1 /*
2 * "streamable kanji code filter and converter"
3 * Copyright (c) 1998-2002 HappySize, Inc. All rights reserved.
4 *
5 * LICENSE NOTICES
6 *
7 * This file is part of "streamable kanji code filter and converter",
8 * which is distributed under the terms of GNU Lesser General Public
9 * License (version 2) as published by the Free Software Foundation.
10 *
11 * This software is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with "streamable kanji code filter and converter";
18 * if not, write to the Free Software Foundation, Inc., 59 Temple Place,
19 * Suite 330, Boston, MA 02111-1307 USA
20 *
21 * The author of this file:
22 *
23 */
24 /*
25 * The source code included in this files was separated from mbfilter_ja.c
26 * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
27 *
28 */
29
30 #include "mbfilter.h"
31 #include "mbfilter_sjis.h"
32
33 #define UNICODE_TABLE_CP932_DEF
34 #define UNICODE_TABLE_JIS_DEF
35
36 #include "unicode_table_cp932_ext.h"
37 #include "unicode_table_jis.h"
38
39 int mbfl_filt_ident_sjis(int c, mbfl_identify_filter *filter);
40
41 const unsigned char mblen_table_sjis[] = { /* 0x80-0x9f,0xE0-0xFF */
42 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
43 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
44 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
45 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
46 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
47 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
48 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
49 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
50 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
51 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
52 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
53 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
54 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
55 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
56 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
57 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2
58 };
59
60 static const char *mbfl_encoding_sjis_aliases[] = {"x-sjis", "SHIFT-JIS", NULL};
61
62 const mbfl_encoding mbfl_encoding_sjis = {
63 mbfl_no_encoding_sjis,
64 "SJIS",
65 "Shift_JIS",
66 (const char *(*)[])&mbfl_encoding_sjis_aliases,
67 mblen_table_sjis,
68 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE,
69 &vtbl_sjis_wchar,
70 &vtbl_wchar_sjis
71 };
72
73 const struct mbfl_identify_vtbl vtbl_identify_sjis = {
74 mbfl_no_encoding_sjis,
75 mbfl_filt_ident_common_ctor,
76 mbfl_filt_ident_sjis
77 };
78
79 const struct mbfl_convert_vtbl vtbl_sjis_wchar = {
80 mbfl_no_encoding_sjis,
81 mbfl_no_encoding_wchar,
82 mbfl_filt_conv_common_ctor,
83 NULL,
84 mbfl_filt_conv_sjis_wchar,
85 mbfl_filt_conv_common_flush,
86 NULL,
87 };
88
89 const struct mbfl_convert_vtbl vtbl_wchar_sjis = {
90 mbfl_no_encoding_wchar,
91 mbfl_no_encoding_sjis,
92 mbfl_filt_conv_common_ctor,
93 NULL,
94 mbfl_filt_conv_wchar_sjis,
95 mbfl_filt_conv_common_flush,
96 NULL,
97 };
98
99 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
100
101 #define SJIS_ENCODE(c1,c2,s1,s2) \
102 do { \
103 s1 = c1; \
104 s1--; \
105 s1 >>= 1; \
106 if ((c1) < 0x5f) { \
107 s1 += 0x71; \
108 } else { \
109 s1 += 0xb1; \
110 } \
111 s2 = c2; \
112 if ((c1) & 1) { \
113 if ((c2) < 0x60) { \
114 s2--; \
115 } \
116 s2 += 0x20; \
117 } else { \
118 s2 += 0x7e; \
119 } \
120 } while (0)
121
122 #define SJIS_DECODE(c1,c2,s1,s2) \
123 do { \
124 s1 = c1; \
125 if (s1 < 0xa0) { \
126 s1 -= 0x81; \
127 } else { \
128 s1 -= 0xc1; \
129 } \
130 s1 <<= 1; \
131 s1 += 0x21; \
132 s2 = c2; \
133 if (s2 < 0x9f) { \
134 if (s2 < 0x7f) { \
135 s2++; \
136 } \
137 s2 -= 0x20; \
138 } else { \
139 s1++; \
140 s2 -= 0x7e; \
141 } \
142 } while (0)
143
144
145 /*
146 * SJIS => wchar
147 */
148 int
mbfl_filt_conv_sjis_wchar(int c,mbfl_convert_filter * filter)149 mbfl_filt_conv_sjis_wchar(int c, mbfl_convert_filter *filter)
150 {
151 int c1, s1, s2, w;
152
153 switch (filter->status) {
154 case 0:
155 if (c >= 0 && c < 0x80) { /* latin */
156 CK((*filter->output_function)(c, filter->data));
157 } else if (c > 0xa0 && c < 0xe0) { /* kana */
158 CK((*filter->output_function)(0xfec0 + c, filter->data));
159 } else if (c > 0x80 && c < 0xfd && c != 0xa0) { /* kanji first char */
160 filter->status = 1;
161 filter->cache = c;
162 } else {
163 w = c & MBFL_WCSGROUP_MASK;
164 w |= MBFL_WCSGROUP_THROUGH;
165 CK((*filter->output_function)(w, filter->data));
166 }
167 break;
168
169 case 1: /* kanji second char */
170 filter->status = 0;
171 c1 = filter->cache;
172 if (c >= 0x40 && c <= 0xfc && c != 0x7f) {
173 SJIS_DECODE(c1, c, s1, s2);
174 w = (s1 - 0x21)*94 + s2 - 0x21;
175 if (w >= 0 && w < jisx0208_ucs_table_size) {
176 w = jisx0208_ucs_table[w];
177 } else {
178 w = 0;
179 }
180 if (w <= 0) {
181 if (s1 < 0x7f && s2 < 0x7f) {
182 w = (s1 << 8) | s2;
183 w &= MBFL_WCSPLANE_MASK;
184 w |= MBFL_WCSPLANE_JIS0208;
185 } else {
186 w = (c1 << 8) | c;
187 w &= MBFL_WCSGROUP_MASK;
188 w |= MBFL_WCSGROUP_THROUGH;
189 }
190 }
191 CK((*filter->output_function)(w, filter->data));
192 } else if ((c >= 0 && c < 0x21) || c == 0x7f) { /* CTLs */
193 CK((*filter->output_function)(c, filter->data));
194 } else {
195 w = (c1 << 8) | c;
196 w &= MBFL_WCSGROUP_MASK;
197 w |= MBFL_WCSGROUP_THROUGH;
198 CK((*filter->output_function)(w, filter->data));
199 }
200 break;
201
202 default:
203 filter->status = 0;
204 break;
205 }
206
207 return c;
208 }
209
210 /*
211 * wchar => SJIS
212 */
213 int
mbfl_filt_conv_wchar_sjis(int c,mbfl_convert_filter * filter)214 mbfl_filt_conv_wchar_sjis(int c, mbfl_convert_filter *filter)
215 {
216 int c1, c2, s1, s2;
217
218 s1 = 0;
219 if (c >= ucs_a1_jis_table_min && c < ucs_a1_jis_table_max) {
220 s1 = ucs_a1_jis_table[c - ucs_a1_jis_table_min];
221 } else if (c >= ucs_a2_jis_table_min && c < ucs_a2_jis_table_max) {
222 s1 = ucs_a2_jis_table[c - ucs_a2_jis_table_min];
223 } else if (c >= ucs_i_jis_table_min && c < ucs_i_jis_table_max) {
224 s1 = ucs_i_jis_table[c - ucs_i_jis_table_min];
225 } else if (c >= ucs_r_jis_table_min && c < ucs_r_jis_table_max) {
226 s1 = ucs_r_jis_table[c - ucs_r_jis_table_min];
227 }
228 if (s1 <= 0) {
229 c1 = c & ~MBFL_WCSPLANE_MASK;
230 if (c1 == MBFL_WCSPLANE_JIS0208) {
231 s1 = c & MBFL_WCSPLANE_MASK;
232 } else if (c == 0xa5) { /* YEN SIGN */
233 s1 = 0x216f; /* FULLWIDTH YEN SIGN */
234 } else if (c == 0x203e) { /* OVER LINE */
235 s1 = 0x2131; /* FULLWIDTH MACRON */
236 } else if (c == 0xff3c) { /* FULLWIDTH REVERSE SOLIDUS */
237 s1 = 0x2140;
238 } else if (c == 0xff5e) { /* FULLWIDTH TILDE */
239 s1 = 0x2141;
240 } else if (c == 0x2225) { /* PARALLEL TO */
241 s1 = 0x2142;
242 } else if (c == 0xff0d) { /* FULLWIDTH HYPHEN-MINUS */
243 s1 = 0x215d;
244 } else if (c == 0xffe0) { /* FULLWIDTH CENT SIGN */
245 s1 = 0x2171;
246 } else if (c == 0xffe1) { /* FULLWIDTH POUND SIGN */
247 s1 = 0x2172;
248 } else if (c == 0xffe2) { /* FULLWIDTH NOT SIGN */
249 s1 = 0x224c;
250 }
251 if (c == 0) {
252 s1 = 0;
253 } else if (s1 <= 0) {
254 s1 = -1;
255 }
256 } else if (s1 >= 0x8080) {
257 s1 = -1;
258 }
259 if (s1 >= 0) {
260 if (s1 < 0x100) { /* latin or kana */
261 CK((*filter->output_function)(s1, filter->data));
262 } else { /* kanji */
263 c1 = (s1 >> 8) & 0xff;
264 c2 = s1 & 0xff;
265 SJIS_ENCODE(c1, c2, s1, s2);
266 CK((*filter->output_function)(s1, filter->data));
267 CK((*filter->output_function)(s2, filter->data));
268 }
269 } else {
270 CK(mbfl_filt_conv_illegal_output(c, filter));
271 }
272
273 return c;
274 }
275
mbfl_filt_ident_sjis(int c,mbfl_identify_filter * filter)276 int mbfl_filt_ident_sjis(int c, mbfl_identify_filter *filter)
277 {
278 if (filter->status) { /* kanji second char */
279 if (c < 0x40 || c > 0xfc || c == 0x7f) { /* bad */
280 filter->flag = 1;
281 }
282 filter->status = 0;
283 } else if (c >= 0 && c < 0x80) { /* latin ok */
284 ;
285 } else if (c > 0xa0 && c < 0xe0) { /* kana ok */
286 ;
287 } else if (c > 0x80 && c < 0xf0 && c != 0xa0) { /* kanji first char */
288 filter->status = 1;
289 } else { /* bad */
290 filter->flag = 1;
291 }
292
293 return c;
294 }
295
