eucjp.go raw
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.
4
5 package japanese
6
7 import (
8 "unicode/utf8"
9
10 "golang.org/x/text/encoding"
11 "golang.org/x/text/encoding/internal"
12 "golang.org/x/text/encoding/internal/identifier"
13 "golang.org/x/text/transform"
14 )
15
16 // EUCJP is the EUC-JP encoding.
17 var EUCJP encoding.Encoding = &eucJP
18
19 var eucJP = internal.Encoding{
20 Encoding: &internal.SimpleEncoding{Decoder: eucJPDecoder{}, Encoder: eucJPEncoder{}},
21 Name: "EUC-JP",
22 MIB: identifier.EUCPkdFmtJapanese,
23 }
24
25 type eucJPDecoder struct{ transform.NopResetter }
26
27 // See https://encoding.spec.whatwg.org/#euc-jp-decoder.
28 func (eucJPDecoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
29 r, size := rune(0), 0
30 loop:
31 for ; nSrc < len(src); nSrc += size {
32 switch c0 := src[nSrc]; {
33 case c0 < utf8.RuneSelf:
34 r, size = rune(c0), 1
35
36 case c0 == 0x8e:
37 if nSrc+1 >= len(src) {
38 if !atEOF {
39 err = transform.ErrShortSrc
40 break loop
41 }
42 r, size = utf8.RuneError, 1
43 break
44 }
45 c1 := src[nSrc+1]
46 switch {
47 case c1 < 0xa1:
48 r, size = utf8.RuneError, 1
49 case c1 > 0xdf:
50 r, size = utf8.RuneError, 2
51 if c1 == 0xff {
52 size = 1
53 }
54 default:
55 r, size = rune(c1)+(0xff61-0xa1), 2
56 }
57 case c0 == 0x8f:
58 if nSrc+2 >= len(src) {
59 if !atEOF {
60 err = transform.ErrShortSrc
61 break loop
62 }
63 r, size = utf8.RuneError, 1
64 if p := nSrc + 1; p < len(src) && 0xa1 <= src[p] && src[p] < 0xfe {
65 size = 2
66 }
67 break
68 }
69 c1 := src[nSrc+1]
70 if c1 < 0xa1 || 0xfe < c1 {
71 r, size = utf8.RuneError, 1
72 break
73 }
74 c2 := src[nSrc+2]
75 if c2 < 0xa1 || 0xfe < c2 {
76 r, size = utf8.RuneError, 2
77 break
78 }
79 r, size = utf8.RuneError, 3
80 if i := int(c1-0xa1)*94 + int(c2-0xa1); i < len(jis0212Decode) {
81 r = rune(jis0212Decode[i])
82 if r == 0 {
83 r = utf8.RuneError
84 }
85 }
86
87 case 0xa1 <= c0 && c0 <= 0xfe:
88 if nSrc+1 >= len(src) {
89 if !atEOF {
90 err = transform.ErrShortSrc
91 break loop
92 }
93 r, size = utf8.RuneError, 1
94 break
95 }
96 c1 := src[nSrc+1]
97 if c1 < 0xa1 || 0xfe < c1 {
98 r, size = utf8.RuneError, 1
99 break
100 }
101 r, size = utf8.RuneError, 2
102 if i := int(c0-0xa1)*94 + int(c1-0xa1); i < len(jis0208Decode) {
103 r = rune(jis0208Decode[i])
104 if r == 0 {
105 r = utf8.RuneError
106 }
107 }
108
109 default:
110 r, size = utf8.RuneError, 1
111 }
112
113 if nDst+utf8.RuneLen(r) > len(dst) {
114 err = transform.ErrShortDst
115 break loop
116 }
117 nDst += utf8.EncodeRune(dst[nDst:], r)
118 }
119 return nDst, nSrc, err
120 }
121
122 type eucJPEncoder struct{ transform.NopResetter }
123
124 func (eucJPEncoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
125 r, size := rune(0), 0
126 for ; nSrc < len(src); nSrc += size {
127 r = rune(src[nSrc])
128
129 // Decode a 1-byte rune.
130 if r < utf8.RuneSelf {
131 size = 1
132
133 } else {
134 // Decode a multi-byte rune.
135 r, size = utf8.DecodeRune(src[nSrc:])
136 if size == 1 {
137 // All valid runes of size 1 (those below utf8.RuneSelf) were
138 // handled above. We have invalid UTF-8 or we haven't seen the
139 // full character yet.
140 if !atEOF && !utf8.FullRune(src[nSrc:]) {
141 err = transform.ErrShortSrc
142 break
143 }
144 }
145
146 // func init checks that the switch covers all tables.
147 switch {
148 case encode0Low <= r && r < encode0High:
149 if r = rune(encode0[r-encode0Low]); r != 0 {
150 goto write2or3
151 }
152 case encode1Low <= r && r < encode1High:
153 if r = rune(encode1[r-encode1Low]); r != 0 {
154 goto write2or3
155 }
156 case encode2Low <= r && r < encode2High:
157 if r = rune(encode2[r-encode2Low]); r != 0 {
158 goto write2or3
159 }
160 case encode3Low <= r && r < encode3High:
161 if r = rune(encode3[r-encode3Low]); r != 0 {
162 goto write2or3
163 }
164 case encode4Low <= r && r < encode4High:
165 if r = rune(encode4[r-encode4Low]); r != 0 {
166 goto write2or3
167 }
168 case encode5Low <= r && r < encode5High:
169 if 0xff61 <= r && r < 0xffa0 {
170 goto write2
171 }
172 if r = rune(encode5[r-encode5Low]); r != 0 {
173 goto write2or3
174 }
175 }
176 err = internal.ErrASCIIReplacement
177 break
178 }
179
180 if nDst >= len(dst) {
181 err = transform.ErrShortDst
182 break
183 }
184 dst[nDst] = uint8(r)
185 nDst++
186 continue
187
188 write2or3:
189 if r>>tableShift == jis0208 {
190 if nDst+2 > len(dst) {
191 err = transform.ErrShortDst
192 break
193 }
194 } else {
195 if nDst+3 > len(dst) {
196 err = transform.ErrShortDst
197 break
198 }
199 dst[nDst] = 0x8f
200 nDst++
201 }
202 dst[nDst+0] = 0xa1 + uint8(r>>codeShift)&codeMask
203 dst[nDst+1] = 0xa1 + uint8(r)&codeMask
204 nDst += 2
205 continue
206
207 write2:
208 if nDst+2 > len(dst) {
209 err = transform.ErrShortDst
210 break
211 }
212 dst[nDst+0] = 0x8e
213 dst[nDst+1] = uint8(r - (0xff61 - 0xa1))
214 nDst += 2
215 continue
216 }
217 return nDst, nSrc, err
218 }
219
220 func init() {
221 // Check that the hard-coded encode switch covers all tables.
222 if numEncodeTables != 6 {
223 panic("bad numEncodeTables")
224 }
225 }
226