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_cp51932.h"
32
33 #include "unicode_table_cp932_ext.h"
34 #include "unicode_table_jis.h"
35 #include "cp932_table.h"
36
37 static int mbfl_filt_conv_cp51932_wchar_flush(mbfl_convert_filter *filter);
38
39 static const unsigned char mblen_table_eucjp[] = { /* 0xA1-0xFE */
40 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
41 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
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, 2, 1,
49 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
50 1, 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 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
53 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
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, 1
56 };
57
58 static const char *mbfl_encoding_cp51932_aliases[] = {"cp51932", NULL};
59
60 const mbfl_encoding mbfl_encoding_cp51932 = {
61 mbfl_no_encoding_cp51932,
62 "CP51932",
63 "CP51932",
64 mbfl_encoding_cp51932_aliases,
65 mblen_table_eucjp,
66 0,
67 &vtbl_cp51932_wchar,
68 &vtbl_wchar_cp51932,
69 NULL
70 };
71
72 const struct mbfl_convert_vtbl vtbl_cp51932_wchar = {
73 mbfl_no_encoding_cp51932,
74 mbfl_no_encoding_wchar,
75 mbfl_filt_conv_common_ctor,
76 NULL,
77 mbfl_filt_conv_cp51932_wchar,
78 mbfl_filt_conv_cp51932_wchar_flush,
79 NULL,
80 };
81
82 const struct mbfl_convert_vtbl vtbl_wchar_cp51932 = {
83 mbfl_no_encoding_wchar,
84 mbfl_no_encoding_cp51932,
85 mbfl_filt_conv_common_ctor,
86 NULL,
87 mbfl_filt_conv_wchar_cp51932,
88 mbfl_filt_conv_common_flush,
89 NULL,
90 };
91
92 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
93
94 /*
95 * cp51932 => wchar
96 */
97 int
mbfl_filt_conv_cp51932_wchar(int c,mbfl_convert_filter * filter)98 mbfl_filt_conv_cp51932_wchar(int c, mbfl_convert_filter *filter)
99 {
100 int c1, s, w;
101
102 switch (filter->status) {
103 case 0:
104 if (c >= 0 && c < 0x80) { /* latin */
105 CK((*filter->output_function)(c, filter->data));
106 } else if (c >= 0xA1 && c <= 0xFE) { /* CP932, first byte */
107 filter->status = 1;
108 filter->cache = c;
109 } else if (c == 0x8e) { /* kana first char */
110 filter->status = 2;
111 } else {
112 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
113 }
114 break;
115
116 case 1: /* got first half */
117 filter->status = 0;
118 c1 = filter->cache;
119 if (c > 0xa0 && c < 0xff) {
120 w = 0;
121 s = (c1 - 0xa1)*94 + c - 0xa1;
122 if (s <= 137) {
123 if (s == 31) {
124 w = 0xff3c; /* FULLWIDTH REVERSE SOLIDUS */
125 } else if (s == 32) {
126 w = 0xff5e; /* FULLWIDTH TILDE */
127 } else if (s == 33) {
128 w = 0x2225; /* PARALLEL TO */
129 } else if (s == 60) {
130 w = 0xff0d; /* FULLWIDTH HYPHEN-MINUS */
131 } else if (s == 80) {
132 w = 0xffe0; /* FULLWIDTH CENT SIGN */
133 } else if (s == 81) {
134 w = 0xffe1; /* FULLWIDTH POUND SIGN */
135 } else if (s == 137) {
136 w = 0xffe2; /* FULLWIDTH NOT SIGN */
137 }
138 }
139 if (w == 0) {
140 if (s >= cp932ext1_ucs_table_min && s < cp932ext1_ucs_table_max) { /* vendor ext1 (13ku) */
141 w = cp932ext1_ucs_table[s - cp932ext1_ucs_table_min];
142 } else if (s >= 0 && s < jisx0208_ucs_table_size) { /* X 0208 */
143 w = jisx0208_ucs_table[s];
144 } else if (s >= cp932ext2_ucs_table_min && s < cp932ext2_ucs_table_max) { /* vendor ext2 (89ku - 92ku) */
145 w = cp932ext2_ucs_table[s - cp932ext2_ucs_table_min];
146 }
147 }
148 if (w <= 0) {
149 w = MBFL_BAD_INPUT;
150 }
151 CK((*filter->output_function)(w, filter->data));
152 } else {
153 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
154 }
155 break;
156
157 case 2: /* got 0x8e, X0201 kana */
158 filter->status = 0;
159 if (c > 0xa0 && c < 0xe0) {
160 w = 0xfec0 + c;
161 CK((*filter->output_function)(w, filter->data));
162 } else {
163 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
164 }
165 break;
166
167 default:
168 filter->status = 0;
169 break;
170 }
171
172 return 0;
173 }
174
mbfl_filt_conv_cp51932_wchar_flush(mbfl_convert_filter * filter)175 static int mbfl_filt_conv_cp51932_wchar_flush(mbfl_convert_filter *filter)
176 {
177 if (filter->status) {
178 /* Input string was truncated */
179 (*filter->output_function)(MBFL_BAD_INPUT, filter->data);
180 filter->status = 0;
181 }
182
183 if (filter->flush_function) {
184 (*filter->flush_function)(filter->data);
185 }
186
187 return 0;
188 }
189
190 /*
191 * wchar => cp51932
192 */
193 int
mbfl_filt_conv_wchar_cp51932(int c,mbfl_convert_filter * filter)194 mbfl_filt_conv_wchar_cp51932(int c, mbfl_convert_filter *filter)
195 {
196 int c1, c2, s1;
197
198 s1 = 0;
199 if (c >= ucs_a1_jis_table_min && c < ucs_a1_jis_table_max) {
200 s1 = ucs_a1_jis_table[c - ucs_a1_jis_table_min];
201 } else if (c >= ucs_a2_jis_table_min && c < ucs_a2_jis_table_max) {
202 s1 = ucs_a2_jis_table[c - ucs_a2_jis_table_min];
203 } else if (c >= ucs_i_jis_table_min && c < ucs_i_jis_table_max) {
204 s1 = ucs_i_jis_table[c - ucs_i_jis_table_min];
205 } else if (c >= ucs_r_jis_table_min && c < ucs_r_jis_table_max) {
206 s1 = ucs_r_jis_table[c - ucs_r_jis_table_min];
207 }
208 if (s1 >= 0x8080) s1 = -1; /* we don't support JIS X0213 */
209 if (s1 <= 0) {
210 if (c == 0xa5) { /* YEN SIGN */
211 s1 = 0x216F; /* FULLWIDTH YEN SIGN */
212 } else if (c == 0xff3c) { /* FULLWIDTH REVERSE SOLIDUS */
213 s1 = 0x2140;
214 } else if (c == 0x2225) { /* PARALLEL TO */
215 s1 = 0x2142;
216 } else if (c == 0xff0d) { /* FULLWIDTH HYPHEN-MINUS */
217 s1 = 0x215d;
218 } else if (c == 0xffe0) { /* FULLWIDTH CENT SIGN */
219 s1 = 0x2171;
220 } else if (c == 0xffe1) { /* FULLWIDTH POUND SIGN */
221 s1 = 0x2172;
222 } else if (c == 0xffe2) { /* FULLWIDTH NOT SIGN */
223 s1 = 0x224c;
224 } else {
225 s1 = -1;
226 c1 = 0;
227 c2 = cp932ext1_ucs_table_max - cp932ext1_ucs_table_min;
228 while (c1 < c2) { /* CP932 vendor ext1 (13ku) */
229 if (c == cp932ext1_ucs_table[c1]) {
230 s1 = ((c1/94 + 0x2d) << 8) + (c1%94 + 0x21);
231 break;
232 }
233 c1++;
234 }
235 if (s1 < 0) {
236 c1 = 0;
237 c2 = cp932ext2_ucs_table_max - cp932ext2_ucs_table_min;
238 while (c1 < c2) { /* CP932 vendor ext3 (115ku - 119ku) */
239 if (c == cp932ext2_ucs_table[c1]) {
240 s1 = ((c1/94 + 0x79) << 8) +(c1%94 + 0x21);
241 break;
242 }
243 c1++;
244 }
245 }
246 }
247 if (c == 0) {
248 s1 = 0;
249 } else if (s1 <= 0) {
250 s1 = -1;
251 }
252 }
253
254 if (s1 >= 0) {
255 if (s1 < 0x80) { /* latin */
256 CK((*filter->output_function)(s1, filter->data));
257 } else if (s1 < 0x100) { /* kana */
258 CK((*filter->output_function)(0x8e, filter->data));
259 CK((*filter->output_function)(s1, filter->data));
260 } else if (s1 < 0x8080) { /* X 0208 */
261 CK((*filter->output_function)(((s1 >> 8) & 0xff) | 0x80, filter->data));
262 CK((*filter->output_function)((s1 & 0xff) | 0x80, filter->data));
263 } else {
264 CK(mbfl_filt_conv_illegal_output(c, filter));
265 }
266 } else {
267 CK(mbfl_filt_conv_illegal_output(c, filter));
268 }
269
270 return 0;
271 }
272
