1 /* Access functions for KS C 5601-1992 based encoding conversion.
2 Copyright (C) 1998-2022 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 #ifndef _KSC5601_H
20 #define _KSC5601_H 1
21
22 #define KSC5601_HANGUL 2350
23 #define KSC5601_HANJA 4888
24 #define KSC5601_SYMBOL 989
25
26 #include <gconv.h>
27 #include <stdint.h>
28
29 /* Structure to map from UCS to KSC. This structure should be packed
30 on all platforms. */
31 struct map
32 {
33 uint16_t ucs;
34 char val[2];
35 };
36
37 /* Conversion table. */
38 extern const uint16_t __ksc5601_hangul_to_ucs[KSC5601_HANGUL];
39 extern const uint16_t __ksc5601_sym_to_ucs[];
40 extern const struct map __ksc5601_sym_from_ucs[KSC5601_SYMBOL];
41 extern const uint16_t __ksc5601_hanja_to_ucs[KSC5601_HANJA];
42 extern const struct map __ksc5601_hanja_from_ucs[KSC5601_HANJA];
43
44
uint32_t(always_inline)45 static inline uint32_t
46 __attribute ((always_inline))
47 ksc5601_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
48 {
49 unsigned char ch = **s;
50 unsigned char ch2;
51 int idx;
52
53 if (avail < 2)
54 return 0;
55
56 /* row 94(0x7e) and row 41(0x49) are user-defined area in KS C 5601 */
57
58 if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) >= 0x7e
59 || (ch - offset) == 0x49)
60 return __UNKNOWN_10646_CHAR;
61
62 ch2 = (*s)[1];
63 if (ch2 < offset || (ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
64 return __UNKNOWN_10646_CHAR;
65
66 idx = (ch - offset - 0x21) * 94 + (ch2 - offset - 0x21);
67
68 /* 1410 = 15 * 94 , 3760 = 40 * 94
69 Hangul in KS C 5601 : row 16 - row 40 */
70
71 *s += 2;
72
73 if (idx >= 1410 && idx < 1410 + KSC5601_HANGUL)
74 return (__ksc5601_hangul_to_ucs[idx - 1410]
75 ?: (*s -= 2, __UNKNOWN_10646_CHAR));
76 else if (idx >= 3854)
77 /* Hanja : row 42 - row 93 : 3854 = 94 * (42-1) */
78 return (__ksc5601_hanja_to_ucs[idx - 3854]
79 ?: (*s -= 2, __UNKNOWN_10646_CHAR));
80 else if (idx <= 1114)
81 return __ksc5601_sym_to_ucs[idx] ?: (*s -= 2, __UNKNOWN_10646_CHAR);
82
83 *s -= 2;
84 return __UNKNOWN_10646_CHAR;
85 }
86
size_t(always_inline)87 static inline size_t
88 __attribute ((always_inline))
89 ucs4_to_ksc5601_hangul (uint32_t wch, unsigned char *s, size_t avail)
90 {
91 int l = 0;
92 int u = KSC5601_HANGUL - 1;
93 uint32_t try;
94
95 while (l <= u)
96 {
97 int m = (l + u) / 2;
98 try = (uint32_t) __ksc5601_hangul_to_ucs[m];
99 if (try > wch)
100 u = m - 1;
101 else if (try < wch)
102 l= m + 1;
103 else
104 {
105 if (avail < 2)
106 return 0;
107
108 s[0] = (m / 94) + 0x30;
109 s[1] = (m % 94) + 0x21;
110
111 return 2;
112 }
113 }
114
115 return __UNKNOWN_10646_CHAR;
116 }
117
118
size_t(always_inline)119 static inline size_t
120 __attribute ((always_inline))
121 ucs4_to_ksc5601_hanja (uint32_t wch, unsigned char *s, size_t avail)
122 {
123 int l = 0;
124 int u = KSC5601_HANJA - 1;
125 uint32_t try;
126
127 while (l <= u)
128 {
129 int m = (l + u) / 2;
130 try = (uint32_t) __ksc5601_hanja_from_ucs[m].ucs;
131 if (try > wch)
132 u=m-1;
133 else if (try < wch)
134 l = m + 1;
135 else
136 {
137 if (avail < 2)
138 return 0;
139
140 s[0] = __ksc5601_hanja_from_ucs[m].val[0];
141 s[1] = __ksc5601_hanja_from_ucs[m].val[1];
142
143 return 2;
144 }
145 }
146
147 return __UNKNOWN_10646_CHAR;
148 }
149
size_t(always_inline)150 static inline size_t
151 __attribute ((always_inline))
152 ucs4_to_ksc5601_sym (uint32_t wch, unsigned char *s, size_t avail)
153 {
154 int l = 0;
155 int u = KSC5601_SYMBOL - 1;
156 uint32_t try;
157
158 while (l <= u)
159 {
160 int m = (l + u) / 2;
161 try = __ksc5601_sym_from_ucs[m].ucs;
162 if (try > wch)
163 u = m - 1;
164 else if (try < wch)
165 l = m + 1;
166 else
167 {
168 if (avail < 2)
169 return 0;
170
171 s[0] = __ksc5601_sym_from_ucs[m].val[0];
172 s[1] = __ksc5601_sym_from_ucs[m].val[1];
173
174 return 2;
175 }
176 }
177
178 return __UNKNOWN_10646_CHAR;
179 }
180
181
size_t(always_inline)182 static inline size_t
183 __attribute ((always_inline))
184 ucs4_to_ksc5601 (uint32_t wch, unsigned char *s, size_t avail)
185 {
186 if (wch >= 0xac00 && wch <= 0xd7a3)
187 return ucs4_to_ksc5601_hangul (wch, s, avail);
188 else if ((wch >= 0x4e00 && wch <= 0x9fff)
189 || (wch >= 0xf900 && wch <= 0xfa0b))
190 return ucs4_to_ksc5601_hanja (wch, s, avail);
191 else
192 return ucs4_to_ksc5601_sym (wch, s, avail);
193 }
194
195 #endif /* ksc5601.h */
196