RDB's Unicode patch. Fonts are now used in Unicode mode where
[u/mdw/putty] / wcwidth.c
CommitLineData
4eeb7d09 1/*
2 * This is an implementation of wcwidth() and wcswidth() as defined in
3 * "The Single UNIX Specification, Version 2, The Open Group, 1997"
4 * <http://www.UNIX-systems.org/online.html>
5 *
6 * Markus Kuhn -- 2000-02-08 -- public domain
7 */
8
9#include <wchar.h>
10
11/* These functions define the column width of an ISO 10646 character
12 * as follows:
13 *
14 * - The null character (U+0000) has a column width of 0.
15 *
16 * - Other C0/C1 control characters and DEL will lead to a return
17 * value of -1.
18 *
19 * - Non-spacing and enclosing combining characters (general
20 * category code Mn or Me in the Unicode database) have a
21 * column width of 0.
22 *
23 * - Spacing characters in the East Asian Wide (W) or East Asian
24 * FullWidth (F) category as defined in Unicode Technical
25 * Report #11 have a column width of 2.
26 *
27 * - All remaining characters (including all printable
28 * ISO 8859-1 and WGL4 characters, Unicode control characters,
29 * etc.) have a column width of 1.
30 *
31 * This implementation assumes that wchar_t characters are encoded
32 * in ISO 10646.
33 */
34
35int wcwidth(wchar_t ucs)
36{
37 /* sorted list of non-overlapping intervals of non-spacing characters */
38 static const struct interval {
39 unsigned short first;
40 unsigned short last;
41 } combining[] = {
42 {
43 0x0300, 0x034E}, {
44 0x0360, 0x0362}, {
45 0x0483, 0x0486}, {
46 0x0488, 0x0489}, {
47 0x0591, 0x05A1}, {
48 0x05A3, 0x05B9}, {
49 0x05BB, 0x05BD}, {
50 0x05BF, 0x05BF}, {
51 0x05C1, 0x05C2}, {
52 0x05C4, 0x05C4}, {
53 0x064B, 0x0655}, {
54 0x0670, 0x0670}, {
55 0x06D6, 0x06E4}, {
56 0x06E7, 0x06E8}, {
57 0x06EA, 0x06ED}, {
58 0x0711, 0x0711}, {
59 0x0730, 0x074A}, {
60 0x07A6, 0x07B0}, {
61 0x0901, 0x0902}, {
62 0x093C, 0x093C}, {
63 0x0941, 0x0948}, {
64 0x094D, 0x094D}, {
65 0x0951, 0x0954}, {
66 0x0962, 0x0963}, {
67 0x0981, 0x0981}, {
68 0x09BC, 0x09BC}, {
69 0x09C1, 0x09C4}, {
70 0x09CD, 0x09CD}, {
71 0x09E2, 0x09E3}, {
72 0x0A02, 0x0A02}, {
73 0x0A3C, 0x0A3C}, {
74 0x0A41, 0x0A42}, {
75 0x0A47, 0x0A48}, {
76 0x0A4B, 0x0A4D}, {
77 0x0A70, 0x0A71}, {
78 0x0A81, 0x0A82}, {
79 0x0ABC, 0x0ABC}, {
80 0x0AC1, 0x0AC5}, {
81 0x0AC7, 0x0AC8}, {
82 0x0ACD, 0x0ACD}, {
83 0x0B01, 0x0B01}, {
84 0x0B3C, 0x0B3C}, {
85 0x0B3F, 0x0B3F}, {
86 0x0B41, 0x0B43}, {
87 0x0B4D, 0x0B4D}, {
88 0x0B56, 0x0B56}, {
89 0x0B82, 0x0B82}, {
90 0x0BC0, 0x0BC0}, {
91 0x0BCD, 0x0BCD}, {
92 0x0C3E, 0x0C40}, {
93 0x0C46, 0x0C48}, {
94 0x0C4A, 0x0C4D}, {
95 0x0C55, 0x0C56}, {
96 0x0CBF, 0x0CBF}, {
97 0x0CC6, 0x0CC6}, {
98 0x0CCC, 0x0CCD}, {
99 0x0D41, 0x0D43}, {
100 0x0D4D, 0x0D4D}, {
101 0x0DCA, 0x0DCA}, {
102 0x0DD2, 0x0DD4}, {
103 0x0DD6, 0x0DD6}, {
104 0x0E31, 0x0E31}, {
105 0x0E34, 0x0E3A}, {
106 0x0E47, 0x0E4E}, {
107 0x0EB1, 0x0EB1}, {
108 0x0EB4, 0x0EB9}, {
109 0x0EBB, 0x0EBC}, {
110 0x0EC8, 0x0ECD}, {
111 0x0F18, 0x0F19}, {
112 0x0F35, 0x0F35}, {
113 0x0F37, 0x0F37}, {
114 0x0F39, 0x0F39}, {
115 0x0F71, 0x0F7E}, {
116 0x0F80, 0x0F84}, {
117 0x0F86, 0x0F87}, {
118 0x0F90, 0x0F97}, {
119 0x0F99, 0x0FBC}, {
120 0x0FC6, 0x0FC6}, {
121 0x102D, 0x1030}, {
122 0x1032, 0x1032}, {
123 0x1036, 0x1037}, {
124 0x1039, 0x1039}, {
125 0x1058, 0x1059}, {
126 0x17B7, 0x17BD}, {
127 0x17C6, 0x17C6}, {
128 0x17C9, 0x17D3}, {
129 0x18A9, 0x18A9}, {
130 0x20D0, 0x20E3}, {
131 0x302A, 0x302F}, {
132 0x3099, 0x309A}, {
133 0xFB1E, 0xFB1E}, {
134 0xFE20, 0xFE23}
135 };
136 int min = 0;
137 int max = sizeof(combining) / sizeof(struct interval) - 1;
138 int mid;
139
140 /* test for 8-bit control characters */
141 if (ucs == 0)
142 return 0;
143 if (ucs < 32 || (ucs >= 0x7f && ucs < 0xa0))
144 return -1;
145
146 /* first quick check for Latin-1 etc. characters */
147 if (ucs < combining[0].first)
148 return 1;
149
150 /* binary search in table of non-spacing characters */
151 while (max >= min) {
152 mid = (min + max) / 2;
153 if (combining[mid].last < ucs)
154 min = mid + 1;
155 else if (combining[mid].first > ucs)
156 max = mid - 1;
157 else if (combining[mid].first <= ucs && combining[mid].last >= ucs)
158 return 0;
159 }
160
161 /* if we arrive here, ucs is not a combining or C0/C1 control character */
162
163 /* fast test for majority of non-wide scripts */
164 if (ucs < 0x1100)
165 return 1;
166
167 return 1 + ((ucs >= 0x1100 && ucs <= 0x115f) || /* Hangul Jamo */
168
169 (ucs >= 0x2e80 && ucs <= 0xa4cf
170 && (ucs & ~0x0011) != 0x300a && ucs != 0x303f) || /* CJK ... Yi */
171 (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
172 (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
173 (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
174 (ucs >= 0xff00 && ucs <= 0xff5f) || /* Fullwidth Forms */
175 (ucs >= 0xffe0 && ucs <= 0xffe6));
176}
177
178
179int wcswidth(const wchar_t * pwcs, size_t n)
180{
181 int w, width = 0;
182
183 for (; *pwcs && n-- > 0; pwcs++)
184 if ((w = wcwidth(*pwcs)) < 0)
185 return -1;
186 else
187 width += w;
188
189 return width;
190}