commit ed93163eb6ead427ecf625a51789793674a01d34
Author: crusader <crusader@loafle.com>
Date:   Thu Oct 26 20:14:00 2017 +0900

    ing

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..a5cf485
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,70 @@
+# Created by .ignore support plugin (hsz.mobi)
+### JetBrains template
+# Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio and Webstorm
+# Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839
+
+# User-specific stuff:
+.idea/**/workspace.xml
+.idea/**/tasks.xml
+.idea/dictionaries
+
+# Sensitive or high-churn files:
+.idea/**/dataSources/
+.idea/**/dataSources.ids
+.idea/**/dataSources.xml
+.idea/**/dataSources.local.xml
+.idea/**/sqlDataSources.xml
+.idea/**/dynamic.xml
+.idea/**/uiDesigner.xml
+
+# Gradle:
+.idea/**/gradle.xml
+.idea/**/libraries
+
+# Mongo Explorer plugin:
+.idea/**/mongoSettings.xml
+
+## File-based project format:
+*.iws
+
+## Plugin-specific files:
+
+# IntelliJ
+/out/
+
+# mpeltonen/sbt-idea plugin
+.idea_modules/
+
+# JIRA plugin
+atlassian-ide-plugin.xml
+
+# Crashlytics plugin (for Android Studio and IntelliJ)
+com_crashlytics_export_strings.xml
+crashlytics.properties
+crashlytics-build.properties
+fabric.properties
+### Go template
+# Binaries for programs and plugins
+*.exe
+*.dll
+*.so
+*.dylib
+
+# Test binary, build with `go test -c`
+*.test
+
+# Output of the go coverage tool, specifically when used with LiteIDE
+*.out
+
+# Project-local glide cache, RE: https://github.com/Masterminds/glide/issues/736
+.glide/
+.idea/
+*.iml
+
+vendor/
+glide.lock
+.DS_Store
+dist/
+debug
+
+
diff --git a/.vscode/launch.json b/.vscode/launch.json
new file mode 100644
index 0000000..2ca2b1d
--- /dev/null
+++ b/.vscode/launch.json
@@ -0,0 +1,32 @@
+{
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Debug",
+            "type": "go",
+            "request": "launch",
+            "mode": "debug",
+            "remotePath": "",
+            "port": 2345,
+            "host": "127.0.0.1",
+            "program": "${workspaceRoot}/main.go",
+            "env": {},
+            "args": [],
+            "showLog": true
+        },
+        {
+            "name": "File Debug",
+            "type": "go",
+            "request": "launch",
+            "mode": "debug",
+            "remotePath": "",
+            "port": 2345,
+            "host": "127.0.0.1",
+            "program": "${fileDirname}",
+            "env": {},
+            "args": [],
+            "showLog": true
+        }
+
+    ]
+}
\ No newline at end of file
diff --git a/.vscode/settings.json b/.vscode/settings.json
new file mode 100644
index 0000000..20af2f6
--- /dev/null
+++ b/.vscode/settings.json
@@ -0,0 +1,3 @@
+// Place your settings in this file to overwrite default and user settings.
+{
+}
\ No newline at end of file
diff --git a/glide.yaml b/glide.yaml
new file mode 100644
index 0000000..2098be0
--- /dev/null
+++ b/glide.yaml
@@ -0,0 +1,3 @@
+package: git.loafle.net/commons_go/server
+import:
+- package: git.loafle.net/commons_go/logging
diff --git a/server.go b/server.go
new file mode 100644
index 0000000..f95c5ce
--- /dev/null
+++ b/server.go
@@ -0,0 +1,124 @@
+package server
+
+import (
+	"fmt"
+	"io"
+	"net"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"git.loafle.net/commons_go/logging"
+)
+
+type Server interface {
+	Start() error
+	Stop()
+	Serve() error
+}
+
+type server struct {
+	serverHandler ServerHandler
+
+	listener net.Listener
+
+	Stats ServerStats
+
+	stopChan chan struct{}
+	stopWg   sync.WaitGroup
+}
+
+func (s *server) Start() error {
+	if nil != s.serverHandler {
+		panic("Server: server handler must be specified.")
+	}
+	s.serverHandler.Validate()
+
+	if s.stopChan != nil {
+		panic("Server: server is already running. Stop it before starting it again")
+	}
+	s.stopChan = make(chan struct{})
+	var err error
+	if s.listener, err = s.serverHandler.Listen(); nil != err {
+		return err
+	}
+
+	s.stopWg.Add(1)
+	go runServer(s)
+	return nil
+}
+
+func (s *server) Stop() {
+	if s.stopChan == nil {
+		panic("Server: server must be started before stopping it")
+	}
+	close(s.stopChan)
+	s.stopWg.Wait()
+	s.stopChan = nil
+}
+
+func (s *server) Serve() error {
+	if err := s.Start(); err != nil {
+		return err
+	}
+	s.stopWg.Wait()
+	return nil
+}
+
+func runServer(s *server) {
+	defer s.stopWg.Done()
+
+	var conn io.ReadWriteCloser
+	var clientAddr string
+	var err error
+	var stopping atomic.Value
+
+	for {
+		acceptChan := make(chan struct{})
+		go func() {
+			if conn, clientAddr, err = s.serverHandler.accept(s.listener); err != nil {
+				if stopping.Load() == nil {
+					logging.Logger.Error(fmt.Sprintf("Server: [%s]. Cannot accept new connection: [%s]", s.serverHandler.GetAddr(), err))
+				}
+			}
+			close(acceptChan)
+		}()
+
+		select {
+		case <-s.stopChan:
+			stopping.Store(true)
+			s.listener.Close()
+			<-acceptChan
+			return
+		case <-acceptChan:
+			s.Stats.incAcceptCalls()
+		}
+
+		if err != nil {
+			s.Stats.incAcceptErrors()
+			select {
+			case <-s.stopChan:
+				return
+			case <-time.After(time.Second):
+			}
+			continue
+		}
+
+		s.stopWg.Add(1)
+		go handleConnection(s, conn, clientAddr)
+	}
+}
+
+func handleConnection(s *server, conn io.ReadWriteCloser, clientAddr string) {
+	defer s.stopWg.Done()
+
+	stopChan := make(chan struct{})
+	go s.serverHandler.Handle(clientAddr, conn, stopChan)
+
+	select {
+	case <-s.stopChan:
+		close(stopChan)
+		conn.Close()
+		return
+	}
+}
diff --git a/server_handler.go b/server_handler.go
new file mode 100644
index 0000000..1517cc1
--- /dev/null
+++ b/server_handler.go
@@ -0,0 +1,18 @@
+package server
+
+import (
+	"io"
+	"net"
+)
+
+type ServerHandler interface {
+	Listen() (net.Listener, error)
+	Handle(remoteAddr string, rwc io.ReadWriteCloser, stopChan chan struct{})
+
+	GetAddr() string
+	GetPendingResponses() int
+
+	Validate()
+
+	accept(l net.Listener) (io.ReadWriteCloser, string, error)
+}
diff --git a/server_handlers.go b/server_handlers.go
new file mode 100644
index 0000000..63dc037
--- /dev/null
+++ b/server_handlers.go
@@ -0,0 +1,162 @@
+package server
+
+import (
+	"errors"
+	"io"
+	"net"
+	"time"
+)
+
+const (
+	// DefaultConcurrency is the default number of concurrent rpc calls
+	// the server can process.
+	DefaultConcurrency = 8 * 1024
+
+	// DefaultRequestTimeout is the default timeout for client request.
+	DefaultRequestTimeout = 20 * time.Second
+
+	// DefaultPendingMessages is the default number of pending messages
+	// handled by Client and Server.
+	DefaultPendingMessages = 32 * 1024
+
+	// DefaultFlushDelay is the default delay between message flushes
+	// on Client and Server.
+	DefaultFlushDelay = -1
+
+	// DefaultBufferSize is the default size for Client and Server buffers.
+	DefaultBufferSize = 64 * 1024
+
+	// DefaultKeepAlivePeriod is the default time for KeepAlivePeriod of connection.
+	DefaultKeepAlivePeriod = 0
+)
+
+type ServerHandlers struct {
+	// Address to listen to for incoming connections.
+	//
+	// The address format depends on the underlying transport provided
+	// by Server.Listener. The following transports are provided
+	// out of the box:
+	//   * TCP - see NewTCPServer() and NewTCPClient().
+	//   * TLS (aka SSL) - see NewTLSServer() and NewTLSClient().
+	//   * Unix sockets - see NewUnixServer() and NewUnixClient().
+	//
+	// By default TCP transport is used.
+	Addr string
+
+	// The maximum delay between response flushes to clients.
+	//
+	// Negative values lead to immediate requests' sending to the client
+	// without their buffering. This minimizes rpc latency at the cost
+	// of higher CPU and network usage.
+	//
+	// Default is DefaultFlushDelay.
+	FlushDelay time.Duration
+
+	// The maximum number of pending responses in the queue.
+	// Default is DefaultPendingMessages.
+	PendingResponses int
+
+	// Size of send buffer per each underlying connection in bytes.
+	// Default is DefaultBufferSize.
+	SendBufferSize int
+
+	// Size of recv buffer per each underlying connection in bytes.
+	// Default is DefaultBufferSize.
+	RecvBufferSize int
+
+	KeepAlivePeriod time.Duration
+}
+
+func (sh *ServerHandlers) Listen() (net.Listener, error) {
+	return nil, errors.New("Server: Handler method[Listen] of Server is not implement")
+}
+
+func (sh *ServerHandlers) Handle(remoteAddr string, rwc io.ReadWriteCloser, stopChan chan struct{}) {
+
+}
+
+func (sh *ServerHandlers) GetAddr() string {
+	return sh.Addr
+}
+
+func (sh *ServerHandlers) GetPendingResponses() int {
+	return sh.PendingResponses
+}
+
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+// func (sh *ServerHandlers)  {
+
+// }
+
+func (sh *ServerHandlers) accept(l net.Listener) (io.ReadWriteCloser, string, error) {
+	conn, err := l.Accept()
+	if nil != err {
+		return nil, "", err
+	}
+
+	if 0 < sh.KeepAlivePeriod {
+		if err = setupKeepalive(conn, sh.KeepAlivePeriod); nil != err {
+			conn.Close()
+			return nil, "", err
+		}
+	}
+
+	return conn, conn.RemoteAddr().String(), nil
+}
+
+func (sh *ServerHandlers) Validate() {
+	if sh.Concurrency <= 0 {
+		sh.Concurrency = DefaultConcurrency
+	}
+	if sh.FlushDelay == 0 {
+		sh.FlushDelay = DefaultFlushDelay
+	}
+	if sh.PendingResponses <= 0 {
+		sh.PendingResponses = DefaultPendingMessages
+	}
+	if sh.SendBufferSize <= 0 {
+		sh.SendBufferSize = DefaultBufferSize
+	}
+	if sh.RecvBufferSize <= 0 {
+		sh.RecvBufferSize = DefaultBufferSize
+	}
+	if sh.KeepAlivePeriod <= 0 {
+		sh.KeepAlivePeriod = DefaultKeepAlivePeriod
+	}
+
+}
+
+func setupKeepalive(conn net.Conn, keepAlivePeriod time.Duration) error {
+	tcpConn, ok := conn.(*net.TCPConn)
+	if !ok {
+		return errors.New("Server: Keepalive is valid when connection is net.TCPConn")
+	}
+
+	if err := tcpConn.SetKeepAlive(true); err != nil {
+		return err
+	}
+	if err := tcpConn.SetKeepAlivePeriod(keepAlivePeriod * time.Second); err != nil {
+		return err
+	}
+	return nil
+}
diff --git a/stats.go b/stats.go
new file mode 100644
index 0000000..c0ce5a4
--- /dev/null
+++ b/stats.go
@@ -0,0 +1,120 @@
+package server
+
+import (
+	"io"
+	"sync"
+	"time"
+)
+
+type ServerStats struct {
+	// The number of request performed.
+	RequestCount uint64
+
+	// The total aggregate time for all request in milliseconds.
+	//
+	// This time can be used for calculating the average response time
+	// per request:
+	//     avgRequesttime = RequestTime / RequestCount
+	RequestTime uint64
+
+	// The number of bytes written to the underlying connections.
+	BytesWritten uint64
+
+	// The number of bytes read from the underlying connections.
+	BytesRead uint64
+
+	// The number of Read() calls.
+	ReadCalls uint64
+
+	// The number of Read() errors.
+	ReadErrors uint64
+
+	// The number of Write() calls.
+	WriteCalls uint64
+
+	// The number of Write() errors.
+	WriteErrors uint64
+
+	// The number of Dial() calls.
+	DialCalls uint64
+
+	// The number of Dial() errors.
+	DialErrors uint64
+
+	// The number of Accept() calls.
+	AcceptCalls uint64
+
+	// The number of Accept() errors.
+	AcceptErrors uint64
+
+	// lock is for 386 builds. See https://github.com/valyala/gorpc/issues/5 .
+	lock sync.Mutex
+}
+
+// AvgRequestTime returns the average Request execution time.
+//
+// Use stats returned from ServerStats.Snapshot() on live Client and / or Server,
+// since the original stats can be updated by concurrently running goroutines.
+func (ss *ServerStats) AvgRequestTime() time.Duration {
+	return time.Duration(float64(ss.RequestTime)/float64(ss.RequestCount)) * time.Millisecond
+}
+
+// AvgRequestBytes returns the average bytes sent / received per Request.
+//
+// Use stats returned from ServerStats.Snapshot() on live Client and / or Server,
+// since the original stats can be updated by concurrently running goroutines.
+func (ss *ServerStats) AvgRequestBytes() (send float64, recv float64) {
+	return float64(ss.BytesWritten) / float64(ss.RequestCount), float64(ss.BytesRead) / float64(ss.RequestCount)
+}
+
+// AvgRequestCalls returns the average number of write() / read() syscalls per Request.
+//
+// Use stats returned from ServerStats.Snapshot() on live Client and / or Server,
+// since the original stats can be updated by concurrently running goroutines.
+func (ss *ServerStats) AvgRequestCalls() (write float64, read float64) {
+	return float64(ss.WriteCalls) / float64(ss.RequestCount), float64(ss.ReadCalls) / float64(ss.RequestCount)
+}
+
+type writerCounter struct {
+	w  io.Writer
+	ss *ServerStats
+}
+
+type readerCounter struct {
+	r  io.Reader
+	ss *ServerStats
+}
+
+func newWriterCounter(w io.Writer, ss *ServerStats) io.Writer {
+	return &writerCounter{
+		w:  w,
+		ss: ss,
+	}
+}
+
+func newReaderCounter(r io.Reader, ss *ServerStats) io.Reader {
+	return &readerCounter{
+		r:  r,
+		ss: ss,
+	}
+}
+
+func (w *writerCounter) Write(p []byte) (int, error) {
+	n, err := w.w.Write(p)
+	w.ss.incWriteCalls()
+	if err != nil {
+		w.ss.incWriteErrors()
+	}
+	w.ss.addBytesWritten(uint64(n))
+	return n, err
+}
+
+func (r *readerCounter) Read(p []byte) (int, error) {
+	n, err := r.r.Read(p)
+	r.ss.incReadCalls()
+	if err != nil {
+		r.ss.incReadErrors()
+	}
+	r.ss.addBytesRead(uint64(n))
+	return n, err
+}
diff --git a/stats_386.go b/stats_386.go
new file mode 100644
index 0000000..091efac
--- /dev/null
+++ b/stats_386.go
@@ -0,0 +1,106 @@
+package server
+
+import "sync"
+
+// Snapshot returns connection statistics' snapshot.
+//
+// Use stats returned from ServerStats.Snapshot() on live Client and / or Server,
+// since the original stats can be updated by concurrently running goroutines.
+func (ss *ServerStats) Snapshot() *ServerStats {
+	ss.lock.Lock()
+	snapshot := *ss
+	ss.lock.Unlock()
+
+	snapshot.lock = sync.Mutex{}
+	return &snapshot
+}
+
+// Reset resets all the stats counters.
+func (ss *ServerStats) Reset() {
+	ss.lock.Lock()
+	ss.RequestCount = 0
+	ss.RequestTime = 0
+	ss.BytesWritten = 0
+	ss.BytesRead = 0
+	ss.WriteCalls = 0
+	ss.WriteErrors = 0
+	ss.ReadCalls = 0
+	ss.ReadErrors = 0
+	ss.DialCalls = 0
+	ss.DialErrors = 0
+	ss.AcceptCalls = 0
+	ss.AcceptErrors = 0
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incRPCCalls() {
+	ss.lock.Lock()
+	ss.RequestCount++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incRPCTime(dt uint64) {
+	ss.lock.Lock()
+	ss.RequestTime += dt
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) addBytesWritten(n uint64) {
+	ss.lock.Lock()
+	ss.BytesWritten += n
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) addBytesRead(n uint64) {
+	ss.lock.Lock()
+	ss.BytesRead += n
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incReadCalls() {
+	ss.lock.Lock()
+	ss.ReadCalls++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incReadErrors() {
+	ss.lock.Lock()
+	ss.ReadErrors++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incWriteCalls() {
+	ss.lock.Lock()
+	ss.WriteCalls++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incWriteErrors() {
+	ss.lock.Lock()
+	ss.WriteErrors++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incDialCalls() {
+	ss.lock.Lock()
+	ss.DialCalls++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incDialErrors() {
+	ss.lock.Lock()
+	ss.DialErrors++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incAcceptCalls() {
+	ss.lock.Lock()
+	ss.AcceptCalls++
+	ss.lock.Unlock()
+}
+
+func (ss *ServerStats) incAcceptErrors() {
+	ss.lock.Lock()
+	ss.AcceptErrors++
+	ss.lock.Unlock()
+}
diff --git a/stats_generic.go b/stats_generic.go
new file mode 100644
index 0000000..dd9cba6
--- /dev/null
+++ b/stats_generic.go
@@ -0,0 +1,88 @@
+package server
+
+import "sync/atomic"
+
+// Snapshot returns connection statistics' snapshot.
+//
+// Use stats returned from ServerStats.Snapshot() on live Client and / or Server,
+// since the original stats can be updated by concurrently running goroutines.
+func (ss *ServerStats) Snapshot() *ServerStats {
+	return &ServerStats{
+		RequestCount: atomic.LoadUint64(&ss.RequestCount),
+		RequestTime:  atomic.LoadUint64(&ss.RequestTime),
+		BytesWritten: atomic.LoadUint64(&ss.BytesWritten),
+		BytesRead:    atomic.LoadUint64(&ss.BytesRead),
+		ReadCalls:    atomic.LoadUint64(&ss.ReadCalls),
+		ReadErrors:   atomic.LoadUint64(&ss.ReadErrors),
+		WriteCalls:   atomic.LoadUint64(&ss.WriteCalls),
+		WriteErrors:  atomic.LoadUint64(&ss.WriteErrors),
+		DialCalls:    atomic.LoadUint64(&ss.DialCalls),
+		DialErrors:   atomic.LoadUint64(&ss.DialErrors),
+		AcceptCalls:  atomic.LoadUint64(&ss.AcceptCalls),
+		AcceptErrors: atomic.LoadUint64(&ss.AcceptErrors),
+	}
+}
+
+// Reset resets all the stats counters.
+func (ss *ServerStats) Reset() {
+	atomic.StoreUint64(&ss.RequestCount, 0)
+	atomic.StoreUint64(&ss.RequestTime, 0)
+	atomic.StoreUint64(&ss.BytesWritten, 0)
+	atomic.StoreUint64(&ss.BytesRead, 0)
+	atomic.StoreUint64(&ss.WriteCalls, 0)
+	atomic.StoreUint64(&ss.WriteErrors, 0)
+	atomic.StoreUint64(&ss.ReadCalls, 0)
+	atomic.StoreUint64(&ss.ReadErrors, 0)
+	atomic.StoreUint64(&ss.DialCalls, 0)
+	atomic.StoreUint64(&ss.DialErrors, 0)
+	atomic.StoreUint64(&ss.AcceptCalls, 0)
+	atomic.StoreUint64(&ss.AcceptErrors, 0)
+}
+
+func (ss *ServerStats) incRPCCalls() {
+	atomic.AddUint64(&ss.RequestCount, 1)
+}
+
+func (ss *ServerStats) incRPCTime(dt uint64) {
+	atomic.AddUint64(&ss.RequestTime, dt)
+}
+
+func (ss *ServerStats) addBytesWritten(n uint64) {
+	atomic.AddUint64(&ss.BytesWritten, n)
+}
+
+func (ss *ServerStats) addBytesRead(n uint64) {
+	atomic.AddUint64(&ss.BytesRead, n)
+}
+
+func (ss *ServerStats) incReadCalls() {
+	atomic.AddUint64(&ss.ReadCalls, 1)
+}
+
+func (ss *ServerStats) incReadErrors() {
+	atomic.AddUint64(&ss.ReadErrors, 1)
+}
+
+func (ss *ServerStats) incWriteCalls() {
+	atomic.AddUint64(&ss.WriteCalls, 1)
+}
+
+func (ss *ServerStats) incWriteErrors() {
+	atomic.AddUint64(&ss.WriteErrors, 1)
+}
+
+func (ss *ServerStats) incDialCalls() {
+	atomic.AddUint64(&ss.DialCalls, 1)
+}
+
+func (ss *ServerStats) incDialErrors() {
+	atomic.AddUint64(&ss.DialErrors, 1)
+}
+
+func (ss *ServerStats) incAcceptCalls() {
+	atomic.AddUint64(&ss.AcceptCalls, 1)
+}
+
+func (ss *ServerStats) incAcceptErrors() {
+	atomic.AddUint64(&ss.AcceptErrors, 1)
+}