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 "gunicodeprivate.h"
35 #define ATTR_TABLE(Page) (((Page) <= G_UNICODE_LAST_PAGE_PART1) \
36 ? attr_table_part1[Page] \
37 : attr_table_part2[(Page) - 0xe00])
39 #define ATTTABLE(Page, Char) \
40 ((ATTR_TABLE(Page) == G_UNICODE_MAX_TABLE_INDEX) ? 0 : (attr_data[ATTR_TABLE(Page)][Char]))
42 #define TTYPE_PART1(Page, Char) \
43 ((type_table_part1[Page] >= G_UNICODE_MAX_TABLE_INDEX) \
44 ? (type_table_part1[Page] - G_UNICODE_MAX_TABLE_INDEX) \
45 : (type_data[type_table_part1[Page]][Char]))
47 #define TTYPE_PART2(Page, Char) \
48 ((type_table_part2[Page] >= G_UNICODE_MAX_TABLE_INDEX) \
49 ? (type_table_part2[Page] - G_UNICODE_MAX_TABLE_INDEX) \
50 : (type_data[type_table_part2[Page]][Char]))
53 (((Char) <= G_UNICODE_LAST_CHAR_PART1) \
54 ? TTYPE_PART1 ((Char) >> 8, (Char) & 0xff) \
55 : (((Char) >= 0xe0000 && (Char) <= G_UNICODE_LAST_CHAR) \
56 ? TTYPE_PART2 (((Char) - 0xe0000) >> 8, (Char) & 0xff) \
57 : G_UNICODE_UNASSIGNED))
60 #define IS(Type, Class) (((guint)1 << (Type)) & (Class))
61 #define OR(Type, Rest) (((guint)1 << (Type)) | (Rest))
65 #define ISDIGIT(Type) IS ((Type), \
66 OR (G_UNICODE_DECIMAL_NUMBER, \
67 OR (G_UNICODE_LETTER_NUMBER, \
68 OR (G_UNICODE_OTHER_NUMBER, 0))))
70 #define ISALPHA(Type) IS ((Type), \
71 OR (G_UNICODE_LOWERCASE_LETTER, \
72 OR (G_UNICODE_UPPERCASE_LETTER, \
73 OR (G_UNICODE_TITLECASE_LETTER, \
74 OR (G_UNICODE_MODIFIER_LETTER, \
75 OR (G_UNICODE_OTHER_LETTER, 0))))))
77 #define ISALDIGIT(Type) IS ((Type), \
78 OR (G_UNICODE_DECIMAL_NUMBER, \
79 OR (G_UNICODE_LETTER_NUMBER, \
80 OR (G_UNICODE_OTHER_NUMBER, \
81 OR (G_UNICODE_LOWERCASE_LETTER, \
82 OR (G_UNICODE_UPPERCASE_LETTER, \
83 OR (G_UNICODE_TITLECASE_LETTER, \
84 OR (G_UNICODE_MODIFIER_LETTER, \
85 OR (G_UNICODE_OTHER_LETTER, 0)))))))))
87 #define ISMARK(Type) IS ((Type), \
88 OR (G_UNICODE_NON_SPACING_MARK, \
89 OR (G_UNICODE_COMBINING_MARK, \
90 OR (G_UNICODE_ENCLOSING_MARK, 0))))
94 * @c: a Unicode character
96 * Determines whether a character is alphanumeric.
97 * Given some UTF-8 text, obtain a character value
98 * with g_utf8_get_char().
100 * Return value: %TRUE if @c is an alphanumeric character
103 g_unichar_isalnum (gunichar c)
105 return ISALDIGIT (TYPE (c)) ? TRUE : FALSE;
110 * @c: a Unicode character
112 * Determines whether a character is alphabetic (i.e. a letter).
113 * Given some UTF-8 text, obtain a character value with
116 * Return value: %TRUE if @c is an alphabetic character
119 g_unichar_isalpha (gunichar c)
121 return ISALPHA (TYPE (c)) ? TRUE : FALSE;
127 * @c: a Unicode character
129 * Determines whether a character is a control character.
130 * Given some UTF-8 text, obtain a character value with
133 * Return value: %TRUE if @c is a control character
136 g_unichar_iscntrl (gunichar c)
138 return TYPE (c) == G_UNICODE_CONTROL;
143 * @c: a Unicode character
145 * Determines whether a character is numeric (i.e. a digit). This
146 * covers ASCII 0-9 and also digits in other languages/scripts. Given
147 * some UTF-8 text, obtain a character value with g_utf8_get_char().
149 * Return value: %TRUE if @c is a digit
152 g_unichar_isdigit (gunichar c)
154 return TYPE (c) == G_UNICODE_DECIMAL_NUMBER;
160 * @c: a Unicode character
162 * Determines whether a character is printable and not a space
163 * (returns %FALSE for control characters, format characters, and
164 * spaces). g_unichar_isprint() is similar, but returns %TRUE for
165 * spaces. Given some UTF-8 text, obtain a character value with
168 * Return value: %TRUE if @c is printable unless it's a space
171 g_unichar_isgraph (gunichar c)
174 OR (G_UNICODE_CONTROL,
175 OR (G_UNICODE_FORMAT,
176 OR (G_UNICODE_UNASSIGNED,
177 OR (G_UNICODE_PRIVATE_USE,
178 OR (G_UNICODE_SURROGATE,
179 OR (G_UNICODE_SPACE_SEPARATOR,
185 * @c: a Unicode character
187 * Determines whether a character is a lowercase letter.
188 * Given some UTF-8 text, obtain a character value with
191 * Return value: %TRUE if @c is a lowercase letter
194 g_unichar_islower (gunichar c)
196 return TYPE (c) == G_UNICODE_LOWERCASE_LETTER;
202 * @c: a Unicode character
204 * Determines whether a character is printable.
205 * Unlike g_unichar_isgraph(), returns %TRUE for spaces.
206 * Given some UTF-8 text, obtain a character value with
209 * Return value: %TRUE if @c is printable
212 g_unichar_isprint (gunichar c)
215 OR (G_UNICODE_CONTROL,
216 OR (G_UNICODE_FORMAT,
217 OR (G_UNICODE_UNASSIGNED,
218 OR (G_UNICODE_PRIVATE_USE,
219 OR (G_UNICODE_SURROGATE,
225 * @c: a Unicode character
227 * Determines whether a character is punctuation or a symbol.
228 * Given some UTF-8 text, obtain a character value with
231 * Return value: %TRUE if @c is a punctuation or symbol character
234 g_unichar_ispunct (gunichar c)
237 OR (G_UNICODE_CONNECT_PUNCTUATION,
238 OR (G_UNICODE_DASH_PUNCTUATION,
239 OR (G_UNICODE_CLOSE_PUNCTUATION,
240 OR (G_UNICODE_FINAL_PUNCTUATION,
241 OR (G_UNICODE_INITIAL_PUNCTUATION,
242 OR (G_UNICODE_OTHER_PUNCTUATION,
243 OR (G_UNICODE_OPEN_PUNCTUATION,
244 OR (G_UNICODE_CURRENCY_SYMBOL,
245 OR (G_UNICODE_MODIFIER_SYMBOL,
246 OR (G_UNICODE_MATH_SYMBOL,
247 OR (G_UNICODE_OTHER_SYMBOL,
248 0)))))))))))) ? TRUE : FALSE;
253 * @c: a Unicode character
255 * Determines whether a character is a space, tab, or line separator
256 * (newline, carriage return, etc.). Given some UTF-8 text, obtain a
257 * character value with g_utf8_get_char().
259 * (Note: don't use this to do word breaking; you have to use
260 * Pango or equivalent to get word breaking right, the algorithm
261 * is fairly complex.)
263 * Return value: %TRUE if @c is a space character
266 g_unichar_isspace (gunichar c)
270 /* special-case these since Unicode thinks they are not spaces */
281 OR (G_UNICODE_SPACE_SEPARATOR,
282 OR (G_UNICODE_LINE_SEPARATOR,
283 OR (G_UNICODE_PARAGRAPH_SEPARATOR,
284 0)))) ? TRUE : FALSE;
292 * @c: a Unicode character
294 * Determines if a character is uppercase.
296 * Return value: %TRUE if @c is an uppercase character
299 g_unichar_isupper (gunichar c)
301 return TYPE (c) == G_UNICODE_UPPERCASE_LETTER;
306 * @c: a Unicode character
308 * Determines if a character is titlecase. Some characters in
309 * Unicode which are composites, such as the DZ digraph
310 * have three case variants instead of just two. The titlecase
311 * form is used at the beginning of a word where only the
312 * first letter is capitalized. The titlecase form of the DZ
313 * digraph is U+01F2 LATIN CAPITAL LETTTER D WITH SMALL LETTER Z.
315 * Return value: %TRUE if the character is titlecase
318 g_unichar_istitle (gunichar c)
321 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
322 if (title_table[i][0] == c)
328 * g_unichar_isxdigit:
329 * @c: a Unicode character.
331 * Determines if a character is a hexidecimal digit.
333 * Return value: %TRUE if the character is a hexadecimal digit
336 g_unichar_isxdigit (gunichar c)
338 return ((c >= 'a' && c <= 'f')
339 || (c >= 'A' && c <= 'F')
340 || ISDIGIT (TYPE (c)));
344 * g_unichar_isdefined:
345 * @c: a Unicode character
347 * Determines if a given character is assigned in the Unicode
350 * Return value: %TRUE if the character has an assigned value
353 g_unichar_isdefined (gunichar c)
355 return TYPE (c) != G_UNICODE_UNASSIGNED;
360 * @c: a Unicode character
362 * Determines if a character is typically rendered in a double-width
365 * Return value: %TRUE if the character is wide
367 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
369 g_unichar_iswide (gunichar c)
374 return (c <= 0x115f /* Hangul Jamo init. consonants */
375 || c == 0x2329 || c == 0x232a /* angle brackets */
376 || (c >= 0x2e80 && c <= 0xa4cf && (c < 0x302a || c > 0x302f)
377 && c != 0x303f && c != 0x3099 && c!= 0x309a) /* CJK ... Yi */
378 || (c >= 0xac00 && c <= 0xd7a3) /* Hangul Syllables */
379 || (c >= 0xf900 && c <= 0xfaff) /* CJK Compatibility Ideographs */
380 || (c >= 0xfe30 && c <= 0xfe6f) /* CJK Compatibility Forms */
381 || (c >= 0xff00 && c <= 0xff60) /* Fullwidth Forms */
382 || (c >= 0xffe0 && c <= 0xffe6) /* Fullwidth Forms */
383 || (c >= 0x20000 && c <= 0x2fffd) /* CJK extra stuff */
384 || (c >= 0x30000 && c <= 0x3fffd));
394 interval_compare (const void *key, const void *elt)
396 gunichar c = GPOINTER_TO_UINT (key);
397 struct Interval *interval = elt;
399 if (c < interval->start)
401 if (c > interval->end)
408 * g_unichar_iswide_cjk:
409 * @c: a Unicode character
411 * Determines if a character is typically rendered in a double-width
412 * cell under legacy East Asian locales. If a character is wide according to
413 * g_unichar_iswide(), then it is also reported wide with this function, but
414 * the converse is not necessarily true. See the
415 * <ulink url="http://www.unicode.org/reports/tr11/">Unicode Standard
416 * Annex #11</ulink> for details.
418 * Return value: %TRUE if the character is wide in legacy East Asian locales
422 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
424 g_unichar_iswide_cjk (gunichar c)
426 /* sorted list of non-overlapping intervals of East Asian Ambiguous
427 * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
428 static const struct Interval ambiguous[] = {
429 { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
430 { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
431 { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
432 { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
433 { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
434 { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
435 { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
436 { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
437 { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
438 { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
439 { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
440 { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
441 { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
442 { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
443 { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
444 { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
445 { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
446 { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
447 { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
448 { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
449 { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
450 { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
451 { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
452 { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
453 { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
454 { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
455 { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
456 { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
457 { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
458 { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
459 { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
460 { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
461 { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
462 { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
463 { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
464 { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
465 { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
466 { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
467 { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
468 { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
469 { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
470 { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
471 { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
472 { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
473 { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
474 { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
475 { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
476 { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
477 { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
478 { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
479 { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
480 { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
483 if (g_unichar_iswide (c))
486 if (bsearch (GUINT_TO_POINTER (c), ambiguous, G_N_ELEMENTS (ambiguous), sizeof ambiguous[0],
496 * @c: a Unicode character
498 * Converts a character to uppercase.
500 * Return value: the result of converting @c to uppercase.
501 * If @c is not an lowercase or titlecase character,
502 * or has no upper case equivalent @c is returned unchanged.
505 g_unichar_toupper (gunichar c)
508 if (t == G_UNICODE_LOWERCASE_LETTER)
510 gunichar val = ATTTABLE (c >> 8, c & 0xff);
511 if (val >= 0x1000000)
513 const gchar *p = special_case_table + val - 0x1000000;
514 return g_utf8_get_char (p);
517 return val ? val : c;
519 else if (t == G_UNICODE_TITLECASE_LETTER)
522 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
524 if (title_table[i][0] == c)
525 return title_table[i][1];
533 * @c: a Unicode character.
535 * Converts a character to lower case.
537 * Return value: the result of converting @c to lower case.
538 * If @c is not an upperlower or titlecase character,
539 * or has no lowercase equivalent @c is returned unchanged.
542 g_unichar_tolower (gunichar c)
545 if (t == G_UNICODE_UPPERCASE_LETTER)
547 gunichar val = ATTTABLE (c >> 8, c & 0xff);
548 if (val >= 0x1000000)
550 const gchar *p = special_case_table + val - 0x1000000;
551 return g_utf8_get_char (p);
554 return val ? val : c;
556 else if (t == G_UNICODE_TITLECASE_LETTER)
559 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
561 if (title_table[i][0] == c)
562 return title_table[i][2];
570 * @c: a Unicode character
572 * Converts a character to the titlecase.
574 * Return value: the result of converting @c to titlecase.
575 * If @c is not an uppercase or lowercase character,
576 * @c is returned unchanged.
579 g_unichar_totitle (gunichar c)
582 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
584 if (title_table[i][0] == c || title_table[i][1] == c
585 || title_table[i][2] == c)
586 return title_table[i][0];
588 return (TYPE (c) == G_UNICODE_LOWERCASE_LETTER
589 ? ATTTABLE (c >> 8, c & 0xff)
594 * g_unichar_digit_value:
595 * @c: a Unicode character
597 * Determines the numeric value of a character as a decimal
600 * Return value: If @c is a decimal digit (according to
601 * g_unichar_isdigit()), its numeric value. Otherwise, -1.
604 g_unichar_digit_value (gunichar c)
606 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
607 return ATTTABLE (c >> 8, c & 0xff);
612 * g_unichar_xdigit_value:
613 * @c: a Unicode character
615 * Determines the numeric value of a character as a hexidecimal
618 * Return value: If @c is a hex digit (according to
619 * g_unichar_isxdigit()), its numeric value. Otherwise, -1.
622 g_unichar_xdigit_value (gunichar c)
624 if (c >= 'A' && c <= 'F')
626 if (c >= 'a' && c <= 'f')
628 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
629 return ATTTABLE (c >> 8, c & 0xff);
635 * @c: a Unicode character
637 * Classifies a Unicode character by type.
639 * Return value: the type of the character.
642 g_unichar_type (gunichar c)
648 * Case mapping functions
658 get_locale_type (void)
661 char *tem = g_win32_getlocale ();
668 const char *locale = setlocale (LC_CTYPE, NULL);
674 if (locale[1] == 'z')
675 return LOCALE_TURKIC;
678 if (locale[1] == 't')
679 return LOCALE_LITHUANIAN;
682 if (locale[1] == 'r')
683 return LOCALE_TURKIC;
687 return LOCALE_NORMAL;
691 output_marks (const char **p_inout,
695 const char *p = *p_inout;
700 gunichar c = g_utf8_get_char (p);
702 if (ISMARK (TYPE (c)))
704 if (!remove_dot || c != 0x307 /* COMBINING DOT ABOVE */)
705 len += g_unichar_to_utf8 (c, out_buffer ? out_buffer + len : NULL);
706 p = g_utf8_next_char (p);
717 output_special_case (gchar *out_buffer,
722 const gchar *p = special_case_table + offset;
725 if (type != G_UNICODE_TITLECASE_LETTER)
726 p = g_utf8_next_char (p);
733 memcpy (out_buffer, p, len);
739 real_toupper (const gchar *str,
742 LocaleType locale_type)
744 const gchar *p = str;
745 const char *last = NULL;
747 gboolean last_was_i = FALSE;
749 while ((max_len < 0 || p < str + max_len) && *p)
751 gunichar c = g_utf8_get_char (p);
756 p = g_utf8_next_char (p);
758 if (locale_type == LOCALE_LITHUANIAN)
766 /* Nasty, need to remove any dot above. Though
767 * I think only E WITH DOT ABOVE occurs in practice
768 * which could simplify this considerably.
773 decomp = g_unicode_canonical_decomposition (c, &decomp_len);
774 for (i=0; i < decomp_len; i++)
776 if (decomp[i] != 0x307 /* COMBINING DOT ABOVE */)
777 len += g_unichar_to_utf8 (g_unichar_toupper (decomp[i]), out_buffer ? out_buffer + len : NULL);
781 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, TRUE);
791 if (locale_type == LOCALE_TURKIC && c == 'i')
793 /* i => LATIN CAPITAL LETTER I WITH DOT ABOVE */
794 len += g_unichar_to_utf8 (0x130, out_buffer ? out_buffer + len : NULL);
796 else if (c == 0x0345) /* COMBINING GREEK YPOGEGRAMMENI */
798 /* Nasty, need to move it after other combining marks .. this would go away if
799 * we normalized first.
801 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, FALSE);
803 /* And output as GREEK CAPITAL LETTER IOTA */
804 len += g_unichar_to_utf8 (0x399, out_buffer ? out_buffer + len : NULL);
807 OR (G_UNICODE_LOWERCASE_LETTER,
808 OR (G_UNICODE_TITLECASE_LETTER,
811 val = ATTTABLE (c >> 8, c & 0xff);
813 if (val >= 0x1000000)
815 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t,
816 t == G_UNICODE_LOWERCASE_LETTER ? 0 : 1);
820 if (t == G_UNICODE_TITLECASE_LETTER)
823 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
825 if (title_table[i][0] == c)
827 val = title_table[i][1];
833 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
838 gsize char_len = g_utf8_skip[*(guchar *)last];
841 memcpy (out_buffer + len, last, char_len);
853 * @str: a UTF-8 encoded string
854 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
856 * Converts all Unicode characters in the string that have a case
857 * to uppercase. The exact manner that this is done depends
858 * on the current locale, and may result in the number of
859 * characters in the string increasing. (For instance, the
860 * German ess-zet will be changed to SS.)
862 * Return value: a newly allocated string, with all characters
863 * converted to uppercase.
866 g_utf8_strup (const gchar *str,
870 LocaleType locale_type;
873 g_return_val_if_fail (str != NULL, NULL);
875 locale_type = get_locale_type ();
878 * We use a two pass approach to keep memory management simple
880 result_len = real_toupper (str, len, NULL, locale_type);
881 result = g_malloc (result_len + 1);
882 real_toupper (str, len, result, locale_type);
883 result[result_len] = '\0';
888 /* traverses the string checking for characters with combining class == 230
889 * until a base character is found */
891 has_more_above (const gchar *str)
893 const gchar *p = str;
894 gint combining_class;
898 combining_class = _g_unichar_combining_class (g_utf8_get_char (p));
899 if (combining_class == 230)
901 else if (combining_class == 0)
904 p = g_utf8_next_char (p);
911 real_tolower (const gchar *str,
914 LocaleType locale_type)
916 const gchar *p = str;
917 const char *last = NULL;
920 while ((max_len < 0 || p < str + max_len) && *p)
922 gunichar c = g_utf8_get_char (p);
927 p = g_utf8_next_char (p);
929 if (locale_type == LOCALE_TURKIC && c == 'I')
931 if (g_utf8_get_char (p) == 0x0307)
933 /* I + COMBINING DOT ABOVE => i (U+0069) */
934 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
935 p = g_utf8_next_char (p);
939 /* I => LATIN SMALL LETTER DOTLESS I */
940 len += g_unichar_to_utf8 (0x131, out_buffer ? out_buffer + len : NULL);
943 /* Introduce an explicit dot above when lowercasing capital I's and J's
944 * whenever there are more accents above. [SpecialCasing.txt] */
945 else if (locale_type == LOCALE_LITHUANIAN &&
946 (c == 0x00cc || c == 0x00cd || c == 0x0128))
948 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
949 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
954 len += g_unichar_to_utf8 (0x0300, out_buffer ? out_buffer + len : NULL);
957 len += g_unichar_to_utf8 (0x0301, out_buffer ? out_buffer + len : NULL);
960 len += g_unichar_to_utf8 (0x0303, out_buffer ? out_buffer + len : NULL);
964 else if (locale_type == LOCALE_LITHUANIAN &&
965 (c == 'I' || c == 'J' || c == 0x012e) &&
968 len += g_unichar_to_utf8 (g_unichar_tolower (c), out_buffer ? out_buffer + len : NULL);
969 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
971 else if (c == 0x03A3) /* GREEK CAPITAL LETTER SIGMA */
973 if ((max_len < 0 || p < str + max_len) && *p)
975 gunichar next_c = g_utf8_get_char (p);
976 int next_type = TYPE(next_c);
978 /* SIGMA mapps differently depending on whether it is
979 * final or not. The following simplified test would
980 * fail in the case of combining marks following the
981 * sigma, but I don't think that occurs in real text.
982 * The test here matches that in ICU.
984 if (ISALPHA (next_type)) /* Lu,Ll,Lt,Lm,Lo */
985 val = 0x3c3; /* GREEK SMALL SIGMA */
987 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
990 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
992 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
995 OR (G_UNICODE_UPPERCASE_LETTER,
996 OR (G_UNICODE_TITLECASE_LETTER,
999 val = ATTTABLE (c >> 8, c & 0xff);
1001 if (val >= 0x1000000)
1003 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t, 0);
1007 if (t == G_UNICODE_TITLECASE_LETTER)
1010 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
1012 if (title_table[i][0] == c)
1014 val = title_table[i][2];
1020 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
1025 gsize char_len = g_utf8_skip[*(guchar *)last];
1028 memcpy (out_buffer + len, last, char_len);
1040 * @str: a UTF-8 encoded string
1041 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1043 * Converts all Unicode characters in the string that have a case
1044 * to lowercase. The exact manner that this is done depends
1045 * on the current locale, and may result in the number of
1046 * characters in the string changing.
1048 * Return value: a newly allocated string, with all characters
1049 * converted to lowercase.
1052 g_utf8_strdown (const gchar *str,
1056 LocaleType locale_type;
1059 g_return_val_if_fail (str != NULL, NULL);
1061 locale_type = get_locale_type ();
1064 * We use a two pass approach to keep memory management simple
1066 result_len = real_tolower (str, len, NULL, locale_type);
1067 result = g_malloc (result_len + 1);
1068 real_tolower (str, len, result, locale_type);
1069 result[result_len] = '\0';
1076 * @str: a UTF-8 encoded string
1077 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1079 * Converts a string into a form that is independent of case. The
1080 * result will not correspond to any particular case, but can be
1081 * compared for equality or ordered with the results of calling
1082 * g_utf8_casefold() on other strings.
1084 * Note that calling g_utf8_casefold() followed by g_utf8_collate() is
1085 * only an approximation to the correct linguistic case insensitive
1086 * ordering, though it is a fairly good one. Getting this exactly
1087 * right would require a more sophisticated collation function that
1088 * takes case sensitivity into account. GLib does not currently
1089 * provide such a function.
1091 * Return value: a newly allocated string, that is a
1092 * case independent form of @str.
1095 g_utf8_casefold (const gchar *str,
1101 g_return_val_if_fail (str != NULL, NULL);
1103 result = g_string_new (NULL);
1105 while ((len < 0 || p < str + len) && *p)
1107 gunichar ch = g_utf8_get_char (p);
1110 int end = G_N_ELEMENTS (casefold_table);
1112 if (ch >= casefold_table[start].ch &&
1113 ch <= casefold_table[end - 1].ch)
1117 int half = (start + end) / 2;
1118 if (ch == casefold_table[half].ch)
1120 g_string_append (result, casefold_table[half].data);
1123 else if (half == start)
1125 else if (ch > casefold_table[half].ch)
1132 g_string_append_unichar (result, g_unichar_tolower (ch));
1135 p = g_utf8_next_char (p);
1138 return g_string_free (result, FALSE);
1142 * g_unichar_get_mirror_char:
1143 * @ch: a Unicode character
1144 * @mirrored_ch: location to store the mirrored character
1146 * In Unicode, some characters are <firstterm>mirrored</firstterm>. This
1147 * means that their images are mirrored horizontally in text that is laid
1148 * out from right to left. For instance, "(" would become its mirror image,
1149 * ")", in right-to-left text.
1151 * If @ch has the Unicode mirrored property and there is another unicode
1152 * character that typically has a glyph that is the mirror image of @ch's
1153 * glyph and @mirrored_ch is set, it puts that character in the address
1154 * pointed to by @mirrored_ch. Otherwise the original character is put.
1156 * Return value: %TRUE if @ch has a mirrored character, %FALSE otherwise
1161 g_unichar_get_mirror_char (gunichar ch,
1162 gunichar *mirrored_ch)
1167 mirrored = GLIB_GET_MIRRORING(ch);
1169 found = ch != mirrored;
1171 *mirrored_ch = mirrored;
1177 #define __G_UNIPROP_C__
1178 #include "galiasdef.c"