1/* Access functions for CNS 11643 handling.
2 Copyright (C) 1998-2020 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
5
6 The GNU C Library is free software; you can redistribute it and/or
7 modify it under the terms of the GNU Lesser General Public
8 License as published by the Free Software Foundation; either
9 version 2.1 of the License, or (at your option) any later version.
10
11 The GNU C Library 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 GNU
14 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 the GNU C Library; if not, see
18 <https://www.gnu.org/licenses/>. */
19
20#include <stdint.h>
21
22
23/* Table for CNS 11643, plane 1 to UCS4 conversion. */
24extern const uint16_t __cns11643l1_to_ucs4_tab[];
25/* Table for CNS 11643, plane 2 to UCS4 conversion. */
26extern const uint16_t __cns11643l2_to_ucs4_tab[];
27/* Table for CNS 11643, plane 3 to UCS4 conversion. */
28extern const uint32_t __cns11643l3_to_ucs4_tab[];
29/* Table for CNS 11643, plane 4 to UCS4 conversion. */
30extern const uint32_t __cns11643l4_to_ucs4_tab[];
31/* Table for CNS 11643, plane 5 to UCS4 conversion. */
32extern const uint32_t __cns11643l5_to_ucs4_tab[];
33/* Table for CNS 11643, plane 6 to UCS4 conversion. */
34extern const uint32_t __cns11643l6_to_ucs4_tab[];
35/* Table for CNS 11643, plane 7 to UCS4 conversion. */
36extern const uint32_t __cns11643l7_to_ucs4_tab[];
37/* Table for CNS 11643, plane 15 (old) to UCS4 conversion. */
38extern const uint32_t __cns11643l15_to_ucs4_tab[];
39
40
41static inline uint32_t
42__attribute ((always_inline))
43cns11643_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
44{
45 unsigned char ch = *(*s);
46 unsigned char ch2;
47 unsigned char ch3;
48 uint32_t result;
49 int idx;
50
51 if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) > 0x30)
52 return __UNKNOWN_10646_CHAR;
53
54 if (avail < 3)
55 return 0;
56
57 ch2 = (*s)[1];
58 if ((ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
59 return __UNKNOWN_10646_CHAR;
60
61 ch3 = (*s)[2];
62 if ((ch3 - offset) <= 0x20 || (ch3 - offset) >= 0x7f)
63 return __UNKNOWN_10646_CHAR;
64
65 idx = (ch2 - 0x21 - offset) * 94 + (ch3 - 0x21 - offset);
66
67 switch (ch - 0x20 - offset)
68 {
69 case 1:
70 if (idx > 0x21f2)
71 return __UNKNOWN_10646_CHAR;
72 result = __cns11643l1_to_ucs4_tab[idx];
73 break;
74 case 2:
75 if (idx > 0x1de1)
76 return __UNKNOWN_10646_CHAR;
77 result = __cns11643l2_to_ucs4_tab[idx];
78 break;
79 case 3:
80 if (idx > 0x19bd)
81 return __UNKNOWN_10646_CHAR;
82 result = __cns11643l3_to_ucs4_tab[idx];
83 break;
84 case 4:
85 if (idx > 0x1c81)
86 return __UNKNOWN_10646_CHAR;
87 result = __cns11643l4_to_ucs4_tab[idx];
88 break;
89 case 5:
90 if (idx > 0x219a)
91 return __UNKNOWN_10646_CHAR;
92 result = __cns11643l5_to_ucs4_tab[idx];
93 break;
94 case 6:
95 if (idx > 0x18f3)
96 return __UNKNOWN_10646_CHAR;
97 result = __cns11643l6_to_ucs4_tab[idx];
98 break;
99 case 7:
100 if (idx > 0x198a)
101 return __UNKNOWN_10646_CHAR;
102 result = __cns11643l7_to_ucs4_tab[idx];
103 break;
104 case 15:
105 if (idx > 0x1c00)
106 return __UNKNOWN_10646_CHAR;
107 result = __cns11643l15_to_ucs4_tab[idx];
108 break;
109 default:
110 return __UNKNOWN_10646_CHAR;
111 }
112
113 if (result != L'\0')
114 (*s) += 3;
115 else
116 result = __UNKNOWN_10646_CHAR;
117
118 return result;
119}
120
121
122/* Tables for the UCS4 -> CNS conversion. */
123extern const char __cns11643l1_from_ucs4_tab1[][2];
124extern const char __cns11643l1_from_ucs4_tab2[][2];
125extern const char __cns11643l1_from_ucs4_tab3[][2];
126extern const char __cns11643l1_from_ucs4_tab4[][2];
127extern const char __cns11643l1_from_ucs4_tab5[][2];
128extern const char __cns11643l1_from_ucs4_tab6[][2];
129extern const char __cns11643l1_from_ucs4_tab7[][2];
130extern const char __cns11643l1_from_ucs4_tab8[][2];
131extern const char __cns11643l1_from_ucs4_tab9[][2];
132extern const char __cns11643l1_from_ucs4_tab10[][2];
133extern const char __cns11643l1_from_ucs4_tab11[][2];
134extern const char __cns11643l1_from_ucs4_tab12[][2];
135extern const char __cns11643l1_from_ucs4_tab13[][2];
136extern const char __cns11643l1_from_ucs4_tab14[][2];
137extern const char __cns11643_from_ucs4p0_tab[][3];
138extern const char __cns11643_from_ucs4p2_tab[][3];
139extern const char __cns11643_from_ucs4p2c_tab[][3];
140
141
142static inline size_t
143__attribute ((always_inline))
144ucs4_to_cns11643 (uint32_t wch, unsigned char *s, size_t avail)
145{
146 unsigned int ch = (unsigned int) wch;
147 char buf[2];
148 const char *cp = buf;
149 size_t needed = 2;
150
151 switch (ch)
152 {
153 case 0xa7 ... 0xf7:
154 cp = __cns11643l1_from_ucs4_tab1[ch - 0xa7];
155 break;
156 case 0x2c7 ... 0x2d9:
157 cp = __cns11643l1_from_ucs4_tab2[ch - 0x2c7];
158 break;
159 case 0x391 ... 0x3c9:
160 cp = __cns11643l1_from_ucs4_tab3[ch - 0x391];
161 break;
162 case 0x2013 ... 0x203e:
163 cp = __cns11643l1_from_ucs4_tab4[ch - 0x2013];
164 break;
165 case 0x2103:
166 cp = "\x22\x6a";
167 break;
168 case 0x2105:
169 cp = "\x22\x22";
170 break;
171 case 0x2109:
172 cp = "\x22\x6b";
173 break;
174 case 0x2160 ... 0x2169:
175 buf[0] = '\x24';
176 buf[1] = '\x2b' + (ch - 0x2160);
177 break;
178 case 0x2170 ... 0x2179:
179 buf[0] = '\x26';
180 buf[1] = '\x35' + (ch - 0x2170);
181 break;
182 case 0x2190 ... 0x2199:
183 cp = __cns11643l1_from_ucs4_tab5[ch - 0x2190];
184 break;
185 case 0x2215 ... 0x2267:
186 cp = __cns11643l1_from_ucs4_tab6[ch - 0x2215];
187 break;
188 case 0x22a5:
189 cp = "\x22\x47";
190 break;
191 case 0x22bf:
192 cp = "\x22\x4a";
193 break;
194 case 0x2400 ... 0x2421:
195 cp = __cns11643l1_from_ucs4_tab7[ch - 0x2400];
196 break;
197 case 0x2460 ... 0x247d:
198 cp = __cns11643l1_from_ucs4_tab8[ch - 0x2460];
199 break;
200 case 0x2500 ... 0x2642:
201 cp = __cns11643l1_from_ucs4_tab9[ch - 0x2500];
202 break;
203 case 0x3000 ... 0x3029:
204 cp = __cns11643l1_from_ucs4_tab10[ch - 0x3000];
205 break;
206 case 0x30fb:
207 cp = "\x21\x26";
208 break;
209 case 0x3105 ... 0x3129:
210 buf[0] = '\x25';
211 buf[1] = '\x47' + (ch - 0x3105);
212 break;
213 case 0x32a3:
214 cp = "\x22\x21";
215 break;
216 case 0x338e ... 0x33d5:
217 cp = __cns11643l1_from_ucs4_tab11[ch - 0x338e];
218 break;
219 case 0x4e00 ... 0x9f9c:
220 cp = __cns11643l1_from_ucs4_tab12[ch - 0x4e00];
221 if (cp[0] != '\0')
222 break;
223 /* Let's try the other planes. */
224 /* Fall through. */
225 case 0x3400 ... 0x4dff:
226 case 0x9f9d ... 0x9fa5:
227 /* Let's try the other planes. */
228 needed = 3;
229 cp = __cns11643_from_ucs4p0_tab[ch - 0x3400];
230 break;
231 case 0xfa28:
232 needed = 3;
233 cp = "\x0f\x58\x4c";
234 break;
235 case 0xfe30 ... 0xfe6b:
236 cp = __cns11643l1_from_ucs4_tab13[ch - 0xfe30];
237 break;
238 case 0xff01 ... 0xff5d:
239 cp = __cns11643l1_from_ucs4_tab14[ch - 0xff01];
240 break;
241 case 0xffe0:
242 cp = "\x22\x66";
243 break;
244 case 0xffe1:
245 cp = "\x22\x67";
246 break;
247 case 0xffe5:
248 cp = "\x22\x64";
249 break;
250 case 0x20000 ... 0x2a6d6:
251 needed = 3;
252 cp = __cns11643_from_ucs4p2_tab[ch - 0x20000];
253 break;
254 case 0x2f800 ... 0x2fa1d:
255 needed = 3;
256 cp = __cns11643_from_ucs4p2c_tab[ch - 0x2f800];
257 break;
258 default:
259 return __UNKNOWN_10646_CHAR;
260 }
261
262 if (cp[0] == '\0')
263 return __UNKNOWN_10646_CHAR;
264
265 if (avail < needed)
266 return 0;
267
268 s[0] = cp[0];
269 s[1] = cp[1];
270 if (needed == 3)
271 s[2] = cp[2];
272
273 return needed;
274}
275