaToPem($data["pubkey"]);
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
}
if (!isset($data["url"]) || ($hcard_url == "")) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
2017-05-08 11:06:57 -04:00
foreach ($webfinger["aliases"] as $alias) {
if (normalise_link($alias) != normalise_link($data["url"]) && ! strstr($alias, "@")) {
2016-07-03 16:27:16 -04:00
$data["alias"] = $alias;
} elseif (substr($alias, 0, 5) == 'acct:') {
$data["addr"] = substr($alias, 5);
2017-05-08 11:06:57 -04:00
}
}
}
2016-07-03 16:27:16 -04:00
2017-10-28 07:48:29 -04:00
if (!empty($webfinger["subject"]) && (substr($webfinger["subject"], 0, 5) == 'acct:')) {
$data["addr"] = substr($webfinger["subject"], 5);
}
2016-07-03 16:27:16 -04:00
// Fetch further information from the hcard
$data = self::pollHcard($hcard_url, $data);
2016-07-03 16:27:16 -04:00
2017-05-08 11:06:57 -04:00
if (!$data) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
2017-05-08 11:36:04 -04:00
if (isset($data["url"])
&& isset($data["guid"])
&& isset($data["baseurl"])
&& isset($data["pubkey"])
&& ($hcard_url != "")
2017-05-08 11:36:04 -04:00
) {
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_DIASPORA;
// The Diaspora handle must always be lowercase
if (!empty($data["addr"])) {
$data["addr"] = strtolower($data["addr"]);
}
2016-07-03 16:27:16 -04:00
// We have to overwrite the detected value for "notify" since Hubzilla doesn't send it
2017-05-08 11:06:57 -04:00
$data["notify"] = $data["baseurl"] . "/receive/users/" . $data["guid"];
2017-05-08 12:28:30 -04:00
$data["batch"] = $data["baseurl"] . "/receive/public";
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for OStatus contact
*
* @param array $webfinger Webfinger data
* @param bool $short Short detection mode
2016-07-07 17:04:30 -04:00
*
* @return array|bool OStatus data or "false" on error or "true" on short mode
2016-07-07 17:04:30 -04:00
*/
private static function ostatus($webfinger, $short = false)
{
$data = [];
2017-10-28 07:48:29 -04:00
if (!empty($webfinger["aliases"]) && is_array($webfinger["aliases"])) {
2017-05-08 11:06:57 -04:00
foreach ($webfinger["aliases"] as $alias) {
2017-10-28 07:48:29 -04:00
if (strstr($alias, "@") && !strstr(normalise_link($alias), "http://")) {
$data["addr"] = str_replace('acct:', '', $alias);
}
}
}
if (!empty($webfinger["subject"]) && strstr($webfinger["subject"], "@")
&& !strstr(normalise_link($webfinger["subject"]), "http://")
) {
$data["addr"] = str_replace('acct:', '', $webfinger["subject"]);
}
2017-10-28 07:48:29 -04:00
$pubkey = "";
2017-08-29 00:51:02 -04:00
if (is_array($webfinger["links"])) {
foreach ($webfinger["links"] as $link) {
if (($link["rel"] == "http://webfinger.net/rel/profile-page")
&& ($link["type"] == "text/html")
&& ($link["href"] != "")
) {
$data["url"] = $link["href"];
} elseif (($link["rel"] == "salmon") && ($link["href"] != "")) {
$data["notify"] = $link["href"];
} elseif (($link["rel"] == NAMESPACE_FEED) && ($link["href"] != "")) {
$data["poll"] = $link["href"];
} elseif (($link["rel"] == "magic-public-key") && ($link["href"] != "")) {
$pubkey = $link["href"];
if (substr($pubkey, 0, 5) === 'data:') {
if (strstr($pubkey, ',')) {
$pubkey = substr($pubkey, strpos($pubkey, ',') + 1);
} else {
$pubkey = substr($pubkey, 5);
}
} elseif (normalise_link($pubkey) == 'http://') {
$ret = Network::curl($pubkey);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-08-29 00:51:02 -04:00
return false;
}
$pubkey = $ret['body'];
}
2016-07-03 16:27:16 -04:00
2017-08-29 00:51:02 -04:00
$key = explode(".", $pubkey);
2016-07-03 16:27:16 -04:00
2017-08-29 00:51:02 -04:00
if (sizeof($key) >= 3) {
$m = base64url_decode($key[1]);
$e = base64url_decode($key[2]);
$data["pubkey"] = Crypto::meToPem($m, $e);
2017-08-29 00:51:02 -04:00
}
2016-07-03 16:27:16 -04:00
}
}
}
if (isset($data["notify"]) && isset($data["pubkey"])
&& isset($data["poll"])
&& isset($data["url"])
2017-05-08 11:36:04 -04:00
) {
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_OSTATUS;
} else {
2016-07-03 16:27:16 -04:00
return false;
}
if ($short) {
return true;
}
2016-07-03 16:27:16 -04:00
// Fetch all additional data from the feed
$ret = Network::curl($data["poll"]);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
return false;
}
$feed = $ret['body'];
$dummy1 = null;
$dummy2 = null;
$dummy2 = null;
2017-12-13 02:02:52 -05:00
$feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
if (!$feed_data) {
2016-07-03 16:27:16 -04:00
return false;
}
2017-08-21 16:21:04 -04:00
if (!empty($feed_data["header"]["author-name"])) {
2016-07-03 16:27:16 -04:00
$data["name"] = $feed_data["header"]["author-name"];
}
if (!empty($feed_data["header"]["author-nick"])) {
2016-07-03 16:27:16 -04:00
$data["nick"] = $feed_data["header"]["author-nick"];
}
if (!empty($feed_data["header"]["author-avatar"])) {
$data["photo"] = self::fixAvatar($feed_data["header"]["author-avatar"], $data["url"]);
}
if (!empty($feed_data["header"]["author-id"])) {
2016-07-03 16:27:16 -04:00
$data["alias"] = $feed_data["header"]["author-id"];
}
if (!empty($feed_data["header"]["author-location"])) {
$data["location"] = $feed_data["header"]["author-location"];
}
if (!empty($feed_data["header"]["author-about"])) {
$data["about"] = $feed_data["header"]["author-about"];
}
2016-07-03 16:27:16 -04:00
// OStatus has serious issues when the the url doesn't fit (ssl vs. non ssl)
// So we take the value that we just fetched, although the other one worked as well
if (!empty($feed_data["header"]["author-link"])) {
2016-07-03 16:27:16 -04:00
$data["url"] = $feed_data["header"]["author-link"];
}
2017-08-21 16:21:04 -04:00
if (($data['poll'] == $data['url']) && ($data["alias"] != '')) {
$data['url'] = $data["alias"];
$data["alias"] = '';
}
2016-07-03 16:27:16 -04:00
/// @todo Fetch location and "about" from the feed as well
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Fetch data from a pump.io profile page
*
* @param string $profile_link Link to the profile page
2016-07-07 17:04:30 -04:00
*
* @return array profile data
*/
private static function pumpioProfileData($profile_link)
{
2017-05-11 11:53:04 -04:00
$doc = new DOMDocument();
if (!@$doc->loadHTMLFile($profile_link)) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
2017-05-11 11:53:04 -04:00
$xpath = new DomXPath($doc);
2016-07-03 16:27:16 -04:00
$data = [];
2016-07-03 16:27:16 -04:00
2018-02-13 00:54:09 -05:00
$data["name"] = $xpath->query("//span[contains(@class, 'p-name')]")->item(0)->nodeValue;
if ($data["name"] == '') {
// This is ugly - but pump.io doesn't seem to know a better way for it
$data["name"] = trim($xpath->query("//h1[@class='media-header']")->item(0)->nodeValue);
$pos = strpos($data["name"], chr(10));
if ($pos) {
$data["name"] = trim(substr($data["name"], 0, $pos));
}
}
$data["location"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'p-locality')]");
2018-02-13 00:54:09 -05:00
if ($data["location"] == '') {
$data["location"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'location')]");
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$data["about"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'p-note')]");
2018-02-13 00:54:09 -05:00
if ($data["about"] == '') {
$data["about"] = XML::getFirstNodeValue($xpath, "//p[contains(@class, 'summary')]");
2018-02-13 00:54:09 -05:00
}
$avatar = $xpath->query("//img[contains(@class, 'u-photo')]")->item(0);
if (!$avatar) {
$avatar = $xpath->query("//img[@class='img-rounded media-object']")->item(0);
}
2017-05-08 11:06:57 -04:00
if ($avatar) {
foreach ($avatar->attributes as $attribute) {
if ($attribute->name == "src") {
2016-07-03 16:27:16 -04:00
$data["photo"] = trim($attribute->value);
2017-05-08 11:06:57 -04:00
}
}
}
2016-07-03 16:27:16 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for pump.io contact
*
* @param array $webfinger Webfinger data
*
* @return array pump.io data
*/
2018-02-13 02:10:05 -05:00
private static function pumpio($webfinger, $addr)
{
$data = [];
2017-05-08 11:06:57 -04:00
foreach ($webfinger["links"] as $link) {
2017-05-08 11:36:04 -04:00
if (($link["rel"] == "http://webfinger.net/rel/profile-page")
&& ($link["type"] == "text/html")
&& ($link["href"] != "")
2017-05-08 11:36:04 -04:00
) {
2016-07-03 16:27:16 -04:00
$data["url"] = $link["href"];
} elseif (($link["rel"] == "activity-inbox") && ($link["href"] != "")) {
$data["notify"] = $link["href"];
} elseif (($link["rel"] == "activity-outbox") && ($link["href"] != "")) {
$data["poll"] = $link["href"];
} elseif (($link["rel"] == "dialback") && ($link["href"] != "")) {
2016-07-03 16:27:16 -04:00
$data["dialback"] = $link["href"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
}
if (isset($data["poll"]) && isset($data["notify"])
&& isset($data["dialback"])
&& isset($data["url"])
2017-05-08 11:36:04 -04:00
) {
2016-07-03 16:27:16 -04:00
// by now we use these fields only for the network type detection
// So we unset all data that isn't used at the moment
unset($data["dialback"]);
$data["network"] = NETWORK_PUMPIO;
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$profile_data = self::pumpioProfileData($data["url"]);
2016-07-03 16:27:16 -04:00
2017-05-08 11:06:57 -04:00
if (!$profile_data) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$data = array_merge($data, $profile_data);
2018-02-13 02:10:05 -05:00
if (($addr != '') && ($data['name'] != '')) {
$name = trim(str_replace($addr, '', $data['name']));
if ($name != '') {
$data['name'] = $name;
}
}
2016-07-03 16:27:16 -04:00
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check page for feed link
*
* @param string $url Page link
*
* @return string feed link
*/
private static function getFeedLink($url)
{
2017-05-11 11:53:04 -04:00
$doc = new DOMDocument();
2017-05-08 11:06:57 -04:00
if (!@$doc->loadHTMLFile($url)) {
return false;
2017-05-08 11:06:57 -04:00
}
2017-05-11 11:53:04 -04:00
$xpath = new DomXPath($doc);
//$feeds = $xpath->query("/html/head/link[@type='application/rss+xml']");
$feeds = $xpath->query("/html/head/link[@type='application/rss+xml' and @rel='alternate']");
2017-05-08 11:06:57 -04:00
if (!is_object($feeds)) {
return false;
2017-05-08 11:06:57 -04:00
}
2017-05-08 11:06:57 -04:00
if ($feeds->length == 0) {
return false;
2017-05-08 11:06:57 -04:00
}
$feed_url = "";
2017-05-08 11:06:57 -04:00
foreach ($feeds as $feed) {
$attr = [];
2017-05-08 11:06:57 -04:00
foreach ($feed->attributes as $attribute) {
$attr[$attribute->name] = trim($attribute->value);
}
2017-05-08 11:06:57 -04:00
if ($feed_url == "") {
$feed_url = $attr["href"];
2017-05-08 11:06:57 -04:00
}
}
return $feed_url;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for feed contact
*
* @param string $url Profile link
2016-07-07 17:04:30 -04:00
* @param boolean $probe Do a probe if the page contains a feed link
*
* @return array feed data
*/
private static function feed($url, $probe = true)
{
$ret = Network::curl($url);
2018-07-20 14:07:54 -04:00
if (!empty($ret["errno"]) && ($ret['errno'] == CURLE_OPERATION_TIMEDOUT)) {
2017-02-01 16:35:01 -05:00
return false;
}
$feed = $ret['body'];
$dummy1 = $dummy2 = $dummy3 = null;
2017-12-13 02:02:52 -05:00
$feed_data = Feed::import($feed, $dummy1, $dummy2, $dummy3, true);
2016-07-03 16:27:16 -04:00
if (!$feed_data) {
2017-05-08 11:06:57 -04:00
if (!$probe) {
return false;
2017-05-08 11:06:57 -04:00
}
$feed_url = self::getFeedLink($url);
2017-05-08 11:06:57 -04:00
if (!$feed_url) {
return false;
2017-05-08 11:06:57 -04:00
}
return self::feed($feed_url, false);
}
2016-07-03 16:27:16 -04:00
if (!empty($feed_data["header"]["author-name"])) {
2016-07-03 16:27:16 -04:00
$data["name"] = $feed_data["header"]["author-name"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!empty($feed_data["header"]["author-nick"])) {
2016-07-03 16:27:16 -04:00
$data["nick"] = $feed_data["header"]["author-nick"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!empty($feed_data["header"]["author-avatar"])) {
2016-07-03 16:27:16 -04:00
$data["photo"] = $feed_data["header"]["author-avatar"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if (!empty($feed_data["header"]["author-id"])) {
2016-07-03 16:27:16 -04:00
$data["alias"] = $feed_data["header"]["author-id"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$data["url"] = $url;
$data["poll"] = $url;
if (!empty($feed_data["header"]["author-link"])) {
2016-07-03 16:27:16 -04:00
$data["baseurl"] = $feed_data["header"]["author-link"];
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
$data["baseurl"] = $data["url"];
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_FEED;
return $data;
}
2016-07-07 17:04:30 -04:00
/**
* @brief Check for mail contact
*
* @param string $uri Profile link
2016-07-07 17:04:30 -04:00
* @param integer $uid User ID
*
* @return array mail data
*/
private static function mail($uri, $uid)
{
if (!Network::isEmailDomainValid($uri)) {
2016-07-03 16:27:16 -04:00
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
if ($uid == 0) {
return false;
}
$x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1", intval($uid));
2016-07-03 16:27:16 -04:00
$r = q("SELECT * FROM `mailacct` WHERE `uid` = %d AND `server` != '' LIMIT 1", intval($uid));
2016-07-03 16:27:16 -04:00
if (!DBA::isResult($x) || !DBA::isResult($r)) {
return false;
}
$mailbox = Email::constructMailboxName($r[0]);
$password = '';
openssl_private_decrypt(hex2bin($r[0]['pass']), $password, $x[0]['prvkey']);
$mbox = Email::connect($mailbox, $r[0]['user'], $password);
if (!$mbox) {
return false;
2017-05-08 11:06:57 -04:00
}
2016-07-03 16:27:16 -04:00
$msgs = Email::poll($mbox, $uri);
logger('searching '.$uri.', '.count($msgs).' messages found.', LOGGER_DEBUG);
if (!count($msgs)) {
return false;
}
2017-05-08 12:28:30 -04:00
$phost = substr($uri, strpos($uri, '@') + 1);
2016-07-03 16:27:16 -04:00
$data = [];
2017-05-08 12:28:30 -04:00
$data["addr"] = $uri;
2016-07-03 16:27:16 -04:00
$data["network"] = NETWORK_MAIL;
2017-05-08 12:28:30 -04:00
$data["name"] = substr($uri, 0, strpos($uri, '@'));
$data["nick"] = $data["name"];
$data["photo"] = Network::lookupAvatarByEmail($uri);
2017-09-23 18:53:44 -04:00
$data["url"] = 'mailto:'.$uri;
2017-05-08 12:28:30 -04:00
$data["notify"] = 'smtp '.random_string();
$data["poll"] = 'email '.random_string();
2016-07-03 16:27:16 -04:00
$x = Email::messageMeta($mbox, $msgs[0]);
2017-05-08 11:06:57 -04:00
if (stristr($x[0]->from, $uri)) {
2016-07-03 16:27:16 -04:00
$adr = imap_rfc822_parse_adrlist($x[0]->from, '');
2017-05-08 11:06:57 -04:00
} elseif (stristr($x[0]->to, $uri)) {
2016-07-03 16:27:16 -04:00
$adr = imap_rfc822_parse_adrlist($x[0]->to, '');
2017-05-08 11:06:57 -04:00
}
if (isset($adr)) {
foreach ($adr as $feadr) {
if ((strcasecmp($feadr->mailbox, $data["name"]) == 0)
2016-07-03 16:27:16 -04:00
&&(strcasecmp($feadr->host, $phost) == 0)
2017-05-08 11:36:04 -04:00
&& (strlen($feadr->personal))
) {
2016-07-03 16:27:16 -04:00
$personal = imap_mime_header_decode($feadr->personal);
$data["name"] = "";
2017-05-08 11:06:57 -04:00
foreach ($personal as $perspart) {
if ($perspart->charset != "default") {
2016-07-03 16:27:16 -04:00
$data["name"] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
2017-05-08 11:06:57 -04:00
} else {
2016-07-03 16:27:16 -04:00
$data["name"] .= $perspart->text;
2017-05-08 11:06:57 -04:00
}
}
2016-07-03 16:27:16 -04:00
$data["name"] = notags($data["name"]);
}
}
}
2017-09-23 18:53:44 -04:00
if (!empty($mbox)) {
imap_close($mbox);
}
2016-07-03 16:27:16 -04:00
return $data;
}
/**
* @brief Mix two paths together to possibly fix missing parts
*
* @param string $avatar Path to the avatar
* @param string $base Another path that is hopefully complete
*
* @return string fixed avatar path
*/
public static function fixAvatar($avatar, $base)
{
$base_parts = parse_url($base);
// Remove all parts that could create a problem
unset($base_parts['path']);
unset($base_parts['query']);
unset($base_parts['fragment']);
$avatar_parts = parse_url($avatar);
// Now we mix them
$parts = array_merge($base_parts, $avatar_parts);
// And put them together again
2017-05-08 11:36:04 -04:00
$scheme = isset($parts['scheme']) ? $parts['scheme'] . '://' : '';
$host = isset($parts['host']) ? $parts['host'] : '';
$port = isset($parts['port']) ? ':' . $parts['port'] : '';
$path = isset($parts['path']) ? $parts['path'] : '';
$query = isset($parts['query']) ? '?' . $parts['query'] : '';
$fragment = isset($parts['fragment']) ? '#' . $parts['fragment'] : '';
$fixed = $scheme.$host.$port.$path.$query.$fragment;
logger('Base: '.$base.' - Avatar: '.$avatar.' - Fixed: '.$fixed, LOGGER_DATA);
return $fixed;
}
2016-07-03 16:27:16 -04:00
}