1
// Copyright 2013 The Go Authors. All rights reserved.
2
// Use of this source code is governed by a BSD-style
3
// license that can be found in the LICENSE file.
11
"code.google.com/p/go.text/encoding"
12
"code.google.com/p/go.text/transform"
15
// EUCKR is the EUC-KR encoding, also known as Code Page 949.
16
var EUCKR encoding.Encoding = eucKR{}
20
func (eucKR) NewDecoder() transform.Transformer {
24
func (eucKR) NewEncoder() transform.Transformer {
28
func (eucKR) String() string {
32
var errInvalidEUCKR = errors.New("korean: invalid EUC-KR encoding")
34
type eucKRDecoder struct{}
36
func (eucKRDecoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
39
for ; nSrc < len(src); nSrc += size {
40
switch c0 := src[nSrc]; {
41
case c0 < utf8.RuneSelf:
44
case 0x81 <= c0 && c0 < 0xff:
45
if nSrc+1 >= len(src) {
46
err = transform.ErrShortSrc
51
r = 178 * rune(c0-0x81)
53
case 0x41 <= c1 && c1 < 0x5b:
54
r += rune(c1) - (0x41 - 0*26)
55
case 0x61 <= c1 && c1 < 0x7b:
56
r += rune(c1) - (0x61 - 1*26)
57
case 0x81 <= c1 && c1 < 0xff:
58
r += rune(c1) - (0x81 - 2*26)
63
} else if 0xa1 <= c1 && c1 < 0xff {
64
r = 178*(0xc7-0x81) + rune(c0-0xc7)*94 + rune(c1-0xa1)
69
if int(r) < len(decode) {
84
if nDst+utf8.RuneLen(r) > len(dst) {
85
err = transform.ErrShortDst
88
nDst += utf8.EncodeRune(dst[nDst:], r)
90
if atEOF && err == transform.ErrShortSrc {
93
return nDst, nSrc, err
96
type eucKREncoder struct{}
98
func (eucKREncoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
100
for ; nSrc < len(src); nSrc += size {
103
// Decode a 1-byte rune.
104
if r < utf8.RuneSelf {
108
// Decode a multi-byte rune.
109
r, size = utf8.DecodeRune(src[nSrc:])
111
// All valid runes of size 1 (those below utf8.RuneSelf) were
112
// handled above. We have invalid UTF-8 or we haven't seen the
113
// full character yet.
114
if !atEOF && !utf8.FullRune(src[nSrc:]) {
115
err = transform.ErrShortSrc
120
// func init checks that the switch covers all tables.
122
case encode0Low <= r && r < encode0High:
123
if r = rune(encode0[r-encode0Low]); r != 0 {
126
case encode1Low <= r && r < encode1High:
127
if r = rune(encode1[r-encode1Low]); r != 0 {
130
case encode2Low <= r && r < encode2High:
131
if r = rune(encode2[r-encode2Low]); r != 0 {
134
case encode3Low <= r && r < encode3High:
135
if r = rune(encode3[r-encode3Low]); r != 0 {
138
case encode4Low <= r && r < encode4High:
139
if r = rune(encode4[r-encode4Low]); r != 0 {
142
case encode5Low <= r && r < encode5High:
143
if r = rune(encode5[r-encode5Low]); r != 0 {
146
case encode6Low <= r && r < encode6High:
147
if r = rune(encode6[r-encode6Low]); r != 0 {
151
r = encoding.ASCIISub
154
if nDst >= len(dst) {
155
err = transform.ErrShortDst
163
if nDst+2 > len(dst) {
164
err = transform.ErrShortDst
167
dst[nDst+0] = uint8(r >> 8)
168
dst[nDst+1] = uint8(r)
172
return nDst, nSrc, err
176
// Check that the hard-coded encode switch covers all tables.
177
if numEncodeTables != 7 {
178
panic("bad numEncodeTables")