1 /* Access functions for CNS 11643 handling.
2 Copyright (C) 1998-2023 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4
5 The GNU C Library is free software; you can redistribute it and/or
6 modify it under the terms of the GNU Lesser General Public
7 License as published by the Free Software Foundation; either
8 version 2.1 of the License, or (at your option) any later version.
9
10 The GNU C Library is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 Lesser General Public License for more details.
14
15 You should have received a copy of the GNU Lesser General Public
16 License along with the GNU C Library; if not, see
17 <https://www.gnu.org/licenses/>. */
18
19 #include <stdint.h>
20
21
22 /* Table for CNS 11643, plane 1 to UCS4 conversion. */
23 extern const uint16_t __cns11643l1_to_ucs4_tab[];
24 /* Table for CNS 11643, plane 2 to UCS4 conversion. */
25 extern const uint16_t __cns11643l2_to_ucs4_tab[];
26 /* Table for CNS 11643, plane 3 to UCS4 conversion. */
27 extern const uint32_t __cns11643l3_to_ucs4_tab[];
28 /* Table for CNS 11643, plane 4 to UCS4 conversion. */
29 extern const uint32_t __cns11643l4_to_ucs4_tab[];
30 /* Table for CNS 11643, plane 5 to UCS4 conversion. */
31 extern const uint32_t __cns11643l5_to_ucs4_tab[];
32 /* Table for CNS 11643, plane 6 to UCS4 conversion. */
33 extern const uint32_t __cns11643l6_to_ucs4_tab[];
34 /* Table for CNS 11643, plane 7 to UCS4 conversion. */
35 extern const uint32_t __cns11643l7_to_ucs4_tab[];
36 /* Table for CNS 11643, plane 15 (old) to UCS4 conversion. */
37 extern const uint32_t __cns11643l15_to_ucs4_tab[];
38
39
40 static inline uint32_t
41 __attribute ((always_inline))
42 cns11643_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
43 {
44 unsigned char ch = *(*s);
45 unsigned char ch2;
46 unsigned char ch3;
47 uint32_t result;
48 int idx;
49
50 if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) > 0x30)
51 return __UNKNOWN_10646_CHAR;
52
53 if (avail < 3)
54 return 0;
55
56 ch2 = (*s)[1];
57 if ((ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
58 return __UNKNOWN_10646_CHAR;
59
60 ch3 = (*s)[2];
61 if ((ch3 - offset) <= 0x20 || (ch3 - offset) >= 0x7f)
62 return __UNKNOWN_10646_CHAR;
63
64 idx = (ch2 - 0x21 - offset) * 94 + (ch3 - 0x21 - offset);
65
66 switch (ch - 0x20 - offset)
67 {
68 case 1:
69 if (idx > 0x21f2)
70 return __UNKNOWN_10646_CHAR;
71 result = __cns11643l1_to_ucs4_tab[idx];
72 break;
73 case 2:
74 if (idx > 0x1de1)
75 return __UNKNOWN_10646_CHAR;
76 result = __cns11643l2_to_ucs4_tab[idx];
77 break;
78 case 3:
79 if (idx > 0x19bd)
80 return __UNKNOWN_10646_CHAR;
81 result = __cns11643l3_to_ucs4_tab[idx];
82 break;
83 case 4:
84 if (idx > 0x1c81)
85 return __UNKNOWN_10646_CHAR;
86 result = __cns11643l4_to_ucs4_tab[idx];
87 break;
88 case 5:
89 if (idx > 0x219a)
90 return __UNKNOWN_10646_CHAR;
91 result = __cns11643l5_to_ucs4_tab[idx];
92 break;
93 case 6:
94 if (idx > 0x18f3)
95 return __UNKNOWN_10646_CHAR;
96 result = __cns11643l6_to_ucs4_tab[idx];
97 break;
98 case 7:
99 if (idx > 0x198a)
100 return __UNKNOWN_10646_CHAR;
101 result = __cns11643l7_to_ucs4_tab[idx];
102 break;
103 case 15:
104 if (idx > 0x1c00)
105 return __UNKNOWN_10646_CHAR;
106 result = __cns11643l15_to_ucs4_tab[idx];
107 break;
108 default:
109 return __UNKNOWN_10646_CHAR;
110 }
111
112 if (result != L'\0')
113 (*s) += 3;
114 else
115 result = __UNKNOWN_10646_CHAR;
116
117 return result;
118 }
119
120
121 /* Tables for the UCS4 -> CNS conversion. */
122 extern const char __cns11643l1_from_ucs4_tab1[][2];
123 extern const char __cns11643l1_from_ucs4_tab2[][2];
124 extern const char __cns11643l1_from_ucs4_tab3[][2];
125 extern const char __cns11643l1_from_ucs4_tab4[][2];
126 extern const char __cns11643l1_from_ucs4_tab5[][2];
127 extern const char __cns11643l1_from_ucs4_tab6[][2];
128 extern const char __cns11643l1_from_ucs4_tab7[][2];
129 extern const char __cns11643l1_from_ucs4_tab8[][2];
130 extern const char __cns11643l1_from_ucs4_tab9[][2];
131 extern const char __cns11643l1_from_ucs4_tab10[][2];
132 extern const char __cns11643l1_from_ucs4_tab11[][2];
133 extern const char __cns11643l1_from_ucs4_tab12[][2];
134 extern const char __cns11643l1_from_ucs4_tab13[][2];
135 extern const char __cns11643l1_from_ucs4_tab14[][2];
136 extern const char __cns11643_from_ucs4p0_tab[][3];
137 extern const char __cns11643_from_ucs4p2_tab[][3];
138 extern const char __cns11643_from_ucs4p2c_tab[][3];
139
140
141 static inline size_t
142 __attribute ((always_inline))
143 ucs4_to_cns11643 (uint32_t wch, unsigned char *s, size_t avail)
144 {
145 unsigned int ch = (unsigned int) wch;
146 char buf[2];
147 const char *cp = buf;
148 size_t needed = 2;
149
150 switch (ch)
151 {
152 case 0xa7 ... 0xf7:
153 cp = __cns11643l1_from_ucs4_tab1[ch - 0xa7];
154 break;
155 case 0x2c7 ... 0x2d9:
156 cp = __cns11643l1_from_ucs4_tab2[ch - 0x2c7];
157 break;
158 case 0x391 ... 0x3c9:
159 cp = __cns11643l1_from_ucs4_tab3[ch - 0x391];
160 break;
161 case 0x2013 ... 0x203e:
162 cp = __cns11643l1_from_ucs4_tab4[ch - 0x2013];
163 break;
164 case 0x2103:
165 cp = "\x22\x6a";
166 break;
167 case 0x2105:
168 cp = "\x22\x22";
169 break;
170 case 0x2109:
171 cp = "\x22\x6b";
172 break;
173 case 0x2160 ... 0x2169:
174 buf[0] = '\x24';
175 buf[1] = '\x2b' + (ch - 0x2160);
176 break;
177 case 0x2170 ... 0x2179:
178 buf[0] = '\x26';
179 buf[1] = '\x35' + (ch - 0x2170);
180 break;
181 case 0x2190 ... 0x2199:
182 cp = __cns11643l1_from_ucs4_tab5[ch - 0x2190];
183 break;
184 case 0x2215 ... 0x2267:
185 cp = __cns11643l1_from_ucs4_tab6[ch - 0x2215];
186 break;
187 case 0x22a5:
188 cp = "\x22\x47";
189 break;
190 case 0x22bf:
191 cp = "\x22\x4a";
192 break;
193 case 0x2400 ... 0x2421:
194 cp = __cns11643l1_from_ucs4_tab7[ch - 0x2400];
195 break;
196 case 0x2460 ... 0x247d:
197 cp = __cns11643l1_from_ucs4_tab8[ch - 0x2460];
198 break;
199 case 0x2500 ... 0x2642:
200 cp = __cns11643l1_from_ucs4_tab9[ch - 0x2500];
201 break;
202 case 0x3000 ... 0x3029:
203 cp = __cns11643l1_from_ucs4_tab10[ch - 0x3000];
204 break;
205 case 0x30fb:
206 cp = "\x21\x26";
207 break;
208 case 0x3105 ... 0x3129:
209 buf[0] = '\x25';
210 buf[1] = '\x47' + (ch - 0x3105);
211 break;
212 case 0x32a3:
213 cp = "\x22\x21";
214 break;
215 case 0x338e ... 0x33d5:
216 cp = __cns11643l1_from_ucs4_tab11[ch - 0x338e];
217 break;
218 case 0x4e00 ... 0x9f9c:
219 cp = __cns11643l1_from_ucs4_tab12[ch - 0x4e00];
220 if (cp[0] != '\0')
221 break;
222 /* Let's try the other planes. */
223 /* Fall through. */
224 case 0x3400 ... 0x4dff:
225 case 0x9f9d ... 0x9fa5:
226 /* Let's try the other planes. */
227 needed = 3;
228 cp = __cns11643_from_ucs4p0_tab[ch - 0x3400];
229 break;
230 case 0xfa28:
231 needed = 3;
232 cp = "\x0f\x58\x4c";
233 break;
234 case 0xfe30 ... 0xfe6b:
235 cp = __cns11643l1_from_ucs4_tab13[ch - 0xfe30];
236 break;
237 case 0xff01 ... 0xff5d:
238 cp = __cns11643l1_from_ucs4_tab14[ch - 0xff01];
239 break;
240 case 0xffe0:
241 cp = "\x22\x66";
242 break;
243 case 0xffe1:
244 cp = "\x22\x67";
245 break;
246 case 0xffe5:
247 cp = "\x22\x64";
248 break;
249 case 0x20000 ... 0x2a6d6:
250 needed = 3;
251 cp = __cns11643_from_ucs4p2_tab[ch - 0x20000];
252 break;
253 case 0x2f800 ... 0x2fa1d:
254 needed = 3;
255 cp = __cns11643_from_ucs4p2c_tab[ch - 0x2f800];
256 break;
257 default:
258 return __UNKNOWN_10646_CHAR;
259 }
260
261 if (cp[0] == '\0')
262 return __UNKNOWN_10646_CHAR;
263
264 if (avail < needed)
265 return 0;
266
267 s[0] = cp[0];
268 s[1] = cp[1];
269 if (needed == 3)
270 s[2] = cp[2];
271
272 return needed;
273 }