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 #ifdef HAVE_CONFIG_H
31 #include "config.h"
32 #endif
33
34 #include "mbfilter.h"
35 #include "mbfilter_sjis.h"
36
37 #define UNICODE_TABLE_CP932_DEF
38 #define UNICODE_TABLE_JIS_DEF
39
40 #include "unicode_table_cp932_ext.h"
41 #include "unicode_table_jis.h"
42
43 int mbfl_filt_ident_sjis(int c, mbfl_identify_filter *filter);
44
45 const unsigned char mblen_table_sjis[] = { /* 0x80-0x9f,0xE0-0xFF */
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 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
51 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
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 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
55 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
56 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
57 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
58 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
59 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
60 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
61 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2
62 };
63
64 static const char *mbfl_encoding_sjis_aliases[] = {"x-sjis", "SHIFT-JIS", NULL};
65
66 const mbfl_encoding mbfl_encoding_sjis = {
67 mbfl_no_encoding_sjis,
68 "SJIS",
69 "Shift_JIS",
70 (const char *(*)[])&mbfl_encoding_sjis_aliases,
71 mblen_table_sjis,
72 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE,
73 &vtbl_sjis_wchar,
74 &vtbl_wchar_sjis
75 };
76
77 const struct mbfl_identify_vtbl vtbl_identify_sjis = {
78 mbfl_no_encoding_sjis,
79 mbfl_filt_ident_common_ctor,
80 mbfl_filt_ident_common_dtor,
81 mbfl_filt_ident_sjis
82 };
83
84 const struct mbfl_convert_vtbl vtbl_sjis_wchar = {
85 mbfl_no_encoding_sjis,
86 mbfl_no_encoding_wchar,
87 mbfl_filt_conv_common_ctor,
88 mbfl_filt_conv_common_dtor,
89 mbfl_filt_conv_sjis_wchar,
90 mbfl_filt_conv_common_flush
91 };
92
93 const struct mbfl_convert_vtbl vtbl_wchar_sjis = {
94 mbfl_no_encoding_wchar,
95 mbfl_no_encoding_sjis,
96 mbfl_filt_conv_common_ctor,
97 mbfl_filt_conv_common_dtor,
98 mbfl_filt_conv_wchar_sjis,
99 mbfl_filt_conv_common_flush
100 };
101
102 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
103
104 #define SJIS_ENCODE(c1,c2,s1,s2) \
105 do { \
106 s1 = c1; \
107 s1--; \
108 s1 >>= 1; \
109 if ((c1) < 0x5f) { \
110 s1 += 0x71; \
111 } else { \
112 s1 += 0xb1; \
113 } \
114 s2 = c2; \
115 if ((c1) & 1) { \
116 if ((c2) < 0x60) { \
117 s2--; \
118 } \
119 s2 += 0x20; \
120 } else { \
121 s2 += 0x7e; \
122 } \
123 } while (0)
124
125 #define SJIS_DECODE(c1,c2,s1,s2) \
126 do { \
127 s1 = c1; \
128 if (s1 < 0xa0) { \
129 s1 -= 0x81; \
130 } else { \
131 s1 -= 0xc1; \
132 } \
133 s1 <<= 1; \
134 s1 += 0x21; \
135 s2 = c2; \
136 if (s2 < 0x9f) { \
137 if (s2 < 0x7f) { \
138 s2++; \
139 } \
140 s2 -= 0x20; \
141 } else { \
142 s1++; \
143 s2 -= 0x7e; \
144 } \
145 } while (0)
146
147
148 /*
149 * SJIS => wchar
150 */
151 int
mbfl_filt_conv_sjis_wchar(int c,mbfl_convert_filter * filter)152 mbfl_filt_conv_sjis_wchar(int c, mbfl_convert_filter *filter)
153 {
154 int c1, s1, s2, w;
155
156 switch (filter->status) {
157 case 0:
158 if (c >= 0 && c < 0x80) { /* latin */
159 CK((*filter->output_function)(c, filter->data));
160 } else if (c > 0xa0 && c < 0xe0) { /* kana */
161 CK((*filter->output_function)(0xfec0 + c, filter->data));
162 } else if (c > 0x80 && c < 0xfd && c != 0xa0) { /* kanji first char */
163 filter->status = 1;
164 filter->cache = c;
165 } else {
166 w = c & MBFL_WCSGROUP_MASK;
167 w |= MBFL_WCSGROUP_THROUGH;
168 CK((*filter->output_function)(w, filter->data));
169 }
170 break;
171
172 case 1: /* kanji second char */
173 filter->status = 0;
174 c1 = filter->cache;
175 if (c >= 0x40 && c <= 0xfc && c != 0x7f) {
176 SJIS_DECODE(c1, c, s1, s2);
177 w = (s1 - 0x21)*94 + s2 - 0x21;
178 if (w >= 0 && w < jisx0208_ucs_table_size) {
179 w = jisx0208_ucs_table[w];
180 } else {
181 w = 0;
182 }
183 if (w <= 0) {
184 if (s1 < 0x7f && s2 < 0x7f) {
185 w = (s1 << 8) | s2;
186 w &= MBFL_WCSPLANE_MASK;
187 w |= MBFL_WCSPLANE_JIS0208;
188 } else {
189 w = (c1 << 8) | c;
190 w &= MBFL_WCSGROUP_MASK;
191 w |= MBFL_WCSGROUP_THROUGH;
192 }
193 }
194 CK((*filter->output_function)(w, filter->data));
195 } else if ((c >= 0 && c < 0x21) || c == 0x7f) { /* CTLs */
196 CK((*filter->output_function)(c, filter->data));
197 } else {
198 w = (c1 << 8) | c;
199 w &= MBFL_WCSGROUP_MASK;
200 w |= MBFL_WCSGROUP_THROUGH;
201 CK((*filter->output_function)(w, filter->data));
202 }
203 break;
204
205 default:
206 filter->status = 0;
207 break;
208 }
209
210 return c;
211 }
212
213 /*
214 * wchar => SJIS
215 */
216 int
mbfl_filt_conv_wchar_sjis(int c,mbfl_convert_filter * filter)217 mbfl_filt_conv_wchar_sjis(int c, mbfl_convert_filter *filter)
218 {
219 int c1, c2, s1, s2;
220
221 s1 = 0;
222 if (c >= ucs_a1_jis_table_min && c < ucs_a1_jis_table_max) {
223 s1 = ucs_a1_jis_table[c - ucs_a1_jis_table_min];
224 } else if (c >= ucs_a2_jis_table_min && c < ucs_a2_jis_table_max) {
225 s1 = ucs_a2_jis_table[c - ucs_a2_jis_table_min];
226 } else if (c >= ucs_i_jis_table_min && c < ucs_i_jis_table_max) {
227 s1 = ucs_i_jis_table[c - ucs_i_jis_table_min];
228 } else if (c >= ucs_r_jis_table_min && c < ucs_r_jis_table_max) {
229 s1 = ucs_r_jis_table[c - ucs_r_jis_table_min];
230 }
231 if (s1 <= 0) {
232 c1 = c & ~MBFL_WCSPLANE_MASK;
233 if (c1 == MBFL_WCSPLANE_JIS0208) {
234 s1 = c & MBFL_WCSPLANE_MASK;
235 } else if (c == 0xa5) { /* YEN SIGN */
236 s1 = 0x216f; /* FULLWIDTH YEN SIGN */
237 } else if (c == 0x203e) { /* OVER LINE */
238 s1 = 0x2131; /* FULLWIDTH MACRON */
239 } else if (c == 0xff3c) { /* FULLWIDTH REVERSE SOLIDUS */
240 s1 = 0x2140;
241 } else if (c == 0xff5e) { /* FULLWIDTH TILDE */
242 s1 = 0x2141;
243 } else if (c == 0x2225) { /* PARALLEL TO */
244 s1 = 0x2142;
245 } else if (c == 0xff0d) { /* FULLWIDTH HYPHEN-MINUS */
246 s1 = 0x215d;
247 } else if (c == 0xffe0) { /* FULLWIDTH CENT SIGN */
248 s1 = 0x2171;
249 } else if (c == 0xffe1) { /* FULLWIDTH POUND SIGN */
250 s1 = 0x2172;
251 } else if (c == 0xffe2) { /* FULLWIDTH NOT SIGN */
252 s1 = 0x224c;
253 }
254 if (c == 0) {
255 s1 = 0;
256 } else if (s1 <= 0) {
257 s1 = -1;
258 }
259 } else if (s1 >= 0x8080) {
260 s1 = -1;
261 }
262 if (s1 >= 0) {
263 if (s1 < 0x100) { /* latin or kana */
264 CK((*filter->output_function)(s1, filter->data));
265 } else { /* kanji */
266 c1 = (s1 >> 8) & 0xff;
267 c2 = s1 & 0xff;
268 SJIS_ENCODE(c1, c2, s1, s2);
269 CK((*filter->output_function)(s1, filter->data));
270 CK((*filter->output_function)(s2, filter->data));
271 }
272 } else {
273 if (filter->illegal_mode != MBFL_OUTPUTFILTER_ILLEGAL_MODE_NONE) {
274 CK(mbfl_filt_conv_illegal_output(c, filter));
275 }
276 }
277
278 return c;
279 }
280
mbfl_filt_ident_sjis(int c,mbfl_identify_filter * filter)281 int mbfl_filt_ident_sjis(int c, mbfl_identify_filter *filter)
282 {
283 if (filter->status) { /* kanji second char */
284 if (c < 0x40 || c > 0xfc || c == 0x7f) { /* bad */
285 filter->flag = 1;
286 }
287 filter->status = 0;
288 } else if (c >= 0 && c < 0x80) { /* latin ok */
289 ;
290 } else if (c > 0xa0 && c < 0xe0) { /* kana ok */
291 ;
292 } else if (c > 0x80 && c < 0xf0 && c != 0xa0) { /* kanji first char */
293 filter->status = 1;
294 } else { /* bad */
295 filter->flag = 1;
296 }
297
298 return c;
299 }
300
