mail: decode RFC 2047 "B" encoding.

R=rsc, r
CC=golang-dev
https://golang.org/cl/4584056

src/pkg/mail/message.go

@@ -18,8 +18,10 @@ package mail
 import (
 	"bufio"
 	"bytes"
+	"encoding/base64"
 	"fmt"
 	"io"
+	"io/ioutil"
 	"log"
 	"net/textproto"
 	"os"
@@ -428,37 +430,68 @@ func decodeRFC2047Word(s string) (string, os.Error) {
 		return "", os.ErrorString("mail: address not RFC 2047 encoded")
 	}
 	charset, enc := strings.ToLower(fields[1]), strings.ToLower(fields[2])
-	// TODO(dsymonds): Support "b" encoding too.
-	if enc != "q" {
-		return "", fmt.Errorf("mail: RFC 2047 encoding not supported: %q", enc)
-	}
 	if charset != "iso-8859-1" && charset != "utf-8" {
 		return "", fmt.Errorf("mail: charset not supported: %q", charset)
 	}
 
-	in := fields[3]
-	b := new(bytes.Buffer)
-	for i := 0; i < len(in); i++ {
-		switch c := in[i]; {
-		case c == '=' && i+2 < len(in):
-			x, err := strconv.Btoi64(in[i+1:i+3], 16)
-			if err != nil {
-				return "", fmt.Errorf("mail: invalid RFC 2047 encoding: %q", in[i:i+3])
+	in := bytes.NewBufferString(fields[3])
+	var r io.Reader
+	switch enc {
+	case "b":
+		r = base64.NewDecoder(base64.StdEncoding, in)
+	case "q":
+		r = qDecoder{r: in}
+	default:
+		return "", fmt.Errorf("mail: RFC 2047 encoding not supported: %q", enc)
 	}
-			i += 2
+
+	dec, err := ioutil.ReadAll(r)
+	if err != nil {
+		return "", err
+	}
+
 	switch charset {
 	case "iso-8859-1":
-				b.WriteRune(int(x))
-			case "utf-8":
-				b.WriteByte(byte(x))
-			}
-		case c == '_':
-			b.WriteByte(' ')
-		default:
-			b.WriteByte(c)
-		}
+		b := new(bytes.Buffer)
+		for _, c := range dec {
+			b.WriteRune(int(c))
 		}
 		return b.String(), nil
+	case "utf-8":
+		return string(dec), nil
+	}
+	panic("unreachable")
+}
+
+type qDecoder struct {
+	r       io.Reader
+	scratch [2]byte
+}
+
+func (qd qDecoder) Read(p []byte) (n int, err os.Error) {
+	// This method writes at most one byte into p.
+	if len(p) == 0 {
+		return 0, nil
+	}
+	if _, err := qd.r.Read(qd.scratch[:1]); err != nil {
+		return 0, err
+	}
+	switch c := qd.scratch[0]; {
+	case c == '=':
+		if _, err := io.ReadFull(qd.r, qd.scratch[:2]); err != nil {
+			return 0, err
+		}
+		x, err := strconv.Btoi64(string(qd.scratch[:2]), 16)
+		if err != nil {
+			return 0, fmt.Errorf("mail: invalid RFC 2047 encoding: %q", qd.scratch[:2])
+		}
+		p[0] = byte(x)
+	case c == '_':
+		p[0] = ' '
+	default:
+		p[0] = c
+	}
+	return 1, nil
 }
 
 var atextChars = []byte("ABCDEFGHIJKLMNOPQRSTUVWXYZ" +

src/pkg/mail/message_test.go

@@ -217,6 +217,27 @@ func TestAddressParsing(t *testing.T) {
 				},
 			},
 		},
+		// Custom example of RFC 2047 "B"-encoded ISO-8859-1 address.
+		{
+			`=?ISO-8859-1?B?SvZyZw==?= <joerg@example.com>`,
+			[]*Address{
+				&Address{
+					Name:    `Jörg`,
+					Address: "joerg@example.com",
+				},
+			},
+		},
+		// Custom example of RFC 2047 "B"-encoded UTF-8 address.
+		{
+			// XXX: a different example
+			`=?UTF-8?B?SsO2cmc=?= <joerg@example.com>`,
+			[]*Address{
+				&Address{
+					Name:    `Jörg`,
+					Address: "joerg@example.com",
+				},
+			},
+		},
 	}
 	for _, test := range tests {
 		addrs, err := newAddrParser(test.addrsStr).parseAddressList()
@@ -225,7 +246,7 @@ func TestAddressParsing(t *testing.T) {
 			continue
 		}
 		if !reflect.DeepEqual(addrs, test.exp) {
-			t.Errorf("Parse of %q: got %+v, want %+v", test.addrsStr, addrs, test.exp)
+			t.Errorf("Parse of %q: got %+v, want %+v", test.addrsStr, *addrs[0], *test.exp[0])
 		}
 	}
 }