| blob | f23d7ac9c0d73c7b795b81d362e96625a8c9784c |
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 */
62 #include <wchar.h>
67 };
69 /* auxiliary function for binary search in interval table */
71 {
83 else
85 }
88 }
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 */
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" */
169 };
171 /* test for 8-bit control characters */
177 /* binary search in table of non-spacing characters */
182 /* if we arrive here, ucs is not a combining or C0/C1 control character */
185 ucs == 0x2329 || ucs == 0x232a || (ucs >= 0x2e80 && ucs <= 0xa4cf && ucs != 0x303f) || /* CJK ... Yi */
193 }
197 {
203 else
207 }
210 /*
211 * The following functions are the same as mk_wcwidth() and
212 * mk_wcwidth_cjk(), except that spacing characters in the East Asian
213 * Ambiguous (A) category as defined in Unicode Technical Report #11
214 * have a column width of 2. This variant might be useful for users of
215 * CJK legacy encodings who want to migrate to UCS without changing
216 * the traditional terminal character-width behaviour. It is not
217 * otherwise recommended for general use.
218 */
220 {
221 /* sorted list of non-overlapping intervals of East Asian Ambiguous
222 * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
276 };
278 /* binary search in table of non-spacing characters */
284 }
288 {
294 else
298 }