]>
Commit | Line | Data |
---|---|---|
1c1af145 | 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 -- 2003-05-20 (Unicode 4.0) | |
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 <wchar.h> | |
63 | ||
64 | #include "putty.h" /* for prototypes */ | |
65 | ||
66 | struct interval { | |
67 | int first; | |
68 | int last; | |
69 | }; | |
70 | ||
71 | /* auxiliary function for binary search in interval table */ | |
72 | static int bisearch(wchar_t ucs, const struct interval *table, int max) { | |
73 | int min = 0; | |
74 | int mid; | |
75 | ||
76 | if (ucs < table[0].first || ucs > table[max].last) | |
77 | return 0; | |
78 | while (max >= min) { | |
79 | mid = (min + max) / 2; | |
80 | if (ucs > table[mid].last) | |
81 | min = mid + 1; | |
82 | else if (ucs < table[mid].first) | |
83 | max = mid - 1; | |
84 | else | |
85 | return 1; | |
86 | } | |
87 | ||
88 | return 0; | |
89 | } | |
90 | ||
91 | ||
92 | /* The following two functions define the column width of an ISO 10646 | |
93 | * character as follows: | |
94 | * | |
95 | * - The null character (U+0000) has a column width of 0. | |
96 | * | |
97 | * - Other C0/C1 control characters and DEL will lead to a return | |
98 | * value of -1. | |
99 | * | |
100 | * - Non-spacing and enclosing combining characters (general | |
101 | * category code Mn or Me in the Unicode database) have a | |
102 | * column width of 0. | |
103 | * | |
104 | * - SOFT HYPHEN (U+00AD) has a column width of 1. | |
105 | * | |
106 | * - Other format characters (general category code Cf in the Unicode | |
107 | * database) and ZERO WIDTH SPACE (U+200B) have a column width of 0. | |
108 | * | |
109 | * - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF) | |
110 | * have a column width of 0. | |
111 | * | |
112 | * - Spacing characters in the East Asian Wide (W) or East Asian | |
113 | * Full-width (F) category as defined in Unicode Technical | |
114 | * Report #11 have a column width of 2. | |
115 | * | |
116 | * - All remaining characters (including all printable | |
117 | * ISO 8859-1 and WGL4 characters, Unicode control characters, | |
118 | * etc.) have a column width of 1. | |
119 | * | |
120 | * This implementation assumes that wchar_t characters are encoded | |
121 | * in ISO 10646. | |
122 | */ | |
123 | ||
124 | int mk_wcwidth(wchar_t ucs) | |
125 | { | |
126 | /* sorted list of non-overlapping intervals of non-spacing characters */ | |
127 | /* generated by "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c" */ | |
128 | static const struct interval combining[] = { | |
129 | { 0x0300, 0x0357 }, { 0x035D, 0x036F }, { 0x0483, 0x0486 }, | |
130 | { 0x0488, 0x0489 }, { 0x0591, 0x05A1 }, { 0x05A3, 0x05B9 }, | |
131 | { 0x05BB, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 }, | |
132 | { 0x05C4, 0x05C4 }, { 0x0600, 0x0603 }, { 0x0610, 0x0615 }, | |
133 | { 0x064B, 0x0658 }, { 0x0670, 0x0670 }, { 0x06D6, 0x06E4 }, | |
134 | { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED }, { 0x070F, 0x070F }, | |
135 | { 0x0711, 0x0711 }, { 0x0730, 0x074A }, { 0x07A6, 0x07B0 }, | |
136 | { 0x0901, 0x0902 }, { 0x093C, 0x093C }, { 0x0941, 0x0948 }, | |
137 | { 0x094D, 0x094D }, { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, | |
138 | { 0x0981, 0x0981 }, { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, | |
139 | { 0x09CD, 0x09CD }, { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, | |
140 | { 0x0A3C, 0x0A3C }, { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, | |
141 | { 0x0A4B, 0x0A4D }, { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, | |
142 | { 0x0ABC, 0x0ABC }, { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, | |
143 | { 0x0ACD, 0x0ACD }, { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, | |
144 | { 0x0B3C, 0x0B3C }, { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, | |
145 | { 0x0B4D, 0x0B4D }, { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, | |
146 | { 0x0BC0, 0x0BC0 }, { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, | |
147 | { 0x0C46, 0x0C48 }, { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, | |
148 | { 0x0CBC, 0x0CBC }, { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, | |
149 | { 0x0CCC, 0x0CCD }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D }, | |
150 | { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 }, | |
151 | { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E }, | |
152 | { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC }, | |
153 | { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 }, | |
154 | { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E }, | |
155 | { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 }, | |
156 | { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 }, | |
157 | { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 }, | |
158 | { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x1712, 0x1714 }, | |
159 | { 0x1732, 0x1734 }, { 0x1752, 0x1753 }, { 0x1772, 0x1773 }, | |
160 | { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD }, { 0x17C6, 0x17C6 }, | |
161 | { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD }, { 0x180B, 0x180D }, | |
162 | { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 }, { 0x1927, 0x1928 }, | |
163 | { 0x1932, 0x1932 }, { 0x1939, 0x193B }, { 0x200B, 0x200F }, | |
164 | { 0x202A, 0x202E }, { 0x2060, 0x2063 }, { 0x206A, 0x206F }, | |
165 | { 0x20D0, 0x20EA }, { 0x302A, 0x302F }, { 0x3099, 0x309A }, | |
166 | { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F }, { 0xFE20, 0xFE23 }, | |
167 | { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB }, { 0x1D167, 0x1D169 }, | |
168 | { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B }, { 0x1D1AA, 0x1D1AD }, | |
169 | { 0xE0001, 0xE0001 }, { 0xE0020, 0xE007F }, { 0xE0100, 0xE01EF } | |
170 | }; | |
171 | ||
172 | /* test for 8-bit control characters */ | |
173 | if (ucs == 0) | |
174 | return 0; | |
175 | if (ucs < 32 || (ucs >= 0x7f && ucs < 0xa0)) | |
176 | return -1; | |
177 | ||
178 | /* binary search in table of non-spacing characters */ | |
179 | if (bisearch(ucs, combining, | |
180 | sizeof(combining) / sizeof(struct interval) - 1)) | |
181 | return 0; | |
182 | ||
183 | /* if we arrive here, ucs is not a combining or C0/C1 control character */ | |
184 | ||
185 | return 1 + | |
186 | (ucs >= 0x1100 && | |
187 | (ucs <= 0x115f || /* Hangul Jamo init. consonants */ | |
188 | ucs == 0x2329 || ucs == 0x232a || | |
189 | (ucs >= 0x2e80 && ucs <= 0xa4cf && | |
190 | ucs != 0x303f) || /* CJK ... Yi */ | |
191 | (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */ | |
192 | (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */ | |
193 | (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */ | |
194 | (ucs >= 0xff00 && ucs <= 0xff60) || /* Fullwidth Forms */ | |
195 | (ucs >= 0xffe0 && ucs <= 0xffe6) || | |
196 | (ucs >= 0x20000 && ucs <= 0x2fffd) || | |
197 | (ucs >= 0x30000 && ucs <= 0x3fffd))); | |
198 | } | |
199 | ||
200 | ||
201 | int mk_wcswidth(const wchar_t *pwcs, size_t n) | |
202 | { | |
203 | int w, width = 0; | |
204 | ||
205 | for (;*pwcs && n-- > 0; pwcs++) | |
206 | if ((w = mk_wcwidth(*pwcs)) < 0) | |
207 | return -1; | |
208 | else | |
209 | width += w; | |
210 | ||
211 | return width; | |
212 | } | |
213 | ||
214 | ||
215 | /* | |
216 | * The following functions are the same as mk_wcwidth() and | |
217 | * mk_wcwidth_cjk(), except that spacing characters in the East Asian | |
218 | * Ambiguous (A) category as defined in Unicode Technical Report #11 | |
219 | * have a column width of 2. This variant might be useful for users of | |
220 | * CJK legacy encodings who want to migrate to UCS without changing | |
221 | * the traditional terminal character-width behaviour. It is not | |
222 | * otherwise recommended for general use. | |
223 | */ | |
224 | int mk_wcwidth_cjk(wchar_t ucs) | |
225 | { | |
226 | /* sorted list of non-overlapping intervals of East Asian Ambiguous | |
227 | * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */ | |
228 | static const struct interval ambiguous[] = { | |
229 | { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 }, | |
230 | { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 }, | |
231 | { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 }, | |
232 | { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 }, | |
233 | { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED }, | |
234 | { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA }, | |
235 | { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 }, | |
236 | { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B }, | |
237 | { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 }, | |
238 | { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 }, | |
239 | { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 }, | |
240 | { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE }, | |
241 | { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 }, | |
242 | { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA }, | |
243 | { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 }, | |
244 | { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB }, | |
245 | { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB }, | |
246 | { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 }, | |
247 | { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 }, | |
248 | { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 }, | |
249 | { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 }, | |
250 | { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 }, | |
251 | { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 }, | |
252 | { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 }, | |
253 | { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC }, | |
254 | { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 }, | |
255 | { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 }, | |
256 | { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 }, | |
257 | { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 }, | |
258 | { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 }, | |
259 | { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 }, | |
260 | { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B }, | |
261 | { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 }, | |
262 | { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 }, | |
263 | { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E }, | |
264 | { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 }, | |
265 | { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 }, | |
266 | { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F }, | |
267 | { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 }, | |
268 | { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF }, | |
269 | { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B }, | |
270 | { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 }, | |
271 | { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 }, | |
272 | { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 }, | |
273 | { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 }, | |
274 | { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 }, | |
275 | { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 }, | |
276 | { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 }, | |
277 | { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 }, | |
278 | { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F }, | |
279 | { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF }, | |
280 | { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD } | |
281 | }; | |
282 | ||
283 | /* binary search in table of non-spacing characters */ | |
284 | if (bisearch(ucs, ambiguous, | |
285 | sizeof(ambiguous) / sizeof(struct interval) - 1)) | |
286 | return 2; | |
287 | ||
288 | return mk_wcwidth(ucs); | |
289 | } | |
290 | ||
291 | ||
292 | int mk_wcswidth_cjk(const wchar_t *pwcs, size_t n) | |
293 | { | |
294 | int w, width = 0; | |
295 | ||
296 | for (;*pwcs && n-- > 0; pwcs++) | |
297 | if ((w = mk_wcwidth_cjk(*pwcs)) < 0) | |
298 | return -1; | |
299 | else | |
300 | width += w; | |
301 | ||
302 | return width; | |
303 | } |