package main import ( "encoding/hex" "fmt" "io/ioutil" "log" "net" "os" "os/exec" "path/filepath" "runtime" "strconv" "sync" "time" "golang.org/x/net/context" "google.golang.org/grpc" "google.golang.org/grpc/grpclog" "bytes" "github.com/go-errors/errors" "github.com/lightningnetwork/lnd/lnrpc" "github.com/roasbeef/btcd/chaincfg" "github.com/roasbeef/btcd/chaincfg/chainhash" "github.com/roasbeef/btcd/rpctest" "github.com/roasbeef/btcd/txscript" "github.com/roasbeef/btcd/wire" "github.com/roasbeef/btcrpcclient" "github.com/roasbeef/btcutil" ) var ( // numActiveNodes is the number of active nodes within the test network. numActiveNodes = 0 // defaultNodePort is the initial p2p port which will be used by the // first created lightning node to listen on for incoming p2p // connections. Subsequent allocated ports for future lighting nodes // instances will be monotonically increasing odd numbers calculated as // such: defaultP2pPort + (2 * harness.nodeNum). defaultNodePort = 19555 // defaultClientPort is the initial rpc port which will be used by the // first created lightning node to listen on for incoming rpc // connections. Subsequent allocated ports for future rpc harness // instances will be monotonically increasing even numbers calculated // as such: defaultP2pPort + (2 * harness.nodeNum). defaultClientPort = 19556 harnessNetParams = &chaincfg.SimNetParams ) // generateListeningPorts returns two strings representing ports to listen on // designated for the current lightning network test. If there haven't been any // test instances created, the default ports are used. Otherwise, in order to // support multiple test nodes running at once, the p2p and rpc port are // incremented after each initialization. func generateListeningPorts() (int, int) { var p2p, rpc int if numActiveNodes == 0 { p2p = defaultNodePort rpc = defaultClientPort } else { p2p = defaultNodePort + (2 * numActiveNodes) rpc = defaultClientPort + (2 * numActiveNodes) } return p2p, rpc } // lightningNode represents an instance of lnd running within our test network // harness. Each lightningNode instance also fully embedds an RPC client in // order to pragmatically drive the node. type lightningNode struct { cfg *config rpcAddr string p2pAddr string rpcCert []byte nodeId int // PubKey is the serialized compressed identity public key of the node. // This field will only be populated once the node itself has been // started via the start() method. PubKey [33]byte PubKeyStr string cmd *exec.Cmd pidFile string // processExit is a channel that's closed once it's detected that the // process this instance of lightningNode is bound to has exited. processExit chan struct{} extraArgs []string lnrpc.LightningClient } // newLightningNode creates a new test lightning node instance from the passed // rpc config and slice of extra arguments. func newLightningNode(rpcConfig *btcrpcclient.ConnConfig, lndArgs []string) (*lightningNode, error) { var err error cfg := &config{ RPCHost: rpcConfig.Host, RPCUser: rpcConfig.User, RPCPass: rpcConfig.Pass, } nodeNum := numActiveNodes cfg.DataDir, err = ioutil.TempDir("", "lndtest-data") if err != nil { return nil, err } cfg.LogDir, err = ioutil.TempDir("", "lndtest-log") if err != nil { return nil, err } cfg.PeerPort, cfg.RPCPort = generateListeningPorts() numActiveNodes++ return &lightningNode{ cfg: cfg, p2pAddr: net.JoinHostPort("127.0.0.1", strconv.Itoa(cfg.PeerPort)), rpcAddr: net.JoinHostPort("127.0.0.1", strconv.Itoa(cfg.RPCPort)), rpcCert: rpcConfig.Certificates, nodeId: nodeNum, processExit: make(chan struct{}), extraArgs: lndArgs, }, nil } // genArgs generates a slice of command line arguments from the lightningNode's // current config struct. func (l *lightningNode) genArgs() []string { var args []string encodedCert := hex.EncodeToString(l.rpcCert) args = append(args, fmt.Sprintf("--btcdhost=%v", l.cfg.RPCHost)) args = append(args, fmt.Sprintf("--rpcuser=%v", l.cfg.RPCUser)) args = append(args, fmt.Sprintf("--rpcpass=%v", l.cfg.RPCPass)) args = append(args, fmt.Sprintf("--rawrpccert=%v", encodedCert)) args = append(args, fmt.Sprintf("--rpcport=%v", l.cfg.RPCPort)) args = append(args, fmt.Sprintf("--peerport=%v", l.cfg.PeerPort)) args = append(args, fmt.Sprintf("--logdir=%v", l.cfg.LogDir)) args = append(args, fmt.Sprintf("--datadir=%v", l.cfg.DataDir)) args = append(args, fmt.Sprintf("--simnet")) if l.extraArgs != nil { args = append(args, l.extraArgs...) } return args } // start launches a new process running lnd. Additionally, the PID of the // launched process is saved in order to possibly kill the process forcibly // later. func (l *lightningNode) start(lndError chan error) error { args := l.genArgs() l.cmd = exec.Command("lnd", args...) // Redirect stderr output to buffer var errb bytes.Buffer l.cmd.Stderr = &errb if err := l.cmd.Start(); err != nil { return err } // Launch a new goroutine which that bubbles up any potential fatal // process errors to the goroutine running the tests. go func() { if err := l.cmd.Wait(); err != nil { lndError <- errors.New(errb.String()) } // Signal any onlookers that this process has exited. close(l.processExit) }() pid, err := os.Create(filepath.Join(l.cfg.DataDir, fmt.Sprintf("%v.pid", l.nodeId))) if err != nil { return err } l.pidFile = pid.Name() if _, err = fmt.Fprintf(pid, "%v\n", l.cmd.Process.Pid); err != nil { return err } if err := pid.Close(); err != nil { return err } opts := []grpc.DialOption{ grpc.WithInsecure(), grpc.WithBlock(), grpc.WithTimeout(time.Second * 20), } conn, err := grpc.Dial(l.rpcAddr, opts...) if err != nil { return err } l.LightningClient = lnrpc.NewLightningClient(conn) // Obtain the lnid of this node for quick identification purposes. ctxb := context.Background() info, err := l.GetInfo(ctxb, &lnrpc.GetInfoRequest{}) if err != nil { return err } l.PubKeyStr = info.IdentityPubkey pubkey, err := hex.DecodeString(info.IdentityPubkey) if err != nil { return err } copy(l.PubKey[:], pubkey) return nil } // cleanup cleans up all the temporary files created by the node's process. func (l *lightningNode) cleanup() error { dirs := []string{ l.cfg.LogDir, l.cfg.DataDir, } var err error for _, dir := range dirs { if err = os.RemoveAll(dir); err != nil { log.Printf("Cannot remove dir %s: %v", dir, err) } } return err } // stop attempts to stop the active lnd process. func (l *lightningNode) stop() error { // We should skip node stop in case: // - start of the node wasn't initiated // - process wasn't spawned // - process already finished processFinished := l.cmd.ProcessState != nil && l.cmd.ProcessState.Exited() if l.cmd == nil || l.cmd.Process == nil || processFinished { return nil } if runtime.GOOS == "windows" { return l.cmd.Process.Signal(os.Kill) } return l.cmd.Process.Signal(os.Interrupt) } // restart attempts to restart a lightning node by shutting it down cleanly, // then restarting the process. This function is fully blocking. Upon restart, // the RPC connection to the node will be re-attempted, continuing iff the // connection attempt is successful. Additionally, if a callback is passed, the // closure will be executed after the node has been shutdown, but before the // process has been started up again. func (l *lightningNode) restart(errChan chan error, callback func() error) error { if err := l.stop(); err != nil { return nil } <-l.processExit l.processExit = make(chan struct{}) if callback != nil { if err := callback(); err != nil { return err } } return l.start(errChan) } // shutdown stops the active lnd process and clean up any temporary directories // created along the way. func (l *lightningNode) shutdown() error { if err := l.stop(); err != nil { return err } if err := l.cleanup(); err != nil { return err } return nil } // networkHarness is an integration testing harness for the lightning network. // The harness by default is created with two active nodes on the network: // Alice and Bob. type networkHarness struct { rpcConfig btcrpcclient.ConnConfig netParams *chaincfg.Params Miner *rpctest.Harness activeNodes map[int]*lightningNode // Alice and Bob are the initial seeder nodes that are automatically // created to be the initial participants of the test network. Alice *lightningNode Bob *lightningNode seenTxns chan chainhash.Hash watchRequests chan *watchRequest // Channel for transmitting stderr output from failed lightning node // to main process. lndErrorChan chan error sync.Mutex } // newNetworkHarness creates a new network test harness. // TODO(roasbeef): add option to use golang's build library to a binary of the // current repo. This'll save developers from having to manually `go install` // within the repo each time before changes func newNetworkHarness() (*networkHarness, error) { return &networkHarness{ activeNodes: make(map[int]*lightningNode), seenTxns: make(chan chainhash.Hash), watchRequests: make(chan *watchRequest), lndErrorChan: make(chan error), }, nil } // InitializeSeedNodes initialized alice and bob nodes given an already // running instance of btcd's rpctest harness and extra command line flags, // which should be formatted properly - "--arg=value". func (n *networkHarness) InitializeSeedNodes(r *rpctest.Harness, lndArgs []string) error { nodeConfig := r.RPCConfig() n.netParams = r.ActiveNet n.Miner = r n.rpcConfig = nodeConfig var err error n.Alice, err = newLightningNode(&nodeConfig, lndArgs) if err != nil { return err } n.Bob, err = newLightningNode(&nodeConfig, lndArgs) if err != nil { return err } n.activeNodes[n.Alice.nodeId] = n.Alice n.activeNodes[n.Bob.nodeId] = n.Bob return err } // ProcessErrors returns a channel used for reporting any fatal process errors. // If any of the active nodes within the harness' test network incur a fatal // error, that error is sent over this channel. func (n *networkHarness) ProcessErrors() chan error { return n.lndErrorChan } // fakeLogger is a fake grpclog.Logger implementation. This is used to stop // grpc's logger from printing directly to stdout. type fakeLogger struct{} func (f *fakeLogger) Fatal(args ...interface{}) {} func (f *fakeLogger) Fatalf(format string, args ...interface{}) {} func (f *fakeLogger) Fatalln(args ...interface{}) {} func (f *fakeLogger) Print(args ...interface{}) {} func (f *fakeLogger) Printf(format string, args ...interface{}) {} func (f *fakeLogger) Println(args ...interface{}) {} // SetUp starts the initial seeder nodes within the test harness. The initial // node's wallets will be funded wallets with ten 1 BTC outputs each. Finally // rpc clients capable of communicating with the initial seeder nodes are // created. func (n *networkHarness) SetUp() error { // Swap out grpc's default logger with out fake logger which drops the // statements on the floor. grpclog.SetLogger(&fakeLogger{}) // Start the initial seeder nodes within the test network, then connect // their respective RPC clients. var wg sync.WaitGroup errChan := make(chan error, 2) wg.Add(2) go func() { var err error defer wg.Done() if err = n.Alice.start(n.lndErrorChan); err != nil { errChan <- err return } }() go func() { var err error defer wg.Done() if err = n.Bob.start(n.lndErrorChan); err != nil { errChan <- err return } }() wg.Wait() select { case err := <-errChan: return err default: } // Load up the wallets of the seeder nodes with 10 outputs of 1 BTC // each. ctxb := context.Background() addrReq := &lnrpc.NewAddressRequest{lnrpc.NewAddressRequest_WITNESS_PUBKEY_HASH} clients := []lnrpc.LightningClient{n.Alice, n.Bob} for _, client := range clients { for i := 0; i < 10; i++ { resp, err := client.NewAddress(ctxb, addrReq) if err != nil { return err } addr, err := btcutil.DecodeAddress(resp.Address, n.netParams) if err != nil { return err } addrScript, err := txscript.PayToAddrScript(addr) if err != nil { return err } output := &wire.TxOut{ PkScript: addrScript, Value: btcutil.SatoshiPerBitcoin, } if _, err := n.Miner.SendOutputs([]*wire.TxOut{output}, 30); err != nil { return err } } } // We generate several blocks in order to give the outputs created // above a good number of confirmations. if _, err := n.Miner.Node.Generate(10); err != nil { return err } // Finally, make a connection between both of the nodes. if err := n.ConnectNodes(ctxb, n.Alice, n.Bob); err != nil { return err } // Now block until both wallets have fully synced up. expectedBalance := btcutil.Amount(btcutil.SatoshiPerBitcoin * 10).ToBTC() balReq := &lnrpc.WalletBalanceRequest{} balanceTicker := time.Tick(time.Millisecond * 50) out: for { select { case <-balanceTicker: aliceResp, err := n.Alice.WalletBalance(ctxb, balReq) if err != nil { return err } bobResp, err := n.Bob.WalletBalance(ctxb, balReq) if err != nil { return err } if aliceResp.Balance == expectedBalance && bobResp.Balance == expectedBalance { break out } case <-time.After(time.Second * 30): return fmt.Errorf("balances not synced after deadline") } } // Now that the initial test network has been initialized, launch the // network wather. go n.networkWatcher() return nil } // TearDownAll tears down all active nodes within the test lightning network. func (n *networkHarness) TearDownAll() error { for _, node := range n.activeNodes { if err := node.shutdown(); err != nil { return err } } return nil } // NewNode fully initializes a returns a new lightningNode binded to the // current instance of the network harness. The created node is running, but // not yet connected to other nodes within the network. func (n *networkHarness) NewNode(extraArgs []string) (*lightningNode, error) { n.Lock() defer n.Unlock() node, err := newLightningNode(&n.rpcConfig, extraArgs) if err != nil { return nil, err } if err := node.start(n.lndErrorChan); err != nil { return nil, err } n.activeNodes[node.nodeId] = node return node, nil } // ConnectNodes establishes an encrypted+authenticated p2p connection from node // a towards node b. func (n *networkHarness) ConnectNodes(ctx context.Context, a, b *lightningNode) error { bobInfo, err := b.GetInfo(ctx, &lnrpc.GetInfoRequest{}) if err != nil { return err } req := &lnrpc.ConnectPeerRequest{ Addr: &lnrpc.LightningAddress{ Pubkey: bobInfo.IdentityPubkey, Host: b.p2pAddr, }, } if _, err := a.ConnectPeer(ctx, req); err != nil { return err } return nil } // RestartNode attempts to restart a lightning node by shutting it down // cleanly, then restarting the process. This function is fully blocking. Upon // restart, the RPC connection to the node will be re-attempted, continuing iff // the connection attempt is successful. If the callback parameter is non-nil, // then the function will be executed after the node shuts down, but *before* // the process has been started up again. // // This method can be useful when testing edge cases such as a node broadcast // and invalidated prior state, or persistent state recovery, simulating node // crashes, etc. func (n *networkHarness) RestartNode(node *lightningNode, callback func() error) error { return node.restart(n.lndErrorChan, callback) } // TODO(roasbeef): add a WithChannel higher-order function? // * python-like context manager w.r.t using a channel within a test // * possibly adds more funds to the target wallet if the funds are not // enough // watchRequest encapsulates a request to the harness' network watcher to // dispatch a notification once a transaction with the target txid is seen // within the test network. type watchRequest struct { txid chainhash.Hash eventChan chan struct{} } // networkWatcher is a goroutine which accepts async notification requests for // the broadcast of a target transaction, and then dispatches the transaction // once its seen on the network. func (n *networkHarness) networkWatcher() { seenTxns := make(map[chainhash.Hash]struct{}) clients := make(map[chainhash.Hash][]chan struct{}) for { select { case req := <-n.watchRequests: // If we've already seen this transaction, then // immediately dispatch the request. Otherwise, append // to the list of clients who are watching for the // broadcast of this transaction. if _, ok := seenTxns[req.txid]; ok { close(req.eventChan) } else { clients[req.txid] = append(clients[req.txid], req.eventChan) } case txid := <-n.seenTxns: // Add this txid to our set of "seen" transactions. So // we're able to dispatch any notifications for this // txid which arrive *after* it's seen within the // network. seenTxns[txid] = struct{}{} // If there isn't a registered notification for this // transaction then ignore it. txClients, ok := clients[txid] if !ok { continue } // Otherwise, dispatch the notification to all clients, // cleaning up the now un-needed state. for _, client := range txClients { close(client) } delete(clients, txid) } } } // OnTxAccepted is a callback to be called each time a new transaction has been // broadcast on the network. func (n *networkHarness) OnTxAccepted(hash *chainhash.Hash, amt btcutil.Amount) { go func() { n.seenTxns <- *hash }() } // WaitForTxBroadcast blocks until the target txid is seen on the network. If // the transaction isn't seen within the network before the passed timeout, // then an error is returned. // TODO(roasbeef): add another method which creates queue of all seen transactions func (n *networkHarness) WaitForTxBroadcast(ctx context.Context, txid chainhash.Hash) error { eventChan := make(chan struct{}) n.watchRequests <- &watchRequest{txid, eventChan} select { case <-eventChan: return nil case <-ctx.Done(): return fmt.Errorf("tx not seen before context timeout") } } // OpenChannel attempts to open a channel between srcNode and destNode with the // passed channel funding parameters. If the passed context has a timeout, then // if the timeout is reached before the channel pending notification is // received, an error is returned. func (n *networkHarness) OpenChannel(ctx context.Context, srcNode, destNode *lightningNode, amt btcutil.Amount, pushAmt btcutil.Amount, numConfs uint32) (lnrpc.Lightning_OpenChannelClient, error) { openReq := &lnrpc.OpenChannelRequest{ NodePubkey: destNode.PubKey[:], LocalFundingAmount: int64(amt), PushSat: int64(pushAmt), NumConfs: numConfs, } respStream, err := srcNode.OpenChannel(ctx, openReq) if err != nil { return nil, fmt.Errorf("unable to open channel between "+ "alice and bob: %v", err) } chanOpen := make(chan struct{}) errChan := make(chan error) go func() { // Consume the "channel pending" update. This waits until the node // notifies us that the final message in the channel funding workflow // has been sent to the remote node. resp, err := respStream.Recv() if err != nil { errChan <- err return } if _, ok := resp.Update.(*lnrpc.OpenStatusUpdate_ChanPending); !ok { errChan <- fmt.Errorf("expected channel pending update, "+ "instead got %v", resp) return } close(chanOpen) }() select { case <-ctx.Done(): return nil, fmt.Errorf("timeout reached before chan pending " + "update sent") case err := <-errChan: return nil, err case <-chanOpen: return respStream, nil } } // WaitForChannelOpen waits for a notification that a channel is open by // consuming a message from the past open channel stream. If the passed context // has a timeout, then if the timeout is reached before the channel has been // opened, then an error is returned. func (n *networkHarness) WaitForChannelOpen(ctx context.Context, openChanStream lnrpc.Lightning_OpenChannelClient) (*lnrpc.ChannelPoint, error) { errChan := make(chan error) respChan := make(chan *lnrpc.ChannelPoint) go func() { resp, err := openChanStream.Recv() if err != nil { errChan <- fmt.Errorf("unable to read rpc resp: %v", err) return } fundingResp, ok := resp.Update.(*lnrpc.OpenStatusUpdate_ChanOpen) if !ok { errChan <- fmt.Errorf("expected channel open update, "+ "instead got %v", resp) return } respChan <- fundingResp.ChanOpen.ChannelPoint }() select { case <-ctx.Done(): return nil, fmt.Errorf("timeout reached while waiting for " + "channel open") case err := <-errChan: return nil, err case chanPoint := <-respChan: return chanPoint, nil } } // CloseChannel close channel attempts to close the channel indicated by the // passed channel point, initiated by the passed lnNode. If the passed context // has a timeout, then if the timeout is reached before the channel close is // pending, then an error is returned. func (n *networkHarness) CloseChannel(ctx context.Context, lnNode *lightningNode, cp *lnrpc.ChannelPoint, force bool) (lnrpc.Lightning_CloseChannelClient, *chainhash.Hash, error) { closeReq := &lnrpc.CloseChannelRequest{ ChannelPoint: cp, Force: force, } closeRespStream, err := lnNode.CloseChannel(ctx, closeReq) if err != nil { return nil, nil, fmt.Errorf("unable to close channel: %v", err) } errChan := make(chan error) fin := make(chan *chainhash.Hash) go func() { // Consume the "channel close" update in order to wait for the closing // transaction to be broadcast, then wait for the closing tx to be seen // within the network. closeResp, err := closeRespStream.Recv() if err != nil { errChan <- err return } pendingClose, ok := closeResp.Update.(*lnrpc.CloseStatusUpdate_ClosePending) if !ok { errChan <- fmt.Errorf("expected channel close update, "+ "instead got %v", pendingClose) return } closeTxid, err := chainhash.NewHash(pendingClose.ClosePending.Txid) if err != nil { errChan <- err return } if err := n.WaitForTxBroadcast(ctx, *closeTxid); err != nil { errChan <- err return } fin <- closeTxid }() // Wait until either the deadline for the context expires, an error // occurs, or the channel close update is received. select { case <-ctx.Done(): return nil, nil, fmt.Errorf("timeout reached before channel close " + "initiated") case err := <-errChan: return nil, nil, err case closeTxid := <-fin: return closeRespStream, closeTxid, nil } } // WaitForChannelClose waits for a notification from the passed channel close // stream that the node has deemed the channel has been fully closed. If the // passed context has a timeout, then if the timeout is reached before the // notification is received then an error is returned. func (n *networkHarness) WaitForChannelClose(ctx context.Context, closeChanStream lnrpc.Lightning_CloseChannelClient) (*chainhash.Hash, error) { errChan := make(chan error) updateChan := make(chan *lnrpc.CloseStatusUpdate_ChanClose) go func() { closeResp, err := closeChanStream.Recv() if err != nil { errChan <- err return } closeFin, ok := closeResp.Update.(*lnrpc.CloseStatusUpdate_ChanClose) if !ok { errChan <- fmt.Errorf("expected channel close update, "+ "instead got %v", closeFin) return } updateChan <- closeFin }() // Wait until either the deadline for the context expires, an error // occurs, or the channel close update is received. select { case <-ctx.Done(): return nil, fmt.Errorf("timeout reached before update sent") case err := <-errChan: return nil, err case update := <-updateChan: return chainhash.NewHash(update.ChanClose.ClosingTxid) } } // AssertChannelExists asserts that an active channel identified by // channelPoint is known to exist from the point-of-view of node.. func (n *networkHarness) AssertChannelExists(ctx context.Context, node *lightningNode, chanPoint *wire.OutPoint) error { req := &lnrpc.ListChannelsRequest{} resp, err := node.ListChannels(ctx, req) if err != nil { return fmt.Errorf("unable fetch node's channels: %v", err) } for _, channel := range resp.Channels { if channel.ChannelPoint == chanPoint.String() { return nil } } return fmt.Errorf("channel not found") } // DumpLogs reads the current logs generated by the passed node, and returns // the logs as a single string. This function is useful for examining the logs // of a particular node in the case of a test failure. // Logs from lightning node being generated with delay - you should // add time.Sleep() in order to get all logs. func (n *networkHarness) DumpLogs(node *lightningNode) (string, error) { logFile := fmt.Sprintf("%v/simnet/lnd.log", node.cfg.LogDir) buf, err := ioutil.ReadFile(logFile) if err != nil { return "", err } return string(buf), nil } // SendCoins attemps to send amt satoshis from the internal mining node to the // targetted lightning node. func (n *networkHarness) SendCoins(ctx context.Context, amt btcutil.Amount, target *lightningNode) error { balReq := &lnrpc.WalletBalanceRequest{} initialBalance, err := target.WalletBalance(ctx, balReq) if err != nil { return err } // First, obtain an address from the target lightning node, preferring // to receive a p2wkh address s.t the output can immediately be used as // an input to a funding transaction. addrReq := &lnrpc.NewAddressRequest{ Type: lnrpc.NewAddressRequest_WITNESS_PUBKEY_HASH, } resp, err := target.NewAddress(ctx, addrReq) if err != nil { return err } addr, err := btcutil.DecodeAddress(resp.Address, n.netParams) if err != nil { return err } addrScript, err := txscript.PayToAddrScript(addr) if err != nil { return err } // Generate a transaction which creates an output to the target // pkScript of the desired amount. output := &wire.TxOut{ PkScript: addrScript, Value: int64(amt), } if _, err := n.Miner.SendOutputs([]*wire.TxOut{output}, 30); err != nil { return err } // Finally, generate 6 new blocks to ensure the output gains a // sufficient number of confirmations. if _, err := n.Miner.Node.Generate(6); err != nil { return err } // Pause until the nodes current wallet balances reflects the amount // sent to it above. // TODO(roasbeef): factor out into helper func for { select { case <-time.Tick(time.Millisecond * 50): currentBal, err := target.WalletBalance(ctx, balReq) if err != nil { return err } if currentBal.Balance == initialBalance.Balance+amt.ToBTC() { return nil } case <-time.After(time.Second * 30): return fmt.Errorf("balances not synced after deadline") } } }