some work on network discovery

2 files changed, 167 insertions, 168 deletions

diff --git a/include/diaspora.php b/include/diaspora.php
index 02d479d95..dda552536 100755
--- a/include/diaspora.php
+++ b/include/diaspora.php
@@ -140,85 +140,30 @@ function find_diaspora_person_by_handle($handle) {
 	$endlessloop = 0;
 	$maxloops = 10;
 
-	do {
-		$r = q("select * from xchan where xchan_addr = '%s' limit 1",
-			dbesc($handle)
-		);
-		if($r) {
-			$person = $r[0];
-			logger('find_diaspora_person_by handle: in cache ' . print_r($r,true), LOGGER_DEBUG);
+
+	$r = q("select * from xchan where xchan_addr = '%s' limit 1",
+		dbesc($handle)
+	);
+	if($r) {
+		$person = $r[0];
+		logger('find_diaspora_person_by handle: in cache ' . print_r($r,true), LOGGER_DATA);
 
 			// update record occasionally so it doesn't get stale
-			$d = strtotime($person['updated'] . ' +00:00');
-			if($d < strtotime('now - 14 days'))
-				$update = true;
-		}
+//			$d = strtotime($person['updated'] . ' +00:00');
+//			if($d < strtotime('now - 14 days'))
+//				$update = true;
+	}
 
+	if((! $person) || ($update))  {
 
-		// FETCHING PERSON INFORMATION FROM REMOTE SERVER
-		//
-		// If the person isn't in our 'fcontact' table, or if he/she is but
-		// his/her information hasn't been updated for more than 14 days, then
-		// we want to fetch the person's information from the remote server.
-		//
-		// Note that $person isn't changed by this block of code unless the
-		// person's information has been successfully fetched from the remote
-		// server. So if $person was 'false' to begin with (because he/she wasn't
-		// in the local cache), it'll stay false, and if $person held the local
-		// cache information to begin with, it'll keep that information. That way
-		// if there's a problem with the remote fetch, we can at least use our
-		// cached information--it's better than nothing.
-
-//fixme!!!
-
-		if((! $person) || ($update))  {
-			// Lock the function to prevent race conditions if multiple items
-			// come in at the same time from a person who doesn't exist in
-			// fcontact
-			//
-			// Don't loop forever. On the last loop, try to create the contact
-			// whether the function is locked or not. Maybe the locking thread
-			// has died or something. At any rate, a duplicate in 'fcontact'
-			// is a much smaller problem than a deadlocked thread
-//			$got_lock = lock_function('find_diaspora_person_by_handle', false);
-			if(($endlessloop + 1) >= $maxloops)
-				$got_lock = true;
-
-			if($got_lock) {
-				logger('find_diaspora_person_by_handle: create or refresh', LOGGER_DEBUG);
-				require_once('include/Scrape.php');
-				$r = probe_url($handle, PROBE_DIASPORA);
-
-				// Note that Friendica contacts can return a "Diaspora person"
-				// if Diaspora connectivity is enabled on their server
-				if((count($r)) && ($r['network'] === NETWORK_DIASPORA)) {
-					add_fcontact($r,$update);
-					$person = ($r);
-				}
+		// try webfinger. Make sure to distinguish between diaspora, 
+		// redmatrix w/diaspora protocol and friendica w/diaspora protocol.
 
-//				unlock_function('find_diaspora_person_by_handle');
-			}
-			else {
-				logger('find_diaspora_person_by_handle: couldn\'t lock function', LOGGER_DEBUG);
-//				if(! $person)
-//					block_on_function_lock('find_diaspora_person_by_handle');
-			}
-		}
-	} while((! $person) && (! $got_lock) && (++$endlessloop < $maxloops));
+		$result = discover_by_webbie($handle);
+
+
+	}
 
-	// We need to try again if the person wasn't in 'fcontact' but the function was locked.
-	// The fact that the function was locked may mean that another process was creating the
-	// person's record. It could also mean another process was creating or updating an unrelated
-	// person.
-	//
-	// At any rate, we need to keep trying until we've either got the person or had a chance to
-	// try to fetch his/her remote information. But we don't want to block on locking the
-	// function, because if the other process is creating the record, then when we acquire the lock
-	// we'll dive right into creating another, duplicate record. We DO want to at least wait
-	// until the lock is released, so we don't flood the database with requests.
-	//
-	// If the person was in the 'fcontact' table, don't try again. It's not worth the time, since
-	// we do have some information for the person
 
 	return $person;
 }
diff --git a/include/network.php b/include/network.php
index 66bba5b38..1a974a681 100644
--- a/include/network.php
+++ b/include/network.php
@@ -337,101 +337,6 @@ function convert_xml_element_to_array($xml_element, &$recursion_depth=0) {
         }
 }
 
-// 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 $s.
-// Return an empty string if email-style addresses but webfinger fails,
-// or if the resultant personal XRD doesn't contain a supported 
-// subscription/friend-request attribute.
-
-// amended 7/9/2011 to return an hcard which could save potentially loading 
-// a lengthy content page to scrape dfrn attributes
-
-
-function webfinger_dfrn($s,&$hcard) {
-	if(! strstr($s,'@')) {
-		return $s;
-	}
-	$profile_link = '';
-
-	$links = webfinger($s);
-	logger('webfinger_dfrn: ' . $s . ':' . print_r($links,true), LOGGER_DATA);
-	if(count($links)) {
-		foreach($links as $link) {
-			if($link['@attributes']['rel'] === NAMESPACE_DFRN)
-				$profile_link = $link['@attributes']['href'];
-			if($link['@attributes']['rel'] === NAMESPACE_OSTATUSSUB)
-				$profile_link = 'stat:' . $link['@attributes']['template'];	
-			if($link['@attributes']['rel'] === 'http://microformats.org/profile/hcard')
-				$hcard = $link['@attributes']['href'];				
-		}
-	}
-	return $profile_link;
-}
-
-// Given an email style address, perform webfinger lookup and 
-// return the array of link attributes from the personal XRD file.
-// On error/failure return an empty array.
-
-
-
-function webfinger($s, $debug = false) {
-	$host = '';
-	if(strstr($s,'@')) {
-		$host = substr($s,strpos($s,'@') + 1);
-	}
-	if(strlen($host)) {
-		$tpl = fetch_lrdd_template($host);
-		logger('webfinger: lrdd template: ' . $tpl);
-		if(strlen($tpl)) {
-			$pxrd = str_replace('{uri}', urlencode('acct:' . $s), $tpl);
-			logger('webfinger: pxrd: ' . $pxrd);
-			$links = fetch_xrd_links($pxrd);
-			if(! count($links)) {
-				// try with double slashes
-				$pxrd = str_replace('{uri}', urlencode('acct://' . $s), $tpl);
-				logger('webfinger: pxrd: ' . $pxrd);
-				$links = fetch_xrd_links($pxrd);
-			}
-			return $links;
-		}
-	}
-	return array();
-}
-
-
-
-
-// Given a host name, locate the LRDD template from that
-// host. Returns the LRDD template or an empty string on
-// error/failure.
-
-
-function fetch_lrdd_template($host) {
-	$tpl = '';
-
-	$url1 = 'https://' . $host . '/.well-known/host-meta' ;
-	$url2 = 'http://' . $host . '/.well-known/host-meta' ;
-	$links = fetch_xrd_links($url1);
-	logger('fetch_lrdd_template from: ' . $url1);
-	logger('template (https): ' . print_r($links,true));
-	if(! count($links)) {
-		logger('fetch_lrdd_template from: ' . $url2);
-		$links = fetch_xrd_links($url2);
-		logger('template (http): ' . print_r($links,true));
-	}
-	if(count($links)) {
-		foreach($links as $link)
-			if($link['@attributes']['rel'] && $link['@attributes']['rel'] === 'lrdd')
-				$tpl = $link['@attributes']['template'];
-	}
-	if(! strpos($tpl,'{uri}'))
-		$tpl = '';
-	return $tpl;
-}
-
 // Take a URL from the wild, prepend http:// if necessary
 // and check DNS to see if it's real (or check if is a valid IP address)
 // return true if it's OK, false if something is wrong with it
@@ -912,3 +817,152 @@ function email_send($addr, $subject, $headers, $item) {
 	logger('notifier: email delivery to ' . $addr);
 	mail($addr, $subject, $body, $headers);
 }
+
+
+function discover_by_webbie($webbie) {
+
+	$x = webfinger_rfc7033($webbie);
+	if($x && array_key_exists('links',$x) && $x['links']) {
+		foreach($x['links'] as $link) {
+			if(array_key_exists('rel',$link) && $link['rel'] == 'http://purl.org/zot/protocol') {
+				logger('discover_by_webbie: zot found for ' . $webbie);
+				$z = z_fetch_url($link['href']);
+				if($z['success']) {
+					$j = json_decode($z['body'],true);
+					$i = import_xchan($j);
+					return true;
+				}
+			}
+		}
+	}
+
+	$x = old_webfinger($webbie);			
+	if($x) {
+		logger('old_webfinger: ' . print_r($x,true));
+	}
+}
+
+
+function webfinger_rfc7033($webbie) {
+
+
+	if(! strpos($webbie,'@'))
+		return false;
+	$lhs = substr($webbie,0,strpos($webbie,'@'));
+	$rhs = substr($webbie,strpos($webbie,'@')+1);
+
+	$resource = 'acct:' . $webbie;
+
+	$s = z_fetch_url('https://' . $rhs . '/.well-known/webfinger?resource=' . $resource);
+
+	if($s['success'])
+		$j = json_decode($s['body'],true);
+	else
+		return false;
+	return($j);
+}
+
+
+function old_webfinger($webbie) {
+
+	$host = '';
+	if(strstr($webbie,'@'))
+		$host = substr($webbie,strpos($webbie,'@') + 1);
+
+	if(strlen($host)) {
+		$tpl = fetch_lrdd_template($host);
+		logger('old_webfinger: lrdd template: ' . $tpl,LOGGER_DATA);
+		if(strlen($tpl)) {
+			$pxrd = str_replace('{uri}', urlencode('acct:' . $webbie), $tpl);
+			logger('old_webfinger: pxrd: ' . $pxrd,LOGGER_DATA);
+			$links = fetch_xrd_links($pxrd);
+			if(! count($links)) {
+				// try with double slashes
+				$pxrd = str_replace('{uri}', urlencode('acct://' . $webbie), $tpl);
+				logger('old_webfinger: pxrd: ' . $pxrd,LOGGER_DATA);
+				$links = fetch_xrd_links($pxrd);
+			}
+			return $links;
+    	}
+	}
+    return array();
+}
+
+
+function fetch_lrdd_template($host) {
+    $tpl = '';
+
+    $url1 = 'https://' . $host . '/.well-known/host-meta' ;
+    $url2 = 'http://' . $host . '/.well-known/host-meta' ;
+    $links = fetch_xrd_links($url1);
+    logger('fetch_lrdd_template from: ' . $url1, LOGGER_DEBUG);
+    logger('template (https): ' . print_r($links,true),LOGGER_DEBUG);
+    if(! count($links)) {
+        logger('fetch_lrdd_template from: ' . $url2);
+        $links = fetch_xrd_links($url2);
+        logger('template (http): ' . print_r($links,true),LOGGER_DEBUG);
+    }
+    if(count($links)) {
+        foreach($links as $link)
+            if($link['@attributes']['rel'] && $link['@attributes']['rel'] === 'lrdd' && (!$link['@attributes']['type'] || $link['@attributes']['type'] === 'application/xrd+xml'))
+                $tpl = $link['@attributes']['template'];
+    }
+    if(! strpos($tpl,'{uri}'))
+        $tpl = '';
+    return $tpl;
+
+}
+
+
+function fetch_xrd_links($url) {
+
+logger('fetch_xrd_links: ' . $url);
+
+	$redirects = 0;
+    $x = z_fetch_url($url,false,$redirects,array('timeout' => 20));
+
+	if(! $x['success'])
+		return array();
+
+	$xml = $x['body'];
+    logger('fetch_xrd_links: ' . $xml, LOGGER_DATA);
+
+    if ((! $xml) || (! stristr($xml,'<xrd')))
+        return array();
+
+    // fix diaspora's bad xml
+    $xml = str_replace(array('href=&quot;','&quot;/>'),array('href="','"/>'),$xml);
+
+    $h = parse_xml_string($xml);
+    if(! $h)
+        return array();
+
+    $arr = convert_xml_element_to_array($h);
+
+    $links = array();
+
+    if(isset($arr['xrd']['link'])) {
+        $link = $arr['xrd']['link'];
+
+        if(! isset($link[0]))
+            $links = array($link);
+        else
+            $links = $link;
+    }
+    if(isset($arr['xrd']['alias'])) {
+        $alias = $arr['xrd']['alias'];
+        if(! isset($alias[0]))
+            $aliases = array($alias);
+        else
+            $aliases = $alias;
+        if(is_array($aliases) && count($aliases)) {
+            foreach($aliases as $alias) {
+                $links[]['@attributes'] = array('rel' => 'alias' , 'href' => $alias);
+            }
+        }
+    }
+
+    logger('fetch_xrd_links: ' . print_r($links,true), LOGGER_DATA);
+
+    return $links;
+}
\ No newline at end of file