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