provisioner/shell: convert windows line endings to Unix [GH-277]

CHANGELOG.md

@@ -13,6 +13,9 @@ IMPROVEMENTS:
 * core: User variables can now be used for integer, boolean, etc.
   values. [GH-418]
 * builder/amazon/all: Interrupts work while waiting for AMI to be ready.
+* provisioner/shell: Script line-endings are automatically converted to
+  Unix-style line-endings. Can be disabled by setting "binary" to "true".
+  [GH-277]
 
 BUG FIXES:
 

provisioner/shell/provisioner.go

@@ -8,6 +8,7 @@ import (
 	"fmt"
 	"github.com/mitchellh/packer/common"
 	"github.com/mitchellh/packer/packer"
+	"io"
 	"io/ioutil"
 	"log"
 	"os"
@@ -20,6 +21,10 @@ const DefaultRemotePath = "/tmp/script.sh"
 type config struct {
 	common.PackerConfig `mapstructure:",squash"`
 
+	// If true, the script contains binary and line endings will not be
+	// converted from Windows to Unix-style.
+	Binary bool
+
 	// An inline script to execute. Multiple strings are all executed
 	// in the context of a single shell.
 	Inline []string
@@ -259,6 +264,11 @@ func (p *Provisioner) Provision(ui packer.Ui, comm packer.Communicator) error {
 				return err
 			}
 
+			var r io.Reader = f
+			if !p.config.Binary {
+				r = &UnixReader{Reader: r}
+			}
+
 			if err := comm.Upload(p.config.RemotePath, f); err != nil {
 				return fmt.Errorf("Error uploading script: %s", err)
 			}

provisioner/shell/unix_reader.go

+package shell
+
+import (
+	"bufio"
+	"bytes"
+	"io"
+	"sync"
+)
+
+// UnixReader is a Reader implementation that automatically converts
+// Windows line endings to Unix line endings.
+type UnixReader struct {
+	Reader io.Reader
+
+	buf     []byte
+	once    sync.Once
+	scanner *bufio.Scanner
+}
+
+func (r *UnixReader) Read(p []byte) (n int, err error) {
+	// Create the buffered reader once
+	r.once.Do(func() {
+		r.scanner = bufio.NewScanner(r.Reader)
+		r.scanner.Split(scanUnixLine)
+	})
+
+	// If we have no data in our buffer, scan to the next token
+	if len(r.buf) == 0 {
+		if !r.scanner.Scan() {
+			err = r.scanner.Err()
+			if err == nil {
+				err = io.EOF
+			}
+
+			return 0, err
+		}
+
+		r.buf = r.scanner.Bytes()
+	}
+
+	// Write out as much data as we can to the buffer, storing the rest
+	// for the next read.
+	n = len(p)
+	if n > len(r.buf) {
+		n = len(r.buf)
+	}
+	copy(p, r.buf)
+	r.buf = r.buf[n:]
+
+	return
+}
+
+// scanUnixLine is a bufio.Scanner SplitFunc. It tokenizes on lines, but
+// only returns unix-style lines. So even if the line is "one\r\n", the
+// token returned will be "one\n".
+func scanUnixLine(data []byte, atEOF bool) (advance int, token []byte, err error) {
+	if atEOF && len(data) == 0 {
+		return 0, nil, nil
+	}
+
+	if i := bytes.IndexByte(data, '\n'); i >= 0 {
+		// We have a new-line terminated line. Return the line with the newline
+		return i + 1, dropCR(data[0 : i+1]), nil
+	}
+
+	if atEOF {
+		// We have a final, non-terminated line
+		return len(data), dropCR(data), nil
+	}
+
+	if data[len(data)-1] != '\r' {
+		// We have a normal line, just let it tokenize
+		return len(data), data, nil
+	}
+
+	// We need more data
+	return 0, nil, nil
+}
+
+func dropCR(data []byte) []byte {
+	if len(data) > 0 && data[len(data)-2] == '\r' {
+		// Trim off the last byte and replace it with a '\n'
+		data = data[0 : len(data)-1]
+		data[len(data)-1] = '\n'
+	}
+
+	return data
+}

provisioner/shell/unix_reader_test.go

+package shell
+
+import (
+	"bytes"
+	"io"
+	"testing"
+)
+
+func TestUnixReader_impl(t *testing.T) {
+	var raw interface{}
+	raw = new(UnixReader)
+	if _, ok := raw.(io.Reader); !ok {
+		t.Fatal("should be reader")
+	}
+}
+
+func TestUnixReader(t *testing.T) {
+	input := "one\r\ntwo\nthree\r\n"
+	expected := "one\ntwo\nthree\n"
+
+	r := &UnixReader{
+		Reader: bytes.NewReader([]byte(input)),
+	}
+
+	result := new(bytes.Buffer)
+	if _, err := io.Copy(result, r); err != nil {
+		t.Fatalf("err: %s", err)
+	}
+
+	if result.String() != expected {
+		t.Fatalf("bad: %#v", result.String())
+	}
+}

website/source/docs/provisioners/shell.html.markdown

@@ -47,6 +47,10 @@ Exactly _one_ of the following is required:
 
 Optional parameters:
 
+* `binary` (boolean) - If true, specifies that the script(s) are binary
+   files, and Packer should therefore not convert Windows line endings to
+   Unix line endings (if there are any). By default this is false.
+
 * `environment_vars` (array of strings) - An array of key/value pairs
   to inject prior to the execute_command. The format should be
   `key=value`. Packer injects some environmental variables by default