fltk  1.3.5-source
About: FLTK (Fast Light Tool Kit) is a cross-platform C++ GUI toolkit for UNIX/Linux (X11), Microsoft Windows, and MacOS X.
  Fossies Dox: fltk-1.3.5-source.tar.bz2  ("inofficial" and yet experimental doxygen-generated source code documentation)  

mk_wcwidth.c
Go to the documentation of this file.
1 /*
2  * FLTK: Important!
3  * This file should remain as close to Markus Kuhn's original source
4  * as possible for easy checking for changes later, however unlikely.
5  * All customisations to work with FLTK shall be annotated!
6  */
7 
8 /*
9  * This is an implementation of wcwidth() and wcswidth() (defined in
10  * IEEE Std 1002.1-2001) for Unicode.
11  *
12  * http://www.opengroup.org/onlinepubs/007904975/functions/wcwidth.html
13  * http://www.opengroup.org/onlinepubs/007904975/functions/wcswidth.html
14  *
15  * In fixed-width output devices, Latin characters all occupy a single
16  * "cell" position of equal width, whereas ideographic CJK characters
17  * occupy two such cells. Interoperability between terminal-line
18  * applications and (teletype-style) character terminals using the
19  * UTF-8 encoding requires agreement on which character should advance
20  * the cursor by how many cell positions. No established formal
21  * standards exist at present on which Unicode character shall occupy
22  * how many cell positions on character terminals. These routines are
23  * a first attempt of defining such behavior based on simple rules
24  * applied to data provided by the Unicode Consortium.
25  *
26  * For some graphical characters, the Unicode standard explicitly
27  * defines a character-cell width via the definition of the East Asian
28  * FullWidth (F), Wide (W), Half-width (H), and Narrow (Na) classes.
29  * In all these cases, there is no ambiguity about which width a
30  * terminal shall use. For characters in the East Asian Ambiguous (A)
31  * class, the width choice depends purely on a preference of backward
32  * compatibility with either historic CJK or Western practice.
33  * Choosing single-width for these characters is easy to justify as
34  * the appropriate long-term solution, as the CJK practice of
35  * displaying these characters as double-width comes from historic
36  * implementation simplicity (8-bit encoded characters were displayed
37  * single-width and 16-bit ones double-width, even for Greek,
38  * Cyrillic, etc.) and not any typographic considerations.
39  *
40  * Much less clear is the choice of width for the Not East Asian
41  * (Neutral) class. Existing practice does not dictate a width for any
42  * of these characters. It would nevertheless make sense
43  * typographically to allocate two character cells to characters such
44  * as for instance EM SPACE or VOLUME INTEGRAL, which cannot be
45  * represented adequately with a single-width glyph. The following
46  * routines at present merely assign a single-cell width to all
47  * neutral characters, in the interest of simplicity. This is not
48  * entirely satisfactory and should be reconsidered before
49  * establishing a formal standard in this area. At the moment, the
50  * decision which Not East Asian (Neutral) characters should be
51  * represented by double-width glyphs cannot yet be answered by
52  * applying a simple rule from the Unicode database content. Setting
53  * up a proper standard for the behavior of UTF-8 character terminals
54  * will require a careful analysis not only of each Unicode character,
55  * but also of each presentation form, something the author of these
56  * routines has avoided to do so far.
57  *
58  * http://www.unicode.org/unicode/reports/tr11/
59  *
60  * Markus Kuhn -- 2007-05-26 (Unicode 5.0)
61  *
62  * Permission to use, copy, modify, and distribute this software
63  * for any purpose and without fee is hereby granted. The author
64  * disclaims all warranties with regard to this software.
65  *
66  * Latest version: http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
67  */
68 
69 /*
70  * FLTK - avoid possible problems on systems with 32-bit wchar_t.
71  * Don't include wchar.h, and change wchar_t to unsigned int.
72  * Can we guarantee sizeof(unsigned int) >= 4 ?
73  */
74 #if 0
75 #include <wchar.h>
76 #endif
77 
78 struct interval {
79  unsigned int first;
80  unsigned int last;
81 };
82 
83 /* auxiliary function for binary search in interval table */
84 /*
85  * FLTK: was
86 static int bisearch(wchar_t ucs, const struct interval *table, int max) {
87  */
88 static int bisearch(unsigned int ucs, const struct interval *table, int max) {
89  int min = 0;
90  int mid;
91 
92  if (ucs < table[0].first || ucs > table[max].last)
93  return 0;
94  while (max >= min) {
95  mid = (min + max) / 2;
96  if (ucs > table[mid].last)
97  min = mid + 1;
98  else if (ucs < table[mid].first)
99  max = mid - 1;
100  else
101  return 1;
102  }
103 
104  return 0;
105 }
106 
107 
108 /* The following two functions define the column width of an ISO 10646
109  * character as follows:
110  *
111  * - The null character (U+0000) has a column width of 0.
112  *
113  * - Other C0/C1 control characters and DEL will lead to a return
114  * value of -1.
115  *
116  * - Non-spacing and enclosing combining characters (general
117  * category code Mn or Me in the Unicode database) have a
118  * column width of 0.
119  *
120  * - SOFT HYPHEN (U+00AD) has a column width of 1.
121  *
122  * - Other format characters (general category code Cf in the Unicode
123  * database) and ZERO WIDTH SPACE (U+200B) have a column width of 0.
124  *
125  * - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF)
126  * have a column width of 0.
127  *
128  * - Spacing characters in the East Asian Wide (W) or East Asian
129  * Full-width (F) category as defined in Unicode Technical
130  * Report #11 have a column width of 2.
131  *
132  * - All remaining characters (including all printable
133  * ISO 8859-1 and WGL4 characters, Unicode control characters,
134  * etc.) have a column width of 1.
135  *
136  * This implementation assumes that wchar_t characters are encoded
137  * in ISO 10646.
138  */
139 
140 /*
141  * FLTK: was
142 int mk_wcwidth(wchar_t ucs)
143  */
144 int mk_wcwidth(unsigned int ucs)
145 {
146  /* sorted list of non-overlapping intervals of non-spacing characters */
147  /* generated by "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c" */
148  static const struct interval combining[] = {
149  { 0x0300, 0x036F }, { 0x0483, 0x0486 }, { 0x0488, 0x0489 },
150  { 0x0591, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 },
151  { 0x05C4, 0x05C5 }, { 0x05C7, 0x05C7 }, { 0x0600, 0x0603 },
152  { 0x0610, 0x0615 }, { 0x064B, 0x065E }, { 0x0670, 0x0670 },
153  { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
154  { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
155  { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0901, 0x0902 },
156  { 0x093C, 0x093C }, { 0x0941, 0x0948 }, { 0x094D, 0x094D },
157  { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, { 0x0981, 0x0981 },
158  { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD },
159  { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, { 0x0A3C, 0x0A3C },
160  { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D },
161  { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC },
162  { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD },
163  { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C },
164  { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, { 0x0B4D, 0x0B4D },
165  { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, { 0x0BC0, 0x0BC0 },
166  { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, { 0x0C46, 0x0C48 },
167  { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, { 0x0CBC, 0x0CBC },
168  { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD },
169  { 0x0CE2, 0x0CE3 }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D },
170  { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
171  { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
172  { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
173  { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
174  { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
175  { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 },
176  { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
177  { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 },
178  { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x135F, 0x135F },
179  { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 },
180  { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD },
181  { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD },
182  { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 },
183  { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B },
184  { 0x1A17, 0x1A18 }, { 0x1B00, 0x1B03 }, { 0x1B34, 0x1B34 },
185  { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C }, { 0x1B42, 0x1B42 },
186  { 0x1B6B, 0x1B73 }, { 0x1DC0, 0x1DCA }, { 0x1DFE, 0x1DFF },
187  { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x2060, 0x2063 },
188  { 0x206A, 0x206F }, { 0x20D0, 0x20EF }, { 0x302A, 0x302F },
189  { 0x3099, 0x309A }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B },
190  { 0xA825, 0xA826 }, { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F },
191  { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB },
192  { 0x10A01, 0x10A03 }, { 0x10A05, 0x10A06 }, { 0x10A0C, 0x10A0F },
193  { 0x10A38, 0x10A3A }, { 0x10A3F, 0x10A3F }, { 0x1D167, 0x1D169 },
194  { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B }, { 0x1D1AA, 0x1D1AD },
195  { 0x1D242, 0x1D244 }, { 0xE0001, 0xE0001 }, { 0xE0020, 0xE007F },
196  { 0xE0100, 0xE01EF }
197  };
198 
199  /* test for 8-bit control characters */
200  if (ucs == 0)
201  return 0;
202  if (ucs < 32 || (ucs >= 0x7f && ucs < 0xa0))
203  return -1;
204 
205  /* binary search in table of non-spacing characters */
206  if (bisearch(ucs, combining,
207  sizeof(combining) / sizeof(struct interval) - 1))
208  return 0;
209 
210  /* if we arrive here, ucs is not a combining or C0/C1 control character */
211 
212  return 1 +
213  (ucs >= 0x1100 &&
214  (ucs <= 0x115f || /* Hangul Jamo init. consonants */
215  ucs == 0x2329 || ucs == 0x232a ||
216  (ucs >= 0x2e80 && ucs <= 0xa4cf &&
217  ucs != 0x303f) || /* CJK ... Yi */
218  (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
219  (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
220  (ucs >= 0xfe10 && ucs <= 0xfe19) || /* Vertical forms */
221  (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
222  (ucs >= 0xff00 && ucs <= 0xff60) || /* Fullwidth Forms */
223  (ucs >= 0xffe0 && ucs <= 0xffe6) ||
224  (ucs >= 0x20000 && ucs <= 0x2fffd) ||
225  (ucs >= 0x30000 && ucs <= 0x3fffd)));
226 }
227 
228 
229 /*
230  * FLTK: comment out the remaining functions, as we don't need themm.
231  */
232 #if 0
233 
234 /*
235  * FLTK: was
236 int mk_wcswidth(const wchar_t *pwcs, size_t n)
237  */
238 int mk_wcswidth(const unsigned int *pwcs, size_t n)
239 {
240  int w, width = 0;
241 
242  for (;*pwcs && n-- > 0; pwcs++)
243  if ((w = mk_wcwidth(*pwcs)) < 0)
244  return -1;
245  else
246  width += w;
247 
248  return width;
249 }
250 
251 
252 /*
253  * The following functions are the same as mk_wcwidth() and
254  * mk_wcswidth(), except that spacing characters in the East Asian
255  * Ambiguous (A) category as defined in Unicode Technical Report #11
256  * have a column width of 2. This variant might be useful for users of
257  * CJK legacy encodings who want to migrate to UCS without changing
258  * the traditional terminal character-width behaviour. It is not
259  * otherwise recommended for general use.
260  */
261 /*
262  * FLTK: was
263 int mk_wcwidth_cjk(wchar_t ucs)
264  */
265 int mk_wcwidth_cjk(unsigned int ucs)
266 {
267  /* sorted list of non-overlapping intervals of East Asian Ambiguous
268  * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
269  static const struct interval ambiguous[] = {
270  { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
271  { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
272  { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
273  { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
274  { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
275  { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
276  { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
277  { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
278  { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
279  { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
280  { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
281  { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
282  { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
283  { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
284  { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
285  { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
286  { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
287  { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
288  { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
289  { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
290  { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
291  { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
292  { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
293  { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
294  { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
295  { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
296  { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
297  { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
298  { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
299  { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
300  { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
301  { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
302  { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
303  { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
304  { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
305  { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
306  { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
307  { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
308  { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
309  { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
310  { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
311  { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
312  { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
313  { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
314  { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
315  { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
316  { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
317  { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
318  { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
319  { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
320  { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
321  { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
322  };
323 
324  /* binary search in table of non-spacing characters */
325  if (bisearch(ucs, ambiguous,
326  sizeof(ambiguous) / sizeof(struct interval) - 1))
327  return 2;
328 
329  return mk_wcwidth(ucs);
330 }
331 
332 
333 /*
334  * FLTK: was
335 int mk_wcswidth_cjk(const wchar_t *pwcs, size_t n)
336  */
337 int mk_wcswidth_cjk(const unsigned int *pwcs, size_t n)
338 {
339  int w, width = 0;
340 
341  for (;*pwcs && n-- > 0; pwcs++)
342  if ((w = mk_wcwidth_cjk(*pwcs)) < 0)
343  return -1;
344  else
345  width += w;
346 
347  return width;
348 }
349 
350 /*
351  * FLTK: end of commented out functions
352  */
353 #endif
min
static int min(int i1, int i2)
Definition: Fl_Text_Buffer.cxx:63
mk_wcwidth
int mk_wcwidth(unsigned int ucs)
Definition: mk_wcwidth.c:144
interval::first
unsigned int first
Definition: mk_wcwidth.c:79
last
static idle_cb * last
Definition: Fl_add_idle.cxx:34
max
static int max(int i1, int i2)
Definition: Fl_Text_Buffer.cxx:58
table
static symbol table[]
Definition: factory.cxx:1109
first
static idle_cb * first
Definition: Fl_add_idle.cxx:33
interval
Definition: mk_wcwidth.c:78
interval::last
unsigned int last
Definition: mk_wcwidth.c:80
bisearch
static int bisearch(unsigned int ucs, const struct interval *table, int max)
Definition: mk_wcwidth.c:88