1 /*
2 * This is an implementation of wcwidth() and wcswidth() (defined in
3 * IEEE Std 1002.1-2001) for Unicode.
4 *
5 * http://www.opengroup.org/onlinepubs/007904975/functions/wcwidth.html
6 * http://www.opengroup.org/onlinepubs/007904975/functions/wcswidth.html
7 *
8 * In fixed-width output devices, Latin characters all occupy a single
9 * "cell" position of equal width, whereas ideographic CJK characters
10 * occupy two such cells. Interoperability between terminal-line
11 * applications and (teletype-style) character terminals using the
12 * UTF-8 encoding requires agreement on which character should advance
13 * the cursor by how many cell positions. No established formal
14 * standards exist at present on which Unicode character shall occupy
15 * how many cell positions on character terminals. These routines are
16 * a first attempt of defining such behavior based on simple rules
17 * applied to data provided by the Unicode Consortium.
18 *
19 * For some graphical characters, the Unicode standard explicitly
20 * defines a character-cell width via the definition of the East Asian
21 * FullWidth (F), Wide (W), Half-width (H), and Narrow (Na) classes.
22 * In all these cases, there is no ambiguity about which width a
23 * terminal shall use. For characters in the East Asian Ambiguous (A)
24 * class, the width choice depends purely on a preference of backward
25 * compatibility with either historic CJK or Western practice.
26 * Choosing single-width for these characters is easy to justify as
27 * the appropriate long-term solution, as the CJK practice of
28 * displaying these characters as double-width comes from historic
29 * implementation simplicity (8-bit encoded characters were displayed
30 * single-width and 16-bit ones double-width, even for Greek,
31 * Cyrillic, etc.) and not any typographic considerations.
32 *
33 * Much less clear is the choice of width for the Not East Asian
34 * (Neutral) class. Existing practice does not dictate a width for any
35 * of these characters. It would nevertheless make sense
36 * typographically to allocate two character cells to characters such
37 * as for instance EM SPACE or VOLUME INTEGRAL, which cannot be
38 * represented adequately with a single-width glyph. The following
39 * routines at present merely assign a single-cell width to all
40 * neutral characters, in the interest of simplicity. This is not
41 * entirely satisfactory and should be reconsidered before
42 * establishing a formal standard in this area. At the moment, the
43 * decision which Not East Asian (Neutral) characters should be
44 * represented by double-width glyphs cannot yet be answered by
45 * applying a simple rule from the Unicode database content. Setting
46 * up a proper standard for the behavior of UTF-8 character terminals
47 * will require a careful analysis not only of each Unicode character,
48 * but also of each presentation form, something the author of these
49 * routines has avoided to do so far.
50 *
51 * http://www.unicode.org/unicode/reports/tr11/
52 *
53 * Markus Kuhn -- 2007-05-26 (Unicode 5.0)
54 *
55 * Permission to use, copy, modify, and distribute this software
56 * for any purpose and without fee is hereby granted. The author
57 * disclaims all warranties with regard to this software.
58 *
59 * Latest version: http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
60 */
61
62 #include <cmake.h>
63 #include <wchar.h>
64
65 struct interval {
66 int first;
67 int last;
68 };
69
70 /* auxiliary function for binary search in interval table */
bisearch(wchar_t ucs,const struct interval * table,int max)71 static int bisearch(wchar_t ucs, const struct interval *table, int max) {
72 int min = 0;
73 int mid;
74
75 if (ucs < table[0].first || ucs > table[max].last)
76 return 0;
77 while (max >= min) {
78 mid = (min + max) / 2;
79 if (ucs > table[mid].last)
80 min = mid + 1;
81 else if (ucs < table[mid].first)
82 max = mid - 1;
83 else
84 return 1;
85 }
86
87 return 0;
88 }
89
90
91 /* The following two functions define the column width of an ISO 10646
92 * character as follows:
93 *
94 * - The null character (U+0000) has a column width of 0.
95 *
96 * - Other C0/C1 control characters and DEL will lead to a return
97 * value of -1.
98 *
99 * - Non-spacing and enclosing combining characters (general
100 * category code Mn or Me in the Unicode database) have a
101 * column width of 0.
102 *
103 * - SOFT HYPHEN (U+00AD) has a column width of 1.
104 *
105 * - Other format characters (general category code Cf in the Unicode
106 * database) and ZERO WIDTH SPACE (U+200B) have a column width of 0.
107 *
108 * - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF)
109 * have a column width of 0.
110 *
111 * - Spacing characters in the East Asian Wide (W) or East Asian
112 * Full-width (F) category as defined in Unicode Technical
113 * Report #11 have a column width of 2.
114 *
115 * - All remaining characters (including all printable
116 * ISO 8859-1 and WGL4 characters, Unicode control characters,
117 * etc.) have a column width of 1.
118 *
119 * This implementation assumes that wchar_t characters are encoded
120 * in ISO 10646.
121 */
122
mk_wcwidth(wchar_t ucs)123 int mk_wcwidth(wchar_t ucs)
124 {
125 /* sorted list of non-overlapping intervals of non-spacing characters */
126 /* generated by "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c" */
127 /* "uniset cat:Me + cat:Mn + cat:Cf - U+00AD + U+1160..U+11FF + U+200B" */
128 static const struct interval combining[] = {
129 { 0x0300, 0x036F }, { 0x0483, 0x0489 }, { 0x0591, 0x05BD },
130 { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 }, { 0x05C4, 0x05C5 },
131 { 0x05C7, 0x05C7 }, { 0x0600, 0x0604 }, { 0x0610, 0x061A },
132 { 0x064B, 0x065F }, { 0x0670, 0x0670 }, { 0x06D6, 0x06DD },
133 { 0x06DF, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
134 { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
135 { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0816, 0x0819 },
136 { 0x081B, 0x0823 }, { 0x0825, 0x0827 }, { 0x0829, 0x082D },
137 { 0x0859, 0x085B }, { 0x08E4, 0x08FE }, { 0x0900, 0x0902 },
138 { 0x093A, 0x093A }, { 0x093C, 0x093C }, { 0x0941, 0x0948 },
139 { 0x094D, 0x094D }, { 0x0951, 0x0957 }, { 0x0962, 0x0963 },
140 { 0x0981, 0x0981 }, { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 },
141 { 0x09CD, 0x09CD }, { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 },
142 { 0x0A3C, 0x0A3C }, { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 },
143 { 0x0A4B, 0x0A4D }, { 0x0A51, 0x0A51 }, { 0x0A70, 0x0A71 },
144 { 0x0A75, 0x0A75 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC },
145 { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD },
146 { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C },
147 { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B44 }, { 0x0B4D, 0x0B4D },
148 { 0x0B56, 0x0B56 }, { 0x0B62, 0x0B63 }, { 0x0B82, 0x0B82 },
149 { 0x0BC0, 0x0BC0 }, { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 },
150 { 0x0C46, 0x0C48 }, { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 },
151 { 0x0C62, 0x0C63 }, { 0x0CBC, 0x0CBC }, { 0x0CBF, 0x0CBF },
152 { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD }, { 0x0CE2, 0x0CE3 },
153 { 0x0D41, 0x0D44 }, { 0x0D4D, 0x0D4D }, { 0x0D62, 0x0D63 },
154 { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
155 { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
156 { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
157 { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
158 { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
159 { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F8D, 0x0F97 },
160 { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
161 { 0x1032, 0x1037 }, { 0x1039, 0x103A }, { 0x103D, 0x103E },
162 { 0x1058, 0x1059 }, { 0x105E, 0x1060 }, { 0x1071, 0x1074 },
163 { 0x1082, 0x1082 }, { 0x1085, 0x1086 }, { 0x108D, 0x108D },
164 { 0x109D, 0x109D }, { 0x1160, 0x11FF }, { 0x135D, 0x135F },
165 { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 },
166 { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD },
167 { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD },
168 { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 },
169 { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B },
170 { 0x1A17, 0x1A18 }, { 0x1A56, 0x1A56 }, { 0x1A58, 0x1A5E },
171 { 0x1A60, 0x1A60 }, { 0x1A62, 0x1A62 }, { 0x1A65, 0x1A6C },
172 { 0x1A73, 0x1A7C }, { 0x1A7F, 0x1A7F }, { 0x1B00, 0x1B03 },
173 { 0x1B34, 0x1B34 }, { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C },
174 { 0x1B42, 0x1B42 }, { 0x1B6B, 0x1B73 }, { 0x1B80, 0x1B81 },
175 { 0x1BA2, 0x1BA5 }, { 0x1BA8, 0x1BA9 }, { 0x1BAB, 0x1BAB },
176 { 0x1BE6, 0x1BE6 }, { 0x1BE8, 0x1BE9 }, { 0x1BED, 0x1BED },
177 { 0x1BEF, 0x1BF1 }, { 0x1C2C, 0x1C33 }, { 0x1C36, 0x1C37 },
178 { 0x1CD0, 0x1CD2 }, { 0x1CD4, 0x1CE0 }, { 0x1CE2, 0x1CE8 },
179 { 0x1CED, 0x1CED }, { 0x1CF4, 0x1CF4 }, { 0x1DC0, 0x1DE6 },
180 { 0x1DFC, 0x1DFF }, { 0x200B, 0x200F }, { 0x202A, 0x202E },
181 { 0x2060, 0x2064 }, { 0x206A, 0x206F }, { 0x20D0, 0x20F0 },
182 { 0x2CEF, 0x2CF1 }, { 0x2D7F, 0x2D7F }, { 0x2DE0, 0x2DFF },
183 { 0x302A, 0x302D }, { 0x3099, 0x309A }, { 0xA66F, 0xA672 },
184 { 0xA674, 0xA67D }, { 0xA69F, 0xA69F }, { 0xA6F0, 0xA6F1 },
185 { 0xA802, 0xA802 }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B },
186 { 0xA825, 0xA826 }, { 0xA8C4, 0xA8C4 }, { 0xA8E0, 0xA8F1 },
187 { 0xA926, 0xA92D }, { 0xA947, 0xA951 }, { 0xA980, 0xA982 },
188 { 0xA9B3, 0xA9B3 }, { 0xA9B6, 0xA9B9 }, { 0xA9BC, 0xA9BC },
189 { 0xAA29, 0xAA2E }, { 0xAA31, 0xAA32 }, { 0xAA35, 0xAA36 },
190 { 0xAA43, 0xAA43 }, { 0xAA4C, 0xAA4C }, { 0xAAB0, 0xAAB0 },
191 { 0xAAB2, 0xAAB4 }, { 0xAAB7, 0xAAB8 }, { 0xAABE, 0xAABF },
192 { 0xAAC1, 0xAAC1 }, { 0xAAEC, 0xAAED }, { 0xAAF6, 0xAAF6 },
193 { 0xABE5, 0xABE5 }, { 0xABE8, 0xABE8 }, { 0xABED, 0xABED },
194 { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F }, { 0xFE20, 0xFE26 },
195 { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB }, { 0x101FD, 0x101FD },
196 { 0x10A01, 0x10A03 }, { 0x10A05, 0x10A06 }, { 0x10A0C, 0x10A0F },
197 { 0x10A38, 0x10A3A }, { 0x10A3F, 0x10A3F }, { 0x11001, 0x11001 },
198 { 0x11038, 0x11046 }, { 0x11080, 0x11081 }, { 0x110B3, 0x110B6 },
199 { 0x110B9, 0x110BA }, { 0x110BD, 0x110BD }, { 0x11100, 0x11102 },
200 { 0x11127, 0x1112B }, { 0x1112D, 0x11134 }, { 0x11180, 0x11181 },
201 { 0x111B6, 0x111BE }, { 0x116AB, 0x116AB }, { 0x116AD, 0x116AD },
202 { 0x116B0, 0x116B5 }, { 0x116B7, 0x116B7 }, { 0x16F8F, 0x16F92 },
203 { 0x1D167, 0x1D169 }, { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B },
204 { 0x1D1AA, 0x1D1AD }, { 0x1D242, 0x1D244 }, { 0xE0001, 0xE0001 },
205 { 0xE0020, 0xE007F }, { 0xE0100, 0xE01EF },
206 };
207
208 /* test for 8-bit control characters */
209 if (ucs == 0)
210 return 0;
211 if (ucs < 32 || (ucs >= 0x7f && ucs < 0xa0))
212 return -1;
213
214 /* binary search in table of non-spacing characters */
215 if (bisearch(ucs, combining,
216 sizeof(combining) / sizeof(struct interval) - 1))
217 return 0;
218
219 /* if we arrive here, ucs is not a combining or C0/C1 control character */
220
221 return 1 +
222 (ucs >= 0x1100 &&
223 (ucs <= 0x115f || /* Hangul Jamo init. consonants */
224 ucs == 0x2329 || ucs == 0x232a ||
225 (ucs >= 0x2e80 && ucs <= 0xa4cf &&
226 ucs != 0x303f) || /* CJK ... Yi */
227 (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
228 (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
229 (ucs >= 0xfe10 && ucs <= 0xfe19) || /* Vertical forms */
230 (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
231 (ucs >= 0xff00 && ucs <= 0xff60) || /* Fullwidth Forms */
232 (ucs >= 0xffe0 && ucs <= 0xffe6)
233 #ifndef CYGWIN
234 ||
235 (ucs >= 0x20000 && ucs <= 0x2fffd) ||
236 (ucs >= 0x30000 && ucs <= 0x3fffd)
237 #endif
238 )
239 );
240 }
241
mk_wcswidth(const wchar_t * pwcs,size_t n)242 int mk_wcswidth(const wchar_t *pwcs, size_t n)
243 {
244 int w, width = 0;
245
246 for (; *pwcs && n-- > 0; pwcs++)
247 if ((w = mk_wcwidth(*pwcs)) < 0)
248 return -1;
249 else
250 width += w;
251
252 return width;
253 }
254