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 if a character is uppercase.
297 * Return value: %TRUE if @c is an uppercase character
300 g_unichar_isupper (gunichar c)
302 return TYPE (c) == G_UNICODE_UPPERCASE_LETTER;
307 * @c: a Unicode character
309 * Determines if a character is titlecase. Some characters in
310 * Unicode which are composites, such as the DZ digraph
311 * have three case variants instead of just two. The titlecase
312 * form is used at the beginning of a word where only the
313 * first letter is capitalized. The titlecase form of the DZ
314 * digraph is U+01F2 LATIN CAPITAL LETTTER D WITH SMALL LETTER Z.
316 * Return value: %TRUE if the character is titlecase
319 g_unichar_istitle (gunichar c)
322 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
323 if (title_table[i][0] == c)
329 * g_unichar_isxdigit:
330 * @c: a Unicode character.
332 * Determines if a character is a hexidecimal digit.
334 * Return value: %TRUE if the character is a hexadecimal digit
337 g_unichar_isxdigit (gunichar c)
339 return ((c >= 'a' && c <= 'f')
340 || (c >= 'A' && c <= 'F')
341 || (TYPE (c) == G_UNICODE_DECIMAL_NUMBER));
345 * g_unichar_isdefined:
346 * @c: a Unicode character
348 * Determines if a given character is assigned in the Unicode
351 * Return value: %TRUE if the character has an assigned value
354 g_unichar_isdefined (gunichar c)
356 return TYPE (c) != G_UNICODE_UNASSIGNED;
360 * g_unichar_iszerowidth:
361 * @c: a Unicode character
363 * Determines if a given character typically takes zero width when rendered.
364 * The return value is %TRUE for all non-spacing and enclosing marks
365 * (e.g., combining accents), format characters, zero-width
366 * space, but not U+00AD SOFT HYPHEN.
368 * A typical use of this function is with one of g_unichar_iswide() or
369 * g_unichar_iswide_cjk() to determine the number of cells a string occupies
370 * when displayed on a grid display (terminals). However, note that not all
371 * terminals support zero-width rendering of zero-width marks.
373 * Return value: %TRUE if the character has zero width
378 g_unichar_iszerowidth (gunichar c)
380 if (G_UNLIKELY (c == 0x00AD))
383 if (G_UNLIKELY (ISZEROWIDTHTYPE (c)))
386 if (G_UNLIKELY ((c >= 0x1160 && c < 0x1200) ||
395 * @c: a Unicode character
397 * Determines if a character is typically rendered in a double-width
400 * Return value: %TRUE if the character is wide
402 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
404 g_unichar_iswide (gunichar c)
409 return (c <= 0x115f /* Hangul Jamo init. consonants */
410 || c == 0x2329 || c == 0x232a /* angle brackets */
411 || (c >= 0x2e80 && c <= 0xa4cf && (c < 0x302a || c > 0x302f)
412 && c != 0x303f && c != 0x3099 && c!= 0x309a) /* CJK ... Yi */
413 || (c >= 0xac00 && c <= 0xd7a3) /* Hangul Syllables */
414 || (c >= 0xf900 && c <= 0xfaff) /* CJK Compatibility Ideographs */
415 || (c >= 0xfe30 && c <= 0xfe6f) /* CJK Compatibility Forms */
416 || (c >= 0xff00 && c <= 0xff60) /* Fullwidth Forms */
417 || (c >= 0xffe0 && c <= 0xffe6) /* Fullwidth Forms */
418 || (c >= 0x20000 && c <= 0x2fffd) /* CJK extra stuff */
419 || (c >= 0x30000 && c <= 0x3fffd));
429 interval_compare (const void *key, const void *elt)
431 gunichar c = GPOINTER_TO_UINT (key);
432 struct Interval *interval = (struct Interval *)elt;
434 if (c < interval->start)
436 if (c > interval->end)
443 * g_unichar_iswide_cjk:
444 * @c: a Unicode character
446 * Determines if a character is typically rendered in a double-width
447 * cell under legacy East Asian locales. If a character is wide according to
448 * g_unichar_iswide(), then it is also reported wide with this function, but
449 * the converse is not necessarily true. See the
450 * <ulink url="http://www.unicode.org/reports/tr11/">Unicode Standard
451 * Annex #11</ulink> for details.
453 * Return value: %TRUE if the character is wide in legacy East Asian locales
457 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
459 g_unichar_iswide_cjk (gunichar c)
461 /* sorted list of non-overlapping intervals of East Asian Ambiguous
462 * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
463 static const struct Interval ambiguous[] = {
464 { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
465 { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
466 { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
467 { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
468 { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
469 { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
470 { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
471 { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
472 { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
473 { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
474 { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
475 { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
476 { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
477 { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
478 { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
479 { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
480 { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
481 { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
482 { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
483 { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
484 { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
485 { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
486 { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
487 { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
488 { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
489 { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
490 { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
491 { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
492 { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
493 { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
494 { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
495 { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
496 { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
497 { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
498 { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
499 { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
500 { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
501 { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
502 { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
503 { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
504 { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
505 { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
506 { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
507 { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
508 { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
509 { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
510 { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
511 { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
512 { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
513 { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
514 { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
515 { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
518 if (g_unichar_iswide (c))
521 if (bsearch (GUINT_TO_POINTER (c), ambiguous, G_N_ELEMENTS (ambiguous), sizeof ambiguous[0],
531 * @c: a Unicode character
533 * Converts a character to uppercase.
535 * Return value: the result of converting @c to uppercase.
536 * If @c is not an lowercase or titlecase character,
537 * or has no upper case equivalent @c is returned unchanged.
540 g_unichar_toupper (gunichar c)
543 if (t == G_UNICODE_LOWERCASE_LETTER)
545 gunichar val = ATTTABLE (c >> 8, c & 0xff);
546 if (val >= 0x1000000)
548 const gchar *p = special_case_table + val - 0x1000000;
549 return g_utf8_get_char (p);
553 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
554 * do not have an uppercase equivalent, in which case val will be
556 return val ? val : c;
559 else if (t == G_UNICODE_TITLECASE_LETTER)
562 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
564 if (title_table[i][0] == c)
565 return title_table[i][1];
573 * @c: a Unicode character.
575 * Converts a character to lower case.
577 * Return value: the result of converting @c to lower case.
578 * If @c is not an upperlower or titlecase character,
579 * or has no lowercase equivalent @c is returned unchanged.
582 g_unichar_tolower (gunichar c)
585 if (t == G_UNICODE_UPPERCASE_LETTER)
587 gunichar val = ATTTABLE (c >> 8, c & 0xff);
588 if (val >= 0x1000000)
590 const gchar *p = special_case_table + val - 0x1000000;
591 return g_utf8_get_char (p);
595 /* Not all uppercase letters are guaranteed to have a lowercase
596 * equivalent. If this is the case, val will be zero. */
597 return val ? val : c;
600 else if (t == G_UNICODE_TITLECASE_LETTER)
603 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
605 if (title_table[i][0] == c)
606 return title_table[i][2];
614 * @c: a Unicode character
616 * Converts a character to the titlecase.
618 * Return value: the result of converting @c to titlecase.
619 * If @c is not an uppercase or lowercase character,
620 * @c is returned unchanged.
623 g_unichar_totitle (gunichar c)
626 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
628 if (title_table[i][0] == c || title_table[i][1] == c
629 || title_table[i][2] == c)
630 return title_table[i][0];
632 return (TYPE (c) == G_UNICODE_LOWERCASE_LETTER
633 ? ATTTABLE (c >> 8, c & 0xff)
638 * g_unichar_digit_value:
639 * @c: a Unicode character
641 * Determines the numeric value of a character as a decimal
644 * Return value: If @c is a decimal digit (according to
645 * g_unichar_isdigit()), its numeric value. Otherwise, -1.
648 g_unichar_digit_value (gunichar c)
650 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
651 return ATTTABLE (c >> 8, c & 0xff);
656 * g_unichar_xdigit_value:
657 * @c: a Unicode character
659 * Determines the numeric value of a character as a hexidecimal
662 * Return value: If @c is a hex digit (according to
663 * g_unichar_isxdigit()), its numeric value. Otherwise, -1.
666 g_unichar_xdigit_value (gunichar c)
668 if (c >= 'A' && c <= 'F')
670 if (c >= 'a' && c <= 'f')
672 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
673 return ATTTABLE (c >> 8, c & 0xff);
679 * @c: a Unicode character
681 * Classifies a Unicode character by type.
683 * Return value: the type of the character.
686 g_unichar_type (gunichar c)
692 * Case mapping functions
702 get_locale_type (void)
705 char *tem = g_win32_getlocale ();
712 const char *locale = setlocale (LC_CTYPE, NULL);
718 if (locale[1] == 'z')
719 return LOCALE_TURKIC;
722 if (locale[1] == 't')
723 return LOCALE_LITHUANIAN;
726 if (locale[1] == 'r')
727 return LOCALE_TURKIC;
731 return LOCALE_NORMAL;
735 output_marks (const char **p_inout,
739 const char *p = *p_inout;
744 gunichar c = g_utf8_get_char (p);
746 if (ISMARK (TYPE (c)))
748 if (!remove_dot || c != 0x307 /* COMBINING DOT ABOVE */)
749 len += g_unichar_to_utf8 (c, out_buffer ? out_buffer + len : NULL);
750 p = g_utf8_next_char (p);
761 output_special_case (gchar *out_buffer,
766 const gchar *p = special_case_table + offset;
769 if (type != G_UNICODE_TITLECASE_LETTER)
770 p = g_utf8_next_char (p);
777 memcpy (out_buffer, p, len);
783 real_toupper (const gchar *str,
786 LocaleType locale_type)
788 const gchar *p = str;
789 const char *last = NULL;
791 gboolean last_was_i = FALSE;
793 while ((max_len < 0 || p < str + max_len) && *p)
795 gunichar c = g_utf8_get_char (p);
800 p = g_utf8_next_char (p);
802 if (locale_type == LOCALE_LITHUANIAN)
810 /* Nasty, need to remove any dot above. Though
811 * I think only E WITH DOT ABOVE occurs in practice
812 * which could simplify this considerably.
817 decomp = g_unicode_canonical_decomposition (c, &decomp_len);
818 for (i=0; i < decomp_len; i++)
820 if (decomp[i] != 0x307 /* COMBINING DOT ABOVE */)
821 len += g_unichar_to_utf8 (g_unichar_toupper (decomp[i]), out_buffer ? out_buffer + len : NULL);
825 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, TRUE);
835 if (locale_type == LOCALE_TURKIC && c == 'i')
837 /* i => LATIN CAPITAL LETTER I WITH DOT ABOVE */
838 len += g_unichar_to_utf8 (0x130, out_buffer ? out_buffer + len : NULL);
840 else if (c == 0x0345) /* COMBINING GREEK YPOGEGRAMMENI */
842 /* Nasty, need to move it after other combining marks .. this would go away if
843 * we normalized first.
845 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, FALSE);
847 /* And output as GREEK CAPITAL LETTER IOTA */
848 len += g_unichar_to_utf8 (0x399, out_buffer ? out_buffer + len : NULL);
851 OR (G_UNICODE_LOWERCASE_LETTER,
852 OR (G_UNICODE_TITLECASE_LETTER,
855 val = ATTTABLE (c >> 8, c & 0xff);
857 if (val >= 0x1000000)
859 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t,
860 t == G_UNICODE_LOWERCASE_LETTER ? 0 : 1);
864 if (t == G_UNICODE_TITLECASE_LETTER)
867 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
869 if (title_table[i][0] == c)
871 val = title_table[i][1];
877 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
878 * do not have an uppercase equivalent, in which case val will be
880 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
885 gsize char_len = g_utf8_skip[*(guchar *)last];
888 memcpy (out_buffer + len, last, char_len);
900 * @str: a UTF-8 encoded string
901 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
903 * Converts all Unicode characters in the string that have a case
904 * to uppercase. The exact manner that this is done depends
905 * on the current locale, and may result in the number of
906 * characters in the string increasing. (For instance, the
907 * German ess-zet will be changed to SS.)
909 * Return value: a newly allocated string, with all characters
910 * converted to uppercase.
913 g_utf8_strup (const gchar *str,
917 LocaleType locale_type;
920 g_return_val_if_fail (str != NULL, NULL);
922 locale_type = get_locale_type ();
925 * We use a two pass approach to keep memory management simple
927 result_len = real_toupper (str, len, NULL, locale_type);
928 result = g_malloc (result_len + 1);
929 real_toupper (str, len, result, locale_type);
930 result[result_len] = '\0';
935 /* traverses the string checking for characters with combining class == 230
936 * until a base character is found */
938 has_more_above (const gchar *str)
940 const gchar *p = str;
941 gint combining_class;
945 combining_class = _g_unichar_combining_class (g_utf8_get_char (p));
946 if (combining_class == 230)
948 else if (combining_class == 0)
951 p = g_utf8_next_char (p);
958 real_tolower (const gchar *str,
961 LocaleType locale_type)
963 const gchar *p = str;
964 const char *last = NULL;
967 while ((max_len < 0 || p < str + max_len) && *p)
969 gunichar c = g_utf8_get_char (p);
974 p = g_utf8_next_char (p);
976 if (locale_type == LOCALE_TURKIC && c == 'I')
978 if (g_utf8_get_char (p) == 0x0307)
980 /* I + COMBINING DOT ABOVE => i (U+0069) */
981 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
982 p = g_utf8_next_char (p);
986 /* I => LATIN SMALL LETTER DOTLESS I */
987 len += g_unichar_to_utf8 (0x131, out_buffer ? out_buffer + len : NULL);
990 /* Introduce an explicit dot above when lowercasing capital I's and J's
991 * whenever there are more accents above. [SpecialCasing.txt] */
992 else if (locale_type == LOCALE_LITHUANIAN &&
993 (c == 0x00cc || c == 0x00cd || c == 0x0128))
995 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
996 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1001 len += g_unichar_to_utf8 (0x0300, out_buffer ? out_buffer + len : NULL);
1004 len += g_unichar_to_utf8 (0x0301, out_buffer ? out_buffer + len : NULL);
1007 len += g_unichar_to_utf8 (0x0303, out_buffer ? out_buffer + len : NULL);
1011 else if (locale_type == LOCALE_LITHUANIAN &&
1012 (c == 'I' || c == 'J' || c == 0x012e) &&
1015 len += g_unichar_to_utf8 (g_unichar_tolower (c), out_buffer ? out_buffer + len : NULL);
1016 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1018 else if (c == 0x03A3) /* GREEK CAPITAL LETTER SIGMA */
1020 if ((max_len < 0 || p < str + max_len) && *p)
1022 gunichar next_c = g_utf8_get_char (p);
1023 int next_type = TYPE(next_c);
1025 /* SIGMA mapps differently depending on whether it is
1026 * final or not. The following simplified test would
1027 * fail in the case of combining marks following the
1028 * sigma, but I don't think that occurs in real text.
1029 * The test here matches that in ICU.
1031 if (ISALPHA (next_type)) /* Lu,Ll,Lt,Lm,Lo */
1032 val = 0x3c3; /* GREEK SMALL SIGMA */
1034 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1037 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1039 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
1042 OR (G_UNICODE_UPPERCASE_LETTER,
1043 OR (G_UNICODE_TITLECASE_LETTER,
1046 val = ATTTABLE (c >> 8, c & 0xff);
1048 if (val >= 0x1000000)
1050 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t, 0);
1054 if (t == G_UNICODE_TITLECASE_LETTER)
1057 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
1059 if (title_table[i][0] == c)
1061 val = title_table[i][2];
1067 /* Not all uppercase letters are guaranteed to have a lowercase
1068 * equivalent. If this is the case, val will be zero. */
1069 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
1074 gsize char_len = g_utf8_skip[*(guchar *)last];
1077 memcpy (out_buffer + len, last, char_len);
1089 * @str: a UTF-8 encoded string
1090 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1092 * Converts all Unicode characters in the string that have a case
1093 * to lowercase. The exact manner that this is done depends
1094 * on the current locale, and may result in the number of
1095 * characters in the string changing.
1097 * Return value: a newly allocated string, with all characters
1098 * converted to lowercase.
1101 g_utf8_strdown (const gchar *str,
1105 LocaleType locale_type;
1108 g_return_val_if_fail (str != NULL, NULL);
1110 locale_type = get_locale_type ();
1113 * We use a two pass approach to keep memory management simple
1115 result_len = real_tolower (str, len, NULL, locale_type);
1116 result = g_malloc (result_len + 1);
1117 real_tolower (str, len, result, locale_type);
1118 result[result_len] = '\0';
1125 * @str: a UTF-8 encoded string
1126 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1128 * Converts a string into a form that is independent of case. The
1129 * result will not correspond to any particular case, but can be
1130 * compared for equality or ordered with the results of calling
1131 * g_utf8_casefold() on other strings.
1133 * Note that calling g_utf8_casefold() followed by g_utf8_collate() is
1134 * only an approximation to the correct linguistic case insensitive
1135 * ordering, though it is a fairly good one. Getting this exactly
1136 * right would require a more sophisticated collation function that
1137 * takes case sensitivity into account. GLib does not currently
1138 * provide such a function.
1140 * Return value: a newly allocated string, that is a
1141 * case independent form of @str.
1144 g_utf8_casefold (const gchar *str,
1150 g_return_val_if_fail (str != NULL, NULL);
1152 result = g_string_new (NULL);
1154 while ((len < 0 || p < str + len) && *p)
1156 gunichar ch = g_utf8_get_char (p);
1159 int end = G_N_ELEMENTS (casefold_table);
1161 if (ch >= casefold_table[start].ch &&
1162 ch <= casefold_table[end - 1].ch)
1166 int half = (start + end) / 2;
1167 if (ch == casefold_table[half].ch)
1169 g_string_append (result, casefold_table[half].data);
1172 else if (half == start)
1174 else if (ch > casefold_table[half].ch)
1181 g_string_append_unichar (result, g_unichar_tolower (ch));
1184 p = g_utf8_next_char (p);
1187 return g_string_free (result, FALSE);
1191 * g_unichar_get_mirror_char:
1192 * @ch: a Unicode character
1193 * @mirrored_ch: location to store the mirrored character
1195 * In Unicode, some characters are <firstterm>mirrored</firstterm>. This
1196 * means that their images are mirrored horizontally in text that is laid
1197 * out from right to left. For instance, "(" would become its mirror image,
1198 * ")", in right-to-left text.
1200 * If @ch has the Unicode mirrored property and there is another unicode
1201 * character that typically has a glyph that is the mirror image of @ch's
1202 * glyph and @mirrored_ch is set, it puts that character in the address
1203 * pointed to by @mirrored_ch. Otherwise the original character is put.
1205 * Return value: %TRUE if @ch has a mirrored character, %FALSE otherwise
1210 g_unichar_get_mirror_char (gunichar ch,
1211 gunichar *mirrored_ch)
1216 mirrored = GLIB_GET_MIRRORING(ch);
1218 found = ch != mirrored;
1220 *mirrored_ch = mirrored;
1226 #define G_SCRIPT_TABLE_MIDPOINT (G_N_ELEMENTS (g_script_table) / 2)
1228 static inline GUnicodeScript
1229 g_unichar_get_script_bsearch (gunichar ch)
1232 int upper = G_N_ELEMENTS (g_script_table) - 1;
1233 static int saved_mid = G_SCRIPT_TABLE_MIDPOINT;
1234 int mid = saved_mid;
1239 if (ch < g_script_table[mid].start)
1241 else if (ch >= g_script_table[mid].start + g_script_table[mid].chars)
1244 return g_script_table[saved_mid = mid].script;
1246 mid = (lower + upper) / 2;
1248 while (lower <= upper);
1250 return G_UNICODE_SCRIPT_UNKNOWN;
1254 * g_unichar_get_script:
1255 * @ch: a Unicode character
1257 * Looks up the #GUnicodeScript for a particular character (as defined
1258 * by Unicode Standard Annex #24). No check is made for @ch being a
1259 * valid Unicode character; if you pass in invalid character, the
1260 * result is undefined.
1262 * Return value: the #GUnicodeScript for the character.
1267 g_unichar_get_script (gunichar ch)
1269 if (ch < G_EASY_SCRIPTS_RANGE)
1270 return g_script_easy_table[ch];
1272 return g_unichar_get_script_bsearch (ch);
1276 #define __G_UNIPROP_C__
1277 #include "galiasdef.c"