1 /* Copyright (c) 2003 Corinna Vinschen <corinna@vinschen.de> */
2 /*
3 FUNCTION
4 <<wcwidth>>---number of column positions of a wide-character code
5
6 INDEX
7 wcwidth
8
9 SYNOPSIS
10 #include <wchar.h>
11 int wcwidth(const wchar_t <[wc]>);
12
13 DESCRIPTION
14 The <<wcwidth>> function shall determine the number of column
15 positions required for the wide character <[wc]>. The application
16 shall ensure that the value of <[wc]> is a character representable
17 as a wint_t (combining Unicode surrogate pairs into single 21-bit
18 Unicode code points), and is a wide-character code corresponding to a
19 valid character in the current locale.
20
21 RETURNS
22 The <<wcwidth>> function shall either return 0 (if <[wc]> is a null
23 wide-character code), or return the number of column positions to
24 be occupied by the wide-character code <[wc]>, or return -1 (if <[wc]>
25 does not correspond to a printable wide-character code).
26
27 PORTABILITY
28 <<wcwidth>> has been introduced in the Single UNIX Specification Volume 2.
29 <<wcwidth>> has been marked as an extension in the Single UNIX Specification Volume 3.
30 */
31
32 /*
33 * This is an implementation of wcwidth() (defined in
34 * IEEE Std 1002.1-2001) for Unicode.
35 *
36 * http://www.opengroup.org/onlinepubs/007904975/functions/wcwidth.html
37 *
38 * In fixed-width output devices, Latin characters all occupy a single
39 * "cell" position of equal width, whereas ideographic CJK characters
40 * occupy two such cells. Interoperability between terminal-line
41 * applications and (teletype-style) character terminals using the
42 * UTF-8 encoding requires agreement on which character should advance
43 * the cursor by how many cell positions. No established formal
44 * standards exist at present on which Unicode character shall occupy
45 * how many cell positions on character terminals. These routines are
46 * a first attempt of defining such behavior based on simple rules
47 * applied to data provided by the Unicode Consortium.
48 *
49 * For some graphical characters, the Unicode standard explicitly
50 * defines a character-cell width via the definition of the East Asian
51 * FullWidth (F), Wide (W), Half-width (H), and Narrow (Na) classes.
52 * In all these cases, there is no ambiguity about which width a
53 * terminal shall use. For characters in the East Asian Ambiguous (A)
54 * class, the width choice depends purely on a preference of backward
55 * compatibility with either historic CJK or Western practice.
56 * Choosing single-width for these characters is easy to justify as
57 * the appropriate long-term solution, as the CJK practice of
58 * displaying these characters as double-width comes from historic
59 * implementation simplicity (8-bit encoded characters were displayed
60 * single-width and 16-bit ones double-width, even for Greek,
61 * Cyrillic, etc.) and not any typographic considerations.
62 *
63 * Much less clear is the choice of width for the Not East Asian
64 * (Neutral) class. Existing practice does not dictate a width for any
65 * of these characters. It would nevertheless make sense
66 * typographically to allocate two character cells to characters such
67 * as for instance EM SPACE or VOLUME INTEGRAL, which cannot be
68 * represented adequately with a single-width glyph. The following
69 * routines at present merely assign a single-cell width to all
70 * neutral characters, in the interest of simplicity. This is not
71 * entirely satisfactory and should be reconsidered before
72 * establishing a formal standard in this area. At the moment, the
73 * decision which Not East Asian (Neutral) characters should be
74 * represented by double-width glyphs cannot yet be answered by
75 * applying a simple rule from the Unicode database content. Setting
76 * up a proper standard for the behavior of UTF-8 character terminals
77 * will require a careful analysis not only of each Unicode character,
78 * but also of each presentation form, something the author of these
79 * routines has avoided to do so far.
80 *
81 * http://www.unicode.org/unicode/reports/tr11/
82 *
83 * Markus Kuhn -- 2007-05-26 (Unicode 5.0)
84 *
85 * Permission to use, copy, modify, and distribute this software
86 * for any purpose and without fee is hereby granted. The author
87 * disclaims all warranties with regard to this software.
88 *
89 * Latest version: http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
90 */
91
92 #include <_ansi.h>
93 #include <wchar.h>
94 #include <stdint.h>
95 #ifndef _MB_CAPABLE
96 #include <wctype.h> /* iswprint, iswcntrl */
97 #endif
98 #include "local.h"
99
100 #ifdef _MB_CAPABLE
101 struct interval
102 {
103 uint32_t first;
104 uint32_t last;
105 };
106
107 /* auxiliary function for binary search in interval table */
108 static int
bisearch(uint32_t ucs,const struct interval * table,int max)109 bisearch(uint32_t ucs, const struct interval *table, int max)
110 {
111 int min = 0;
112 int mid;
113
114 if (ucs < table[0].first || ucs > table[max].last)
115 return 0;
116 while (max >= min)
117 {
118 mid = (min + max) / 2;
119 if (ucs > table[mid].last)
120 min = mid + 1;
121 else if (ucs < table[mid].first)
122 max = mid - 1;
123 else
124 return 1;
125 }
126
127 return 0;
128 }
129 #endif /* _MB_CAPABLE */
130
131 /* The following function defines the column width of an ISO 10646
132 * character as follows:
133 *
134 * - The null character (U+0000) has a column width of 0.
135 *
136 * - Other C0/C1 control characters and DEL will lead to a return
137 * value of -1.
138 *
139 * - If the current language is recognized as a language usually using
140 * CJK fonts, spacing characters in the East Asian Ambiguous (A)
141 * category as defined in Unicode Technical Report #11 have a column
142 * width of 2.
143 *
144 * - Non-spacing and enclosing combining characters (general
145 * category code Mn or Me in the Unicode database) have a
146 * column width of 0.
147 *
148 * - SOFT HYPHEN (U+00AD) has a column width of 1.
149 *
150 * - Other format characters (general category code Cf in the Unicode
151 * database) and ZERO WIDTH SPACE (U+200B) have a column width of 0.
152 *
153 * - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF)
154 * have a column width of 0.
155 *
156 * - Spacing characters in the East Asian Wide (W) or East Asian
157 * Full-width (F) category as defined in Unicode Technical
158 * Report #11 have a column width of 2.
159 *
160 * - All remaining characters (including all printable
161 * ISO 8859-1 and WGL4 characters, Unicode control characters,
162 * etc.) have a column width of 1.
163 *
164 * This implementation assumes that wint_t characters are encoded
165 * in ISO 10646.
166 */
167
168 int
__wcwidth(const wint_t _ucs)169 __wcwidth (const wint_t _ucs)
170 {
171 uint32_t ucs = (uint32_t) _ucs;
172 #ifdef _MB_CAPABLE
173 /* sorted list of non-overlapping intervals of East Asian Ambiguous chars */
174 static const struct interval ambiguous[] =
175 #include "ambiguous.t"
176
177 /* sorted list of non-overlapping intervals of non-spacing characters */
178 static const struct interval combining[] =
179 #include "combining.t"
180
181 /* sorted list of non-overlapping intervals of wide characters,
182 ranges extended to Blocks where possible
183 */
184 static const struct interval wide[] =
185 #include "wide.t"
186
187 /* Test for NUL character */
188 if (ucs == 0)
189 return 0;
190
191 /* Test for printable ASCII characters */
192 if (ucs >= 0x20 && ucs < 0x7f)
193 return 1;
194
195 /* Test for control characters */
196 if (ucs < 0xa0)
197 return -1;
198
199 /* Test for surrogate pair values. */
200 if (ucs >= (uint32_t) 0xd800 && ucs <= (uint32_t) 0xdfff)
201 return -1;
202
203 /* check CJK width mode (1: ambiguous-wide, 0: normal, -1: disabled) */
204 int cjk_lang = __locale_cjk_lang ();
205
206 /* binary search in table of ambiguous characters */
207 if (cjk_lang > 0
208 && bisearch(ucs, ambiguous,
209 sizeof(ambiguous) / sizeof(struct interval) - 1))
210 return 2;
211
212 /* binary search in table of non-spacing characters */
213 if (bisearch(ucs, combining,
214 sizeof(combining) / sizeof(struct interval) - 1))
215 return 0;
216
217 /* if we arrive here, ucs is not a combining or C0/C1 control character */
218
219 /* binary search in table of wide character codes */
220 if (cjk_lang >= 0
221 && bisearch(ucs, wide,
222 sizeof(wide) / sizeof(struct interval) - 1))
223 return 2;
224 else
225 return 1;
226 #else /* !_MB_CAPABLE */
227 if (iswprint (ucs))
228 return 1;
229 if (iswcntrl (ucs) || ucs == L'\0')
230 return 0;
231 return -1;
232 #endif /* _MB_CAPABLE */
233 }
234
235 int
wcwidth(const wchar_t wc)236 wcwidth (const wchar_t wc)
237 {
238 wint_t wi = wc;
239
240 #ifdef _MB_CAPABLE
241 wi = _jp2uc (wi);
242 #endif /* _MB_CAPABLE */
243 return __wcwidth (wi);
244 }
245
