put can read from stdin by specifying the '-' source file

.travis.yml

@@ -3,7 +3,7 @@ os:
 - osx
 language: go
 go_import_path: github.com/colinmarc/hdfs
-go: 1.7
+go: 1.x
 before_install:
 - git clone https://github.com/sstephenson/bats $HOME/bats
 - mkdir -p $HOME/bats/build
@@ -19,11 +19,12 @@ before_script:
 - "./setup_test_env.sh"
 before_deploy: make release
 script:
+- find protocol -name *.pb.go | xargs touch # so make doesn't try to regen protobuf files
 - make test
 - cat minicluster.log
 sudo: false
 cache:
-- "$HADOOP_HOME"
+- "$HOME/hadoop-$HADOOP_DISTRO"
 - "$HOME/bats"
 deploy:
   skip_cleanup: true

client.go

@@ -98,6 +98,8 @@ func getNameNodeFromConf() ([]string, error) {
 
 // NewForUser returns a connected Client with the user specified, or an error if
 // it can't connect.
+//
+// Deprecated: Use NewClient with ClientOptions instead.
 func NewForUser(address string, user string) (*Client, error) {
 	return NewClient(ClientOptions{
 		Addresses: []string{address},
@@ -107,6 +109,8 @@ func NewForUser(address string, user string) (*Client, error) {
 
 // NewForConnection returns Client with the specified, underlying rpc.NamenodeConnection.
 // You can use rpc.WrapNamenodeConnection to wrap your own net.Conn.
+//
+// Deprecated: Use NewClient with ClientOptions instead.
 func NewForConnection(namenode *rpc.NamenodeConnection) *Client {
 	client, _ := NewClient(ClientOptions{Namenode: namenode})
 	return client
@@ -143,7 +147,6 @@ func (c *Client) CopyToLocal(src string, dst string) error {
 }
 
 // CopyToRemote copies the local file specified by src to the HDFS file at dst.
-// If dst already exists, it will be overwritten.
 func (c *Client) CopyToRemote(src string, dst string) error {
 	local, err := os.Open(src)
 	if err != nil {

cmd/hdfs/put.go

@@ -6,6 +6,8 @@ import (
 	"os"
 	"path"
 	"path/filepath"
+
+	"github.com/colinmarc/hdfs"
 )
 
 func put(args []string) {
@@ -29,6 +31,40 @@ func put(args []string) {
 		fatal(err)
 	}
 
+	if filepath.Base(source) == "-" {
+		putFromStdin(client, dest)
+	} else {
+		putFromFile(client, source, dest)
+	}
+}
+
+func putFromStdin(client *hdfs.Client, dest string) {
+	// If the destination exists, regardless of what it is, bail out.
+	_, err := client.Stat(dest)
+	if err == nil {
+		fatal(&os.PathError{"put", dest, os.ErrExist})
+	} else if !os.IsNotExist(err) {
+		fatal(err)
+	}
+
+	mode := 0755 | os.ModeDir
+	parentDir := filepath.Dir(dest)
+	if parentDir != "." && parentDir != "/" {
+		if err := client.MkdirAll(parentDir, mode); err != nil {
+			fatal(err)
+		}
+	}
+
+	writer, err := client.Create(dest)
+	if err != nil {
+		fatal(err)
+	}
+	defer writer.Close()
+
+	io.Copy(writer, os.Stdin)
+}
+
+func putFromFile(client *hdfs.Client, source string, dest string) {
 	// If the destination is an existing directory, place it inside. Otherwise,
 	// the destination is really the parent directory, and we need to rename the
 	// source directory as we copy.

cmd/hdfs/test/put.bats

@@ -64,6 +64,33 @@ mkdir /_test_cmd/put/existing.txt: file already exists
 OUT
 }
 
+@test "put from stdin" {
+  run bash -c "cat $ROOT_TEST_DIR/test/mobydick.txt | $HDFS put - /_test_cmd/put_stdin/mobydick_stdin.txt"
+  assert_success
+
+  run bash -c "$HDFS cat /_test_cmd/put_stdin/mobydick_stdin.txt > $BATS_TMPDIR/mobydick_stdin_test.txt"
+  assert_success
+
+  SHA=`shasum < $ROOT_TEST_DIR/test/mobydick.txt | awk '{ print $1 }'`
+  assert_equal $SHA `shasum < $BATS_TMPDIR/mobydick_stdin_test.txt | awk '{ print $1 }'`
+}
+
+@test "put from stdin into existing file" {
+  run bash -c "cat $ROOT_TEST_DIR/test/mobydick.txt | $HDFS put - /_test_cmd/put/existing.txt"
+  assert_failure
+  assert_output <<OUT
+put /_test_cmd/put/existing.txt: file already exists
+OUT
+}
+
+@test "put from stdin into dir" {
+  run bash -c "cat $ROOT_TEST_DIR/test/mobydick.txt | $HDFS put - /_test_cmd/put/"
+  assert_failure
+  assert_output <<OUT
+put /_test_cmd/put: file already exists
+OUT
+}
+
 teardown() {
   $HDFS rm -r /_test_cmd/put
 }