diff options
Diffstat (limited to 'libbb/unicode_wcwidth.c')
-rw-r--r-- | libbb/unicode_wcwidth.c | 217 |
1 files changed, 217 insertions, 0 deletions
diff --git a/libbb/unicode_wcwidth.c b/libbb/unicode_wcwidth.c new file mode 100644 index 000000000..515c084c1 --- /dev/null +++ b/libbb/unicode_wcwidth.c | |||
@@ -0,0 +1,217 @@ | |||
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 | struct interval { | ||
63 | uint16_t first; | ||
64 | uint16_t last; | ||
65 | }; | ||
66 | |||
67 | /* auxiliary function for binary search in interval table */ | ||
68 | static int in_table(unsigned ucs, const struct interval *table, unsigned max) | ||
69 | { | ||
70 | unsigned min = 0; | ||
71 | unsigned mid; | ||
72 | |||
73 | if (ucs < table[0].first || ucs > table[max].last) | ||
74 | return 0; | ||
75 | |||
76 | while (max >= min) { | ||
77 | mid = (min + max) / 2; | ||
78 | if (ucs > table[mid].last) | ||
79 | min = mid + 1; | ||
80 | else if (ucs < table[mid].first) | ||
81 | max = mid - 1; | ||
82 | else | ||
83 | return 1; | ||
84 | } | ||
85 | return 0; | ||
86 | } | ||
87 | |||
88 | |||
89 | /* The following two functions define the column width of an ISO 10646 | ||
90 | * character as follows: | ||
91 | * | ||
92 | * - The null character (U+0000) has a column width of 0. | ||
93 | * | ||
94 | * - Other C0/C1 control characters and DEL will lead to a return | ||
95 | * value of -1. | ||
96 | * | ||
97 | * - Non-spacing and enclosing combining characters (general | ||
98 | * category code Mn or Me in the Unicode database) have a | ||
99 | * column width of 0. | ||
100 | * | ||
101 | * - SOFT HYPHEN (U+00AD) has a column width of 1. | ||
102 | * | ||
103 | * - Other format characters (general category code Cf in the Unicode | ||
104 | * database) and ZERO WIDTH SPACE (U+200B) have a column width of 0. | ||
105 | * | ||
106 | * - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF) | ||
107 | * have a column width of 0. | ||
108 | * | ||
109 | * - Spacing characters in the East Asian Wide (W) or East Asian | ||
110 | * Full-width (F) category as defined in Unicode Technical | ||
111 | * Report #11 have a column width of 2. | ||
112 | * | ||
113 | * - All remaining characters (including all printable | ||
114 | * ISO 8859-1 and WGL4 characters, Unicode control characters, | ||
115 | * etc.) have a column width of 1. | ||
116 | * | ||
117 | * This implementation assumes that wchar_t characters are encoded | ||
118 | * in ISO 10646. | ||
119 | */ | ||
120 | static int wcwidth(unsigned ucs) | ||
121 | { | ||
122 | /* sorted list of non-overlapping intervals of non-spacing characters */ | ||
123 | /* generated by "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c" */ | ||
124 | static const struct interval combining[] = { | ||
125 | { 0x0300, 0x036F }, { 0x0483, 0x0486 }, { 0x0488, 0x0489 }, | ||
126 | { 0x0591, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 }, | ||
127 | { 0x05C4, 0x05C5 }, { 0x05C7, 0x05C7 }, { 0x0600, 0x0603 }, | ||
128 | { 0x0610, 0x0615 }, { 0x064B, 0x065E }, { 0x0670, 0x0670 }, | ||
129 | { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED }, | ||
130 | { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A }, | ||
131 | { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0901, 0x0902 }, | ||
132 | { 0x093C, 0x093C }, { 0x0941, 0x0948 }, { 0x094D, 0x094D }, | ||
133 | { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, { 0x0981, 0x0981 }, | ||
134 | { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD }, | ||
135 | { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, { 0x0A3C, 0x0A3C }, | ||
136 | { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D }, | ||
137 | { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC }, | ||
138 | { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD }, | ||
139 | { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C }, | ||
140 | { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, { 0x0B4D, 0x0B4D }, | ||
141 | { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, { 0x0BC0, 0x0BC0 }, | ||
142 | { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, { 0x0C46, 0x0C48 }, | ||
143 | { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, { 0x0CBC, 0x0CBC }, | ||
144 | { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD }, | ||
145 | { 0x0CE2, 0x0CE3 }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D }, | ||
146 | { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 }, | ||
147 | { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E }, | ||
148 | { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC }, | ||
149 | { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 }, | ||
150 | { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E }, | ||
151 | { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 }, | ||
152 | { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 }, | ||
153 | { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 }, | ||
154 | { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x135F, 0x135F }, | ||
155 | { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 }, | ||
156 | { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD }, | ||
157 | { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD }, | ||
158 | { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 }, | ||
159 | { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B }, | ||
160 | { 0x1A17, 0x1A18 }, { 0x1B00, 0x1B03 }, { 0x1B34, 0x1B34 }, | ||
161 | { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C }, { 0x1B42, 0x1B42 }, | ||
162 | { 0x1B6B, 0x1B73 }, { 0x1DC0, 0x1DCA }, { 0x1DFE, 0x1DFF }, | ||
163 | { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x2060, 0x2063 }, | ||
164 | { 0x206A, 0x206F }, { 0x20D0, 0x20EF }, { 0x302A, 0x302F }, | ||
165 | { 0x3099, 0x309A }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B }, | ||
166 | { 0xA825, 0xA826 }, { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F }, | ||
167 | { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB } | ||
168 | }; | ||
169 | static const struct interval combining0x10000[] = { | ||
170 | { 0x0A01, 0x0A03 }, { 0x0A05, 0x0A06 }, { 0x0A0C, 0x0A0F }, | ||
171 | { 0x0A38, 0x0A3A }, { 0x0A3F, 0x0A3F }, { 0xD167, 0xD169 }, | ||
172 | { 0xD173, 0xD182 }, { 0xD185, 0xD18B }, { 0xD1AA, 0xD1AD }, | ||
173 | { 0xD242, 0xD244 } | ||
174 | }; | ||
175 | |||
176 | if (ucs == 0) | ||
177 | return 0; | ||
178 | /* test for 8-bit control characters (00-1f, 80-9f, 7f) */ | ||
179 | if ((ucs & ~0x80) < 0x20 || ucs == 0x7f) | ||
180 | return -1; | ||
181 | if (ucs < 0x0300) /* optimization */ | ||
182 | return 1; | ||
183 | |||
184 | /* binary search in table of non-spacing characters */ | ||
185 | if (in_table(ucs, combining, ARRAY_SIZE(combining) - 1)) | ||
186 | return 0; | ||
187 | |||
188 | if (ucs < 0x1100) /* optimization */ | ||
189 | return 1; | ||
190 | |||
191 | /* binary search in table of non-spacing characters, cont. */ | ||
192 | if (in_table(ucs ^ 0x10000, combining0x10000, ARRAY_SIZE(combining0x10000) - 1)) | ||
193 | return 0; | ||
194 | if (ucs == 0xE0001 | ||
195 | || (ucs >= 0xE0020 && ucs <= 0xE007F) | ||
196 | || (ucs >= 0xE0100 && ucs <= 0xE01EF) | ||
197 | ) { | ||
198 | return 0; | ||
199 | } | ||
200 | |||
201 | /* if we arrive here, ucs is not a combining or C0/C1 control character */ | ||
202 | |||
203 | return 1 + | ||
204 | ( (/*ucs >= 0x1100 &&*/ ucs <= 0x115f) /* Hangul Jamo init. consonants */ | ||
205 | || ucs == 0x2329 | ||
206 | || ucs == 0x232a | ||
207 | || (ucs >= 0x2e80 && ucs <= 0xa4cf && ucs != 0x303f) /* CJK ... Yi */ | ||
208 | || (ucs >= 0xac00 && ucs <= 0xd7a3) /* Hangul Syllables */ | ||
209 | || (ucs >= 0xf900 && ucs <= 0xfaff) /* CJK Compatibility Ideographs */ | ||
210 | || (ucs >= 0xfe10 && ucs <= 0xfe19) /* Vertical forms */ | ||
211 | || (ucs >= 0xfe30 && ucs <= 0xfe6f) /* CJK Compatibility Forms */ | ||
212 | || (ucs >= 0xff00 && ucs <= 0xff60) /* Fullwidth Forms */ | ||
213 | || (ucs >= 0xffe0 && ucs <= 0xffe6) | ||
214 | || (ucs >= 0x20000 && ucs <= 0x2fffd) | ||
215 | || (ucs >= 0x30000 && ucs <= 0x3fffd) | ||
216 | ); | ||
217 | } | ||