本文整理汇总了Golang中github.com/ipfs/go-ipfs/Godeps/_workspace/src/github.com/jbenet/go-peerstream.Conn类的典型用法代码示例。如果您正苦于以下问题:Golang Conn类的具体用法?Golang Conn怎么用?Golang Conn使用的例子?那么恭喜您, 这里精选的类代码示例或许可以为您提供帮助。
在下文中一共展示了Conn类的3个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的Golang代码示例。
示例1: newConnSetup
// newConnSetup does the swarm's "setup" for a connection. returns the underlying
// conn.Conn this method is used by both swarm.Dial and ps.Swarm connHandler
func (s *Swarm) newConnSetup(ctx context.Context, psConn *ps.Conn) (*Conn, error) {
// wrap with a Conn
sc, err := wrapConn(psConn)
if err != nil {
return nil, err
}
// if we have a public key, make sure we add it to our peerstore!
// This is an important detail. Otherwise we must fetch the public
// key from the DHT or some other system.
if pk := sc.RemotePublicKey(); pk != nil {
s.peers.AddPubKey(sc.RemotePeer(), pk)
}
// ok great! we can use it. add it to our group.
// set the RemotePeer as a group on the conn. this lets us group
// connections in the StreamSwarm by peer, and get a streams from
// any available connection in the group (better multiconn):
// swarm.StreamSwarm().NewStreamWithGroup(remotePeer)
psConn.AddGroup(sc.RemotePeer())
return sc, nil
}
示例2: wrapConn
func wrapConn(psc *ps.Conn) (*Conn, error) {
// grab the underlying connection.
if _, ok := psc.NetConn().(conn.Conn); !ok {
// this should never happen. if we see it ocurring it means that we added
// a Listener to the ps.Swarm that is NOT one of our net/conn.Listener.
return nil, fmt.Errorf("swarm connHandler: invalid conn (not a conn.Conn): %s", psc)
}
return (*Conn)(psc), nil
}
示例3: connHandler
// connHandler is called by the StreamSwarm whenever a new connection is added
// here we configure it slightly. Note that this is sequential, so if anything
// will take a while do it in a goroutine.
// See https://godoc.org/github.com/jbenet/go-peerstream for more information
func (s *Swarm) connHandler(c *ps.Conn) *Conn {
ctx := context.Background()
// this context is for running the handshake, which -- when receiveing connections
// -- we have no bound on beyond what the transport protocol bounds it at.
// note that setup + the handshake are bounded by underlying io.
// (i.e. if TCP or UDP disconnects (or the swarm closes), we're done.
// Q: why not have a shorter handshake? think about an HTTP server on really slow conns.
// as long as the conn is live (TCP says its online), it tries its best. we follow suit.)
sc, err := s.newConnSetup(ctx, c)
if err != nil {
log.Debug(err)
log.Event(ctx, "newConnHandlerDisconnect", lgbl.NetConn(c.NetConn()), lgbl.Error(err))
c.Close() // boom. close it.
return nil
}
return sc
}