1 /**********************************************************************
2 koi8.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2008 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 */
29
30 #include "regenc.h"
31
32 #define ENC_KOI8_TO_LOWER_CASE(c) EncKOI8_ToLowerCaseTable[c]
33 #define ENC_IS_KOI8_CTYPE(code,ctype) \
34 ((EncKOI8_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
35
36 static const UChar EncKOI8_ToLowerCaseTable[256] = {
37 '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
38 '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
39 '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
40 '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
41 '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
42 '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
43 '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
44 '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
45 '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
46 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
47 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
48 '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
49 '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
50 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
51 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
52 '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
53 '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
54 '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
55 '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
56 '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
57 '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
58 '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257',
59 '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
60 '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277',
61 '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
62 '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
63 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
64 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337',
65 '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
66 '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
67 '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
68 '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337'
69 };
70
71 static const unsigned short EncKOI8_CtypeTable[256] = {
72 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
73 0x4008, 0x420c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
74 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
75 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
76 0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
77 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
78 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
79 0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
80 0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
81 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
82 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
83 0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
84 0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
85 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
86 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
87 0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
88 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
89 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
90 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
91 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
92 0x0284, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000,
93 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000,
94 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000,
95 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000, 0x0000,
96 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
97 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
98 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
99 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
100 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
101 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
102 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
103 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2
104 };
105
106
107 static int
koi8_mbc_case_fold(OnigCaseFoldType flag ARG_UNUSED,const UChar ** pp,const UChar * end ARG_UNUSED,UChar * lower)108 koi8_mbc_case_fold(OnigCaseFoldType flag ARG_UNUSED,
109 const UChar** pp, const UChar* end ARG_UNUSED, UChar* lower)
110 {
111 const UChar* p = *pp;
112
113 *lower = ENC_KOI8_TO_LOWER_CASE(*p);
114 (*pp)++;
115 return 1;
116 }
117
118 #if 0
119 static int
120 koi8_is_mbc_ambiguous(OnigAmbigType flag, const OnigUChar** pp, const OnigUChar* end)
121 {
122 const OnigUChar* p = *pp;
123
124 (*pp)++;
125 if (((flag & ONIGENC_CASE_FOLD_ASCII_CASE) != 0 &&
126 ONIGENC_IS_MBC_ASCII(p)) ||
127 ((flag & ONIGENC_CASE_FOLD_NONASCII_CASE) != 0 &&
128 !ONIGENC_IS_MBC_ASCII(p))) {
129 int v = (EncKOI8_CtypeTable[*p] &
130 (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
131 return (v != 0 ? TRUE : FALSE);
132 }
133 return FALSE;
134 }
135 #endif
136
137 static int
koi8_is_code_ctype(OnigCodePoint code,unsigned int ctype)138 koi8_is_code_ctype(OnigCodePoint code, unsigned int ctype)
139 {
140 if (code < 256)
141 return ENC_IS_KOI8_CTYPE(code, ctype);
142 else
143 return FALSE;
144 }
145
146 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
147 { 0xc0, 0xe0 },
148 { 0xc1, 0xe1 },
149 { 0xc2, 0xe2 },
150 { 0xc3, 0xe3 },
151 { 0xc4, 0xe4 },
152 { 0xc5, 0xe5 },
153 { 0xc6, 0xe6 },
154 { 0xc7, 0xe7 },
155 { 0xc8, 0xe8 },
156 { 0xc9, 0xe9 },
157 { 0xca, 0xea },
158 { 0xcb, 0xeb },
159 { 0xcc, 0xec },
160 { 0xcd, 0xed },
161 { 0xce, 0xee },
162 { 0xcf, 0xef },
163
164 { 0xd0, 0xf0 },
165 { 0xd1, 0xf1 },
166 { 0xd2, 0xf2 },
167 { 0xd3, 0xf3 },
168 { 0xd4, 0xf4 },
169 { 0xd5, 0xf5 },
170 { 0xd6, 0xf6 },
171 { 0xd7, 0xf7 },
172 { 0xd8, 0xf8 },
173 { 0xd9, 0xf9 },
174 { 0xda, 0xfa },
175 { 0xdb, 0xfb },
176 { 0xdc, 0xfc },
177 { 0xdd, 0xfd },
178 { 0xde, 0xfe },
179 { 0xdf, 0xff },
180
181 { 0xe0, 0xc0 },
182 { 0xe1, 0xc1 },
183 { 0xe2, 0xc2 },
184 { 0xe3, 0xc3 },
185 { 0xe4, 0xc4 },
186 { 0xe5, 0xc5 },
187 { 0xe6, 0xc6 },
188 { 0xe7, 0xc7 },
189 { 0xe8, 0xc8 },
190 { 0xe9, 0xc9 },
191 { 0xea, 0xca },
192 { 0xeb, 0xcb },
193 { 0xec, 0xcc },
194 { 0xed, 0xcd },
195 { 0xee, 0xce },
196 { 0xef, 0xcf },
197
198 { 0xf0, 0xd0 },
199 { 0xf1, 0xd1 },
200 { 0xf2, 0xd2 },
201 { 0xf3, 0xd3 },
202 { 0xf4, 0xd4 },
203 { 0xf5, 0xd5 },
204 { 0xf6, 0xd6 },
205 { 0xf7, 0xd7 },
206 { 0xf8, 0xd8 },
207 { 0xf9, 0xd9 },
208 { 0xfa, 0xda },
209 { 0xfb, 0xdb },
210 { 0xfc, 0xdc },
211 { 0xfe, 0xde },
212 { 0xff, 0xdf }
213 };
214
215 static int
koi8_apply_all_case_fold(OnigCaseFoldType flag,OnigApplyAllCaseFoldFunc f,void * arg)216 koi8_apply_all_case_fold(OnigCaseFoldType flag,
217 OnigApplyAllCaseFoldFunc f, void* arg)
218 {
219 return onigenc_apply_all_case_fold_with_map(
220 sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 0,
221 flag, f, arg);
222 }
223
224 static int
koi8_get_case_fold_codes_by_str(OnigCaseFoldType flag,const OnigUChar * p,const OnigUChar * end,OnigCaseFoldCodeItem items[])225 koi8_get_case_fold_codes_by_str(OnigCaseFoldType flag,
226 const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[])
227 {
228 return onigenc_get_case_fold_codes_by_str_with_map(
229 sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 0,
230 flag, p, end, items);
231 }
232
233 OnigEncodingType OnigEncodingKOI8 = {
234 onigenc_single_byte_mbc_enc_len,
235 "KOI8", /* name */
236 1, /* max enc length */
237 1, /* min enc length */
238 onigenc_is_mbc_newline_0x0a,
239 onigenc_single_byte_mbc_to_code,
240 onigenc_single_byte_code_to_mbclen,
241 onigenc_single_byte_code_to_mbc,
242 koi8_mbc_case_fold,
243 koi8_apply_all_case_fold,
244 koi8_get_case_fold_codes_by_str,
245 onigenc_minimum_property_name_to_ctype,
246 koi8_is_code_ctype,
247 onigenc_not_support_get_ctype_code_range,
248 onigenc_single_byte_left_adjust_char_head,
249 onigenc_always_true_is_allowed_reverse_match
250 };
251
