1 /* guniprop.c - Unicode character properties.
3 * Copyright (C) 1999 Tom Tromey
4 * Copyright (C) 2000 Red Hat, Inc.
6 * This library is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2 of the License, or (at your option) any later version.
11 * This library is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with this library; if not, write to the
18 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
19 * Boston, MA 02111-1307, USA.
30 #include "gunichartables.h"
31 #include "gmirroringtable.h"
32 #include "gscripttable.h"
33 #include "gunicodeprivate.h"
36 #define ATTR_TABLE(Page) (((Page) <= G_UNICODE_LAST_PAGE_PART1) \
37 ? attr_table_part1[Page] \
38 : attr_table_part2[(Page) - 0xe00])
40 #define ATTTABLE(Page, Char) \
41 ((ATTR_TABLE(Page) == G_UNICODE_MAX_TABLE_INDEX) ? 0 : (attr_data[ATTR_TABLE(Page)][Char]))
43 #define TTYPE_PART1(Page, Char) \
44 ((type_table_part1[Page] >= G_UNICODE_MAX_TABLE_INDEX) \
45 ? (type_table_part1[Page] - G_UNICODE_MAX_TABLE_INDEX) \
46 : (type_data[type_table_part1[Page]][Char]))
48 #define TTYPE_PART2(Page, Char) \
49 ((type_table_part2[Page] >= G_UNICODE_MAX_TABLE_INDEX) \
50 ? (type_table_part2[Page] - G_UNICODE_MAX_TABLE_INDEX) \
51 : (type_data[type_table_part2[Page]][Char]))
54 (((Char) <= G_UNICODE_LAST_CHAR_PART1) \
55 ? TTYPE_PART1 ((Char) >> 8, (Char) & 0xff) \
56 : (((Char) >= 0xe0000 && (Char) <= G_UNICODE_LAST_CHAR) \
57 ? TTYPE_PART2 (((Char) - 0xe0000) >> 8, (Char) & 0xff) \
58 : G_UNICODE_UNASSIGNED))
61 #define IS(Type, Class) (((guint)1 << (Type)) & (Class))
62 #define OR(Type, Rest) (((guint)1 << (Type)) | (Rest))
66 #define ISALPHA(Type) IS ((Type), \
67 OR (G_UNICODE_LOWERCASE_LETTER, \
68 OR (G_UNICODE_UPPERCASE_LETTER, \
69 OR (G_UNICODE_TITLECASE_LETTER, \
70 OR (G_UNICODE_MODIFIER_LETTER, \
71 OR (G_UNICODE_OTHER_LETTER, 0))))))
73 #define ISALDIGIT(Type) IS ((Type), \
74 OR (G_UNICODE_DECIMAL_NUMBER, \
75 OR (G_UNICODE_LETTER_NUMBER, \
76 OR (G_UNICODE_OTHER_NUMBER, \
77 OR (G_UNICODE_LOWERCASE_LETTER, \
78 OR (G_UNICODE_UPPERCASE_LETTER, \
79 OR (G_UNICODE_TITLECASE_LETTER, \
80 OR (G_UNICODE_MODIFIER_LETTER, \
81 OR (G_UNICODE_OTHER_LETTER, 0)))))))))
83 #define ISMARK(Type) IS ((Type), \
84 OR (G_UNICODE_NON_SPACING_MARK, \
85 OR (G_UNICODE_COMBINING_MARK, \
86 OR (G_UNICODE_ENCLOSING_MARK, 0))))
88 #define ISZEROWIDTHTYPE(Type) IS ((Type), \
89 OR (G_UNICODE_NON_SPACING_MARK, \
90 OR (G_UNICODE_ENCLOSING_MARK, \
91 OR (G_UNICODE_FORMAT, 0))))
95 * @c: a Unicode character
97 * Determines whether a character is alphanumeric.
98 * Given some UTF-8 text, obtain a character value
99 * with g_utf8_get_char().
101 * Return value: %TRUE if @c is an alphanumeric character
104 g_unichar_isalnum (gunichar c)
106 return ISALDIGIT (TYPE (c)) ? TRUE : FALSE;
111 * @c: a Unicode character
113 * Determines whether a character is alphabetic (i.e. a letter).
114 * Given some UTF-8 text, obtain a character value with
117 * Return value: %TRUE if @c is an alphabetic character
120 g_unichar_isalpha (gunichar c)
122 return ISALPHA (TYPE (c)) ? TRUE : FALSE;
128 * @c: a Unicode character
130 * Determines whether a character is a control character.
131 * Given some UTF-8 text, obtain a character value with
134 * Return value: %TRUE if @c is a control character
137 g_unichar_iscntrl (gunichar c)
139 return TYPE (c) == G_UNICODE_CONTROL;
144 * @c: a Unicode character
146 * Determines whether a character is numeric (i.e. a digit). This
147 * covers ASCII 0-9 and also digits in other languages/scripts. Given
148 * some UTF-8 text, obtain a character value with g_utf8_get_char().
150 * Return value: %TRUE if @c is a digit
153 g_unichar_isdigit (gunichar c)
155 return TYPE (c) == G_UNICODE_DECIMAL_NUMBER;
161 * @c: a Unicode character
163 * Determines whether a character is printable and not a space
164 * (returns %FALSE for control characters, format characters, and
165 * spaces). g_unichar_isprint() is similar, but returns %TRUE for
166 * spaces. Given some UTF-8 text, obtain a character value with
169 * Return value: %TRUE if @c is printable unless it's a space
172 g_unichar_isgraph (gunichar c)
175 OR (G_UNICODE_CONTROL,
176 OR (G_UNICODE_FORMAT,
177 OR (G_UNICODE_UNASSIGNED,
178 OR (G_UNICODE_PRIVATE_USE,
179 OR (G_UNICODE_SURROGATE,
180 OR (G_UNICODE_SPACE_SEPARATOR,
186 * @c: a Unicode character
188 * Determines whether a character is a lowercase letter.
189 * Given some UTF-8 text, obtain a character value with
192 * Return value: %TRUE if @c is a lowercase letter
195 g_unichar_islower (gunichar c)
197 return TYPE (c) == G_UNICODE_LOWERCASE_LETTER;
203 * @c: a Unicode character
205 * Determines whether a character is printable.
206 * Unlike g_unichar_isgraph(), returns %TRUE for spaces.
207 * Given some UTF-8 text, obtain a character value with
210 * Return value: %TRUE if @c is printable
213 g_unichar_isprint (gunichar c)
216 OR (G_UNICODE_CONTROL,
217 OR (G_UNICODE_FORMAT,
218 OR (G_UNICODE_UNASSIGNED,
219 OR (G_UNICODE_PRIVATE_USE,
220 OR (G_UNICODE_SURROGATE,
226 * @c: a Unicode character
228 * Determines whether a character is punctuation or a symbol.
229 * Given some UTF-8 text, obtain a character value with
232 * Return value: %TRUE if @c is a punctuation or symbol character
235 g_unichar_ispunct (gunichar c)
238 OR (G_UNICODE_CONNECT_PUNCTUATION,
239 OR (G_UNICODE_DASH_PUNCTUATION,
240 OR (G_UNICODE_CLOSE_PUNCTUATION,
241 OR (G_UNICODE_FINAL_PUNCTUATION,
242 OR (G_UNICODE_INITIAL_PUNCTUATION,
243 OR (G_UNICODE_OTHER_PUNCTUATION,
244 OR (G_UNICODE_OPEN_PUNCTUATION,
245 OR (G_UNICODE_CURRENCY_SYMBOL,
246 OR (G_UNICODE_MODIFIER_SYMBOL,
247 OR (G_UNICODE_MATH_SYMBOL,
248 OR (G_UNICODE_OTHER_SYMBOL,
249 0)))))))))))) ? TRUE : FALSE;
254 * @c: a Unicode character
256 * Determines whether a character is a space, tab, or line separator
257 * (newline, carriage return, etc.). Given some UTF-8 text, obtain a
258 * character value with g_utf8_get_char().
260 * (Note: don't use this to do word breaking; you have to use
261 * Pango or equivalent to get word breaking right, the algorithm
262 * is fairly complex.)
264 * Return value: %TRUE if @c is a space character
267 g_unichar_isspace (gunichar c)
271 /* special-case these since Unicode thinks they are not spaces */
282 OR (G_UNICODE_SPACE_SEPARATOR,
283 OR (G_UNICODE_LINE_SEPARATOR,
284 OR (G_UNICODE_PARAGRAPH_SEPARATOR,
285 0)))) ? TRUE : FALSE;
293 * @c: a Unicode character
295 * Determines whether a character is a mark (non-spacing mark,
296 * combining mark, or enclosing mark in Unicode speak).
297 * Given some UTF-8 text, obtain a character value
298 * with g_utf8_get_char().
300 * Note: in most cases where isalpha characters are allowed,
301 * ismark characters should be allowed to as they are essential
302 * for writing most European languages as well as many non-Latin
305 * Return value: %TRUE if @c is a mark character
310 g_unichar_ismark (gunichar c)
312 return ISMARK (TYPE (c));
317 * @c: a Unicode character
319 * Determines if a character is uppercase.
321 * Return value: %TRUE if @c is an uppercase character
324 g_unichar_isupper (gunichar c)
326 return TYPE (c) == G_UNICODE_UPPERCASE_LETTER;
331 * @c: a Unicode character
333 * Determines if a character is titlecase. Some characters in
334 * Unicode which are composites, such as the DZ digraph
335 * have three case variants instead of just two. The titlecase
336 * form is used at the beginning of a word where only the
337 * first letter is capitalized. The titlecase form of the DZ
338 * digraph is U+01F2 LATIN CAPITAL LETTTER D WITH SMALL LETTER Z.
340 * Return value: %TRUE if the character is titlecase
343 g_unichar_istitle (gunichar c)
346 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
347 if (title_table[i][0] == c)
353 * g_unichar_isxdigit:
354 * @c: a Unicode character.
356 * Determines if a character is a hexidecimal digit.
358 * Return value: %TRUE if the character is a hexadecimal digit
361 g_unichar_isxdigit (gunichar c)
363 return ((c >= 'a' && c <= 'f')
364 || (c >= 'A' && c <= 'F')
365 || (TYPE (c) == G_UNICODE_DECIMAL_NUMBER));
369 * g_unichar_isdefined:
370 * @c: a Unicode character
372 * Determines if a given character is assigned in the Unicode
375 * Return value: %TRUE if the character has an assigned value
378 g_unichar_isdefined (gunichar c)
380 return TYPE (c) != G_UNICODE_UNASSIGNED;
384 * g_unichar_iszerowidth:
385 * @c: a Unicode character
387 * Determines if a given character typically takes zero width when rendered.
388 * The return value is %TRUE for all non-spacing and enclosing marks
389 * (e.g., combining accents), format characters, zero-width
390 * space, but not U+00AD SOFT HYPHEN.
392 * A typical use of this function is with one of g_unichar_iswide() or
393 * g_unichar_iswide_cjk() to determine the number of cells a string occupies
394 * when displayed on a grid display (terminals). However, note that not all
395 * terminals support zero-width rendering of zero-width marks.
397 * Return value: %TRUE if the character has zero width
402 g_unichar_iszerowidth (gunichar c)
404 if (G_UNLIKELY (c == 0x00AD))
407 if (G_UNLIKELY (ISZEROWIDTHTYPE (TYPE (c))))
410 if (G_UNLIKELY ((c >= 0x1160 && c < 0x1200) ||
419 * @c: a Unicode character
421 * Determines if a character is typically rendered in a double-width
424 * Return value: %TRUE if the character is wide
427 g_unichar_iswide (gunichar c)
431 /* The following block is stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>'s
432 * wcwidth implementation. */
435 (ucs <= 0x115f || /* Hangul Jamo init. consonants */
436 ucs == 0x2329 || ucs == 0x232a ||
437 (ucs >= 0x2e80 && ucs <= 0xa4cf &&
438 ucs != 0x303f) || /* CJK ... Yi */
439 (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
440 (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
441 (ucs >= 0xfe10 && ucs <= 0xfe19) || /* Vertical forms */
442 (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
443 (ucs >= 0xff00 && ucs <= 0xff60) || /* Fullwidth Forms */
444 (ucs >= 0xffe0 && ucs <= 0xffe6) ||
445 (ucs >= 0x20000 && ucs <= 0x2fffd) ||
446 (ucs >= 0x30000 && ucs <= 0x3fffd)));
450 /* The following block is stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>'s
451 * wcwidth implementation. */
457 /* The following block is stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>'s
458 * wcwidth implementation. */
460 interval_compare (const void *key, const void *elt)
462 gunichar c = GPOINTER_TO_UINT (key);
463 struct Interval *interval = (struct Interval *)elt;
465 if (c < interval->start)
467 if (c > interval->end)
474 * g_unichar_iswide_cjk:
475 * @c: a Unicode character
477 * Determines if a character is typically rendered in a double-width
478 * cell under legacy East Asian locales. If a character is wide according to
479 * g_unichar_iswide(), then it is also reported wide with this function, but
480 * the converse is not necessarily true. See the
481 * <ulink url="http://www.unicode.org/reports/tr11/">Unicode Standard
482 * Annex #11</ulink> for details.
484 * Return value: %TRUE if the character is wide in legacy East Asian locales
489 g_unichar_iswide_cjk (gunichar c)
491 /* The following block is stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>'s
492 * wcwidth implementation. */
493 /* sorted list of non-overlapping intervals of East Asian Ambiguous
494 * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
495 static const struct Interval ambiguous[] = {
496 { 0x0300, 0x036F }, { 0x0483, 0x0486 }, { 0x0488, 0x0489 },
497 { 0x0591, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 },
498 { 0x05C4, 0x05C5 }, { 0x05C7, 0x05C7 }, { 0x0600, 0x0603 },
499 { 0x0610, 0x0615 }, { 0x064B, 0x065E }, { 0x0670, 0x0670 },
500 { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
501 { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
502 { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0901, 0x0902 },
503 { 0x093C, 0x093C }, { 0x0941, 0x0948 }, { 0x094D, 0x094D },
504 { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, { 0x0981, 0x0981 },
505 { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD },
506 { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, { 0x0A3C, 0x0A3C },
507 { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D },
508 { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC },
509 { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD },
510 { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C },
511 { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, { 0x0B4D, 0x0B4D },
512 { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, { 0x0BC0, 0x0BC0 },
513 { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, { 0x0C46, 0x0C48 },
514 { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, { 0x0CBC, 0x0CBC },
515 { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD },
516 { 0x0CE2, 0x0CE3 }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D },
517 { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
518 { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
519 { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
520 { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
521 { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
522 { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 },
523 { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
524 { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 },
525 { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x135F, 0x135F },
526 { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 },
527 { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD },
528 { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD },
529 { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 },
530 { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B },
531 { 0x1A17, 0x1A18 }, { 0x1B00, 0x1B03 }, { 0x1B34, 0x1B34 },
532 { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C }, { 0x1B42, 0x1B42 },
533 { 0x1B6B, 0x1B73 }, { 0x1DC0, 0x1DCA }, { 0x1DFE, 0x1DFF },
534 { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x2060, 0x2063 },
535 { 0x206A, 0x206F }, { 0x20D0, 0x20EF }, { 0x302A, 0x302F },
536 { 0x3099, 0x309A }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B },
537 { 0xA825, 0xA826 }, { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F },
538 { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB },
539 { 0x10A01, 0x10A03 }, { 0x10A05, 0x10A06 }, { 0x10A0C, 0x10A0F },
540 { 0x10A38, 0x10A3A }, { 0x10A3F, 0x10A3F }, { 0x1D167, 0x1D169 },
541 { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B }, { 0x1D1AA, 0x1D1AD },
542 { 0x1D242, 0x1D244 }, { 0xE0001, 0xE0001 }, { 0xE0020, 0xE007F },
546 if (g_unichar_iswide (c))
549 if (bsearch (GUINT_TO_POINTER (c), ambiguous, G_N_ELEMENTS (ambiguous), sizeof ambiguous[0],
559 * @c: a Unicode character
561 * Converts a character to uppercase.
563 * Return value: the result of converting @c to uppercase.
564 * If @c is not an lowercase or titlecase character,
565 * or has no upper case equivalent @c is returned unchanged.
568 g_unichar_toupper (gunichar c)
571 if (t == G_UNICODE_LOWERCASE_LETTER)
573 gunichar val = ATTTABLE (c >> 8, c & 0xff);
574 if (val >= 0x1000000)
576 const gchar *p = special_case_table + val - 0x1000000;
577 val = g_utf8_get_char (p);
579 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
580 * do not have an uppercase equivalent, in which case val will be
583 return val ? val : c;
585 else if (t == G_UNICODE_TITLECASE_LETTER)
588 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
590 if (title_table[i][0] == c)
591 return title_table[i][1];
599 * @c: a Unicode character.
601 * Converts a character to lower case.
603 * Return value: the result of converting @c to lower case.
604 * If @c is not an upperlower or titlecase character,
605 * or has no lowercase equivalent @c is returned unchanged.
608 g_unichar_tolower (gunichar c)
611 if (t == G_UNICODE_UPPERCASE_LETTER)
613 gunichar val = ATTTABLE (c >> 8, c & 0xff);
614 if (val >= 0x1000000)
616 const gchar *p = special_case_table + val - 0x1000000;
617 return g_utf8_get_char (p);
621 /* Not all uppercase letters are guaranteed to have a lowercase
622 * equivalent. If this is the case, val will be zero. */
623 return val ? val : c;
626 else if (t == G_UNICODE_TITLECASE_LETTER)
629 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
631 if (title_table[i][0] == c)
632 return title_table[i][2];
640 * @c: a Unicode character
642 * Converts a character to the titlecase.
644 * Return value: the result of converting @c to titlecase.
645 * If @c is not an uppercase or lowercase character,
646 * @c is returned unchanged.
649 g_unichar_totitle (gunichar c)
652 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
654 if (title_table[i][0] == c || title_table[i][1] == c
655 || title_table[i][2] == c)
656 return title_table[i][0];
659 if (TYPE (c) == G_UNICODE_LOWERCASE_LETTER)
660 return g_unichar_toupper (c);
666 * g_unichar_digit_value:
667 * @c: a Unicode character
669 * Determines the numeric value of a character as a decimal
672 * Return value: If @c is a decimal digit (according to
673 * g_unichar_isdigit()), its numeric value. Otherwise, -1.
676 g_unichar_digit_value (gunichar c)
678 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
679 return ATTTABLE (c >> 8, c & 0xff);
684 * g_unichar_xdigit_value:
685 * @c: a Unicode character
687 * Determines the numeric value of a character as a hexidecimal
690 * Return value: If @c is a hex digit (according to
691 * g_unichar_isxdigit()), its numeric value. Otherwise, -1.
694 g_unichar_xdigit_value (gunichar c)
696 if (c >= 'A' && c <= 'F')
698 if (c >= 'a' && c <= 'f')
700 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
701 return ATTTABLE (c >> 8, c & 0xff);
707 * @c: a Unicode character
709 * Classifies a Unicode character by type.
711 * Return value: the type of the character.
714 g_unichar_type (gunichar c)
720 * Case mapping functions
730 get_locale_type (void)
733 char *tem = g_win32_getlocale ();
740 const char *locale = setlocale (LC_CTYPE, NULL);
746 if (locale[1] == 'z')
747 return LOCALE_TURKIC;
750 if (locale[1] == 't')
751 return LOCALE_LITHUANIAN;
754 if (locale[1] == 'r')
755 return LOCALE_TURKIC;
759 return LOCALE_NORMAL;
763 output_marks (const char **p_inout,
767 const char *p = *p_inout;
772 gunichar c = g_utf8_get_char (p);
774 if (ISMARK (TYPE (c)))
776 if (!remove_dot || c != 0x307 /* COMBINING DOT ABOVE */)
777 len += g_unichar_to_utf8 (c, out_buffer ? out_buffer + len : NULL);
778 p = g_utf8_next_char (p);
789 output_special_case (gchar *out_buffer,
794 const gchar *p = special_case_table + offset;
797 if (type != G_UNICODE_TITLECASE_LETTER)
798 p = g_utf8_next_char (p);
805 memcpy (out_buffer, p, len);
811 real_toupper (const gchar *str,
814 LocaleType locale_type)
816 const gchar *p = str;
817 const char *last = NULL;
819 gboolean last_was_i = FALSE;
821 while ((max_len < 0 || p < str + max_len) && *p)
823 gunichar c = g_utf8_get_char (p);
828 p = g_utf8_next_char (p);
830 if (locale_type == LOCALE_LITHUANIAN)
838 /* Nasty, need to remove any dot above. Though
839 * I think only E WITH DOT ABOVE occurs in practice
840 * which could simplify this considerably.
845 decomp = g_unicode_canonical_decomposition (c, &decomp_len);
846 for (i=0; i < decomp_len; i++)
848 if (decomp[i] != 0x307 /* COMBINING DOT ABOVE */)
849 len += g_unichar_to_utf8 (g_unichar_toupper (decomp[i]), out_buffer ? out_buffer + len : NULL);
853 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, TRUE);
863 if (locale_type == LOCALE_TURKIC && c == 'i')
865 /* i => LATIN CAPITAL LETTER I WITH DOT ABOVE */
866 len += g_unichar_to_utf8 (0x130, out_buffer ? out_buffer + len : NULL);
868 else if (c == 0x0345) /* COMBINING GREEK YPOGEGRAMMENI */
870 /* Nasty, need to move it after other combining marks .. this would go away if
871 * we normalized first.
873 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, FALSE);
875 /* And output as GREEK CAPITAL LETTER IOTA */
876 len += g_unichar_to_utf8 (0x399, out_buffer ? out_buffer + len : NULL);
879 OR (G_UNICODE_LOWERCASE_LETTER,
880 OR (G_UNICODE_TITLECASE_LETTER,
883 val = ATTTABLE (c >> 8, c & 0xff);
885 if (val >= 0x1000000)
887 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t,
888 t == G_UNICODE_LOWERCASE_LETTER ? 0 : 1);
892 if (t == G_UNICODE_TITLECASE_LETTER)
895 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
897 if (title_table[i][0] == c)
899 val = title_table[i][1];
905 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
906 * do not have an uppercase equivalent, in which case val will be
908 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
913 gsize char_len = g_utf8_skip[*(guchar *)last];
916 memcpy (out_buffer + len, last, char_len);
928 * @str: a UTF-8 encoded string
929 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
931 * Converts all Unicode characters in the string that have a case
932 * to uppercase. The exact manner that this is done depends
933 * on the current locale, and may result in the number of
934 * characters in the string increasing. (For instance, the
935 * German ess-zet will be changed to SS.)
937 * Return value: a newly allocated string, with all characters
938 * converted to uppercase.
941 g_utf8_strup (const gchar *str,
945 LocaleType locale_type;
948 g_return_val_if_fail (str != NULL, NULL);
950 locale_type = get_locale_type ();
953 * We use a two pass approach to keep memory management simple
955 result_len = real_toupper (str, len, NULL, locale_type);
956 result = g_malloc (result_len + 1);
957 real_toupper (str, len, result, locale_type);
958 result[result_len] = '\0';
963 /* traverses the string checking for characters with combining class == 230
964 * until a base character is found */
966 has_more_above (const gchar *str)
968 const gchar *p = str;
969 gint combining_class;
973 combining_class = _g_unichar_combining_class (g_utf8_get_char (p));
974 if (combining_class == 230)
976 else if (combining_class == 0)
979 p = g_utf8_next_char (p);
986 real_tolower (const gchar *str,
989 LocaleType locale_type)
991 const gchar *p = str;
992 const char *last = NULL;
995 while ((max_len < 0 || p < str + max_len) && *p)
997 gunichar c = g_utf8_get_char (p);
1002 p = g_utf8_next_char (p);
1004 if (locale_type == LOCALE_TURKIC && c == 'I')
1006 if (g_utf8_get_char (p) == 0x0307)
1008 /* I + COMBINING DOT ABOVE => i (U+0069) */
1009 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
1010 p = g_utf8_next_char (p);
1014 /* I => LATIN SMALL LETTER DOTLESS I */
1015 len += g_unichar_to_utf8 (0x131, out_buffer ? out_buffer + len : NULL);
1018 /* Introduce an explicit dot above when lowercasing capital I's and J's
1019 * whenever there are more accents above. [SpecialCasing.txt] */
1020 else if (locale_type == LOCALE_LITHUANIAN &&
1021 (c == 0x00cc || c == 0x00cd || c == 0x0128))
1023 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
1024 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1029 len += g_unichar_to_utf8 (0x0300, out_buffer ? out_buffer + len : NULL);
1032 len += g_unichar_to_utf8 (0x0301, out_buffer ? out_buffer + len : NULL);
1035 len += g_unichar_to_utf8 (0x0303, out_buffer ? out_buffer + len : NULL);
1039 else if (locale_type == LOCALE_LITHUANIAN &&
1040 (c == 'I' || c == 'J' || c == 0x012e) &&
1043 len += g_unichar_to_utf8 (g_unichar_tolower (c), out_buffer ? out_buffer + len : NULL);
1044 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1046 else if (c == 0x03A3) /* GREEK CAPITAL LETTER SIGMA */
1048 if ((max_len < 0 || p < str + max_len) && *p)
1050 gunichar next_c = g_utf8_get_char (p);
1051 int next_type = TYPE(next_c);
1053 /* SIGMA mapps differently depending on whether it is
1054 * final or not. The following simplified test would
1055 * fail in the case of combining marks following the
1056 * sigma, but I don't think that occurs in real text.
1057 * The test here matches that in ICU.
1059 if (ISALPHA (next_type)) /* Lu,Ll,Lt,Lm,Lo */
1060 val = 0x3c3; /* GREEK SMALL SIGMA */
1062 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1065 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1067 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
1070 OR (G_UNICODE_UPPERCASE_LETTER,
1071 OR (G_UNICODE_TITLECASE_LETTER,
1074 val = ATTTABLE (c >> 8, c & 0xff);
1076 if (val >= 0x1000000)
1078 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t, 0);
1082 if (t == G_UNICODE_TITLECASE_LETTER)
1085 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
1087 if (title_table[i][0] == c)
1089 val = title_table[i][2];
1095 /* Not all uppercase letters are guaranteed to have a lowercase
1096 * equivalent. If this is the case, val will be zero. */
1097 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
1102 gsize char_len = g_utf8_skip[*(guchar *)last];
1105 memcpy (out_buffer + len, last, char_len);
1117 * @str: a UTF-8 encoded string
1118 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1120 * Converts all Unicode characters in the string that have a case
1121 * to lowercase. The exact manner that this is done depends
1122 * on the current locale, and may result in the number of
1123 * characters in the string changing.
1125 * Return value: a newly allocated string, with all characters
1126 * converted to lowercase.
1129 g_utf8_strdown (const gchar *str,
1133 LocaleType locale_type;
1136 g_return_val_if_fail (str != NULL, NULL);
1138 locale_type = get_locale_type ();
1141 * We use a two pass approach to keep memory management simple
1143 result_len = real_tolower (str, len, NULL, locale_type);
1144 result = g_malloc (result_len + 1);
1145 real_tolower (str, len, result, locale_type);
1146 result[result_len] = '\0';
1153 * @str: a UTF-8 encoded string
1154 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1156 * Converts a string into a form that is independent of case. The
1157 * result will not correspond to any particular case, but can be
1158 * compared for equality or ordered with the results of calling
1159 * g_utf8_casefold() on other strings.
1161 * Note that calling g_utf8_casefold() followed by g_utf8_collate() is
1162 * only an approximation to the correct linguistic case insensitive
1163 * ordering, though it is a fairly good one. Getting this exactly
1164 * right would require a more sophisticated collation function that
1165 * takes case sensitivity into account. GLib does not currently
1166 * provide such a function.
1168 * Return value: a newly allocated string, that is a
1169 * case independent form of @str.
1172 g_utf8_casefold (const gchar *str,
1178 g_return_val_if_fail (str != NULL, NULL);
1180 result = g_string_new (NULL);
1182 while ((len < 0 || p < str + len) && *p)
1184 gunichar ch = g_utf8_get_char (p);
1187 int end = G_N_ELEMENTS (casefold_table);
1189 if (ch >= casefold_table[start].ch &&
1190 ch <= casefold_table[end - 1].ch)
1194 int half = (start + end) / 2;
1195 if (ch == casefold_table[half].ch)
1197 g_string_append (result, casefold_table[half].data);
1200 else if (half == start)
1202 else if (ch > casefold_table[half].ch)
1209 g_string_append_unichar (result, g_unichar_tolower (ch));
1212 p = g_utf8_next_char (p);
1215 return g_string_free (result, FALSE);
1219 * g_unichar_get_mirror_char:
1220 * @ch: a Unicode character
1221 * @mirrored_ch: location to store the mirrored character
1223 * In Unicode, some characters are <firstterm>mirrored</firstterm>. This
1224 * means that their images are mirrored horizontally in text that is laid
1225 * out from right to left. For instance, "(" would become its mirror image,
1226 * ")", in right-to-left text.
1228 * If @ch has the Unicode mirrored property and there is another unicode
1229 * character that typically has a glyph that is the mirror image of @ch's
1230 * glyph and @mirrored_ch is set, it puts that character in the address
1231 * pointed to by @mirrored_ch. Otherwise the original character is put.
1233 * Return value: %TRUE if @ch has a mirrored character, %FALSE otherwise
1238 g_unichar_get_mirror_char (gunichar ch,
1239 gunichar *mirrored_ch)
1244 mirrored = GLIB_GET_MIRRORING(ch);
1246 found = ch != mirrored;
1248 *mirrored_ch = mirrored;
1254 #define G_SCRIPT_TABLE_MIDPOINT (G_N_ELEMENTS (g_script_table) / 2)
1256 static inline GUnicodeScript
1257 g_unichar_get_script_bsearch (gunichar ch)
1260 int upper = G_N_ELEMENTS (g_script_table) - 1;
1261 static int saved_mid = G_SCRIPT_TABLE_MIDPOINT;
1262 int mid = saved_mid;
1267 if (ch < g_script_table[mid].start)
1269 else if (ch >= g_script_table[mid].start + g_script_table[mid].chars)
1272 return g_script_table[saved_mid = mid].script;
1274 mid = (lower + upper) / 2;
1276 while (lower <= upper);
1278 return G_UNICODE_SCRIPT_UNKNOWN;
1282 * g_unichar_get_script:
1283 * @ch: a Unicode character
1285 * Looks up the #GUnicodeScript for a particular character (as defined
1286 * by Unicode Standard Annex #24). No check is made for @ch being a
1287 * valid Unicode character; if you pass in invalid character, the
1288 * result is undefined.
1290 * Return value: the #GUnicodeScript for the character.
1295 g_unichar_get_script (gunichar ch)
1297 if (ch < G_EASY_SCRIPTS_RANGE)
1298 return g_script_easy_table[ch];
1300 return g_unichar_get_script_bsearch (ch);
1304 #define __G_UNIPROP_C__
1305 #include "galiasdef.c"