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
426 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
428 g_unichar_iswide (gunichar c)
433 return (c <= 0x115f /* Hangul Jamo init. consonants */
434 || c == 0x2329 || c == 0x232a /* angle brackets */
435 || (c >= 0x2e80 && c <= 0xa4cf && (c < 0x302a || c > 0x302f)
436 && c != 0x303f && c != 0x3099 && c!= 0x309a) /* CJK ... Yi */
437 || (c >= 0xac00 && c <= 0xd7a3) /* Hangul Syllables */
438 || (c >= 0xf900 && c <= 0xfaff) /* CJK Compatibility Ideographs */
439 || (c >= 0xfe30 && c <= 0xfe6f) /* CJK Compatibility Forms */
440 || (c >= 0xff00 && c <= 0xff60) /* Fullwidth Forms */
441 || (c >= 0xffe0 && c <= 0xffe6) /* Fullwidth Forms */
442 || (c >= 0x20000 && c <= 0x2fffd) /* CJK extra stuff */
443 || (c >= 0x30000 && c <= 0x3fffd));
453 interval_compare (const void *key, const void *elt)
455 gunichar c = GPOINTER_TO_UINT (key);
456 struct Interval *interval = (struct Interval *)elt;
458 if (c < interval->start)
460 if (c > interval->end)
467 * g_unichar_iswide_cjk:
468 * @c: a Unicode character
470 * Determines if a character is typically rendered in a double-width
471 * cell under legacy East Asian locales. If a character is wide according to
472 * g_unichar_iswide(), then it is also reported wide with this function, but
473 * the converse is not necessarily true. See the
474 * <ulink url="http://www.unicode.org/reports/tr11/">Unicode Standard
475 * Annex #11</ulink> for details.
477 * Return value: %TRUE if the character is wide in legacy East Asian locales
481 /* This function stolen from Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk>. */
483 g_unichar_iswide_cjk (gunichar c)
485 /* sorted list of non-overlapping intervals of East Asian Ambiguous
486 * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
487 static const struct Interval ambiguous[] = {
488 { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
489 { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
490 { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
491 { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
492 { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
493 { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
494 { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
495 { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
496 { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
497 { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
498 { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
499 { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
500 { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
501 { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
502 { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
503 { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
504 { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
505 { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
506 { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
507 { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
508 { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
509 { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
510 { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
511 { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
512 { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
513 { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
514 { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
515 { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
516 { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
517 { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
518 { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
519 { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
520 { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
521 { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
522 { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
523 { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
524 { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
525 { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
526 { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
527 { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
528 { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
529 { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
530 { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
531 { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
532 { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
533 { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
534 { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
535 { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
536 { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
537 { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
538 { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
539 { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
542 if (g_unichar_iswide (c))
545 if (bsearch (GUINT_TO_POINTER (c), ambiguous, G_N_ELEMENTS (ambiguous), sizeof ambiguous[0],
555 * @c: a Unicode character
557 * Converts a character to uppercase.
559 * Return value: the result of converting @c to uppercase.
560 * If @c is not an lowercase or titlecase character,
561 * or has no upper case equivalent @c is returned unchanged.
564 g_unichar_toupper (gunichar c)
567 if (t == G_UNICODE_LOWERCASE_LETTER)
569 gunichar val = ATTTABLE (c >> 8, c & 0xff);
570 if (val >= 0x1000000)
572 const gchar *p = special_case_table + val - 0x1000000;
573 val = g_utf8_get_char (p);
575 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
576 * do not have an uppercase equivalent, in which case val will be
579 return val ? val : c;
581 else if (t == G_UNICODE_TITLECASE_LETTER)
584 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
586 if (title_table[i][0] == c)
587 return title_table[i][1];
595 * @c: a Unicode character.
597 * Converts a character to lower case.
599 * Return value: the result of converting @c to lower case.
600 * If @c is not an upperlower or titlecase character,
601 * or has no lowercase equivalent @c is returned unchanged.
604 g_unichar_tolower (gunichar c)
607 if (t == G_UNICODE_UPPERCASE_LETTER)
609 gunichar val = ATTTABLE (c >> 8, c & 0xff);
610 if (val >= 0x1000000)
612 const gchar *p = special_case_table + val - 0x1000000;
613 return g_utf8_get_char (p);
617 /* Not all uppercase letters are guaranteed to have a lowercase
618 * equivalent. If this is the case, val will be zero. */
619 return val ? val : c;
622 else if (t == G_UNICODE_TITLECASE_LETTER)
625 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
627 if (title_table[i][0] == c)
628 return title_table[i][2];
636 * @c: a Unicode character
638 * Converts a character to the titlecase.
640 * Return value: the result of converting @c to titlecase.
641 * If @c is not an uppercase or lowercase character,
642 * @c is returned unchanged.
645 g_unichar_totitle (gunichar c)
648 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
650 if (title_table[i][0] == c || title_table[i][1] == c
651 || title_table[i][2] == c)
652 return title_table[i][0];
655 if (TYPE (c) == G_UNICODE_LOWERCASE_LETTER)
656 return g_unichar_toupper (c);
662 * g_unichar_digit_value:
663 * @c: a Unicode character
665 * Determines the numeric value of a character as a decimal
668 * Return value: If @c is a decimal digit (according to
669 * g_unichar_isdigit()), its numeric value. Otherwise, -1.
672 g_unichar_digit_value (gunichar c)
674 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
675 return ATTTABLE (c >> 8, c & 0xff);
680 * g_unichar_xdigit_value:
681 * @c: a Unicode character
683 * Determines the numeric value of a character as a hexidecimal
686 * Return value: If @c is a hex digit (according to
687 * g_unichar_isxdigit()), its numeric value. Otherwise, -1.
690 g_unichar_xdigit_value (gunichar c)
692 if (c >= 'A' && c <= 'F')
694 if (c >= 'a' && c <= 'f')
696 if (TYPE (c) == G_UNICODE_DECIMAL_NUMBER)
697 return ATTTABLE (c >> 8, c & 0xff);
703 * @c: a Unicode character
705 * Classifies a Unicode character by type.
707 * Return value: the type of the character.
710 g_unichar_type (gunichar c)
716 * Case mapping functions
726 get_locale_type (void)
729 char *tem = g_win32_getlocale ();
736 const char *locale = setlocale (LC_CTYPE, NULL);
742 if (locale[1] == 'z')
743 return LOCALE_TURKIC;
746 if (locale[1] == 't')
747 return LOCALE_LITHUANIAN;
750 if (locale[1] == 'r')
751 return LOCALE_TURKIC;
755 return LOCALE_NORMAL;
759 output_marks (const char **p_inout,
763 const char *p = *p_inout;
768 gunichar c = g_utf8_get_char (p);
770 if (ISMARK (TYPE (c)))
772 if (!remove_dot || c != 0x307 /* COMBINING DOT ABOVE */)
773 len += g_unichar_to_utf8 (c, out_buffer ? out_buffer + len : NULL);
774 p = g_utf8_next_char (p);
785 output_special_case (gchar *out_buffer,
790 const gchar *p = special_case_table + offset;
793 if (type != G_UNICODE_TITLECASE_LETTER)
794 p = g_utf8_next_char (p);
801 memcpy (out_buffer, p, len);
807 real_toupper (const gchar *str,
810 LocaleType locale_type)
812 const gchar *p = str;
813 const char *last = NULL;
815 gboolean last_was_i = FALSE;
817 while ((max_len < 0 || p < str + max_len) && *p)
819 gunichar c = g_utf8_get_char (p);
824 p = g_utf8_next_char (p);
826 if (locale_type == LOCALE_LITHUANIAN)
834 /* Nasty, need to remove any dot above. Though
835 * I think only E WITH DOT ABOVE occurs in practice
836 * which could simplify this considerably.
841 decomp = g_unicode_canonical_decomposition (c, &decomp_len);
842 for (i=0; i < decomp_len; i++)
844 if (decomp[i] != 0x307 /* COMBINING DOT ABOVE */)
845 len += g_unichar_to_utf8 (g_unichar_toupper (decomp[i]), out_buffer ? out_buffer + len : NULL);
849 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, TRUE);
859 if (locale_type == LOCALE_TURKIC && c == 'i')
861 /* i => LATIN CAPITAL LETTER I WITH DOT ABOVE */
862 len += g_unichar_to_utf8 (0x130, out_buffer ? out_buffer + len : NULL);
864 else if (c == 0x0345) /* COMBINING GREEK YPOGEGRAMMENI */
866 /* Nasty, need to move it after other combining marks .. this would go away if
867 * we normalized first.
869 len += output_marks (&p, out_buffer ? out_buffer + len : NULL, FALSE);
871 /* And output as GREEK CAPITAL LETTER IOTA */
872 len += g_unichar_to_utf8 (0x399, out_buffer ? out_buffer + len : NULL);
875 OR (G_UNICODE_LOWERCASE_LETTER,
876 OR (G_UNICODE_TITLECASE_LETTER,
879 val = ATTTABLE (c >> 8, c & 0xff);
881 if (val >= 0x1000000)
883 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t,
884 t == G_UNICODE_LOWERCASE_LETTER ? 0 : 1);
888 if (t == G_UNICODE_TITLECASE_LETTER)
891 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
893 if (title_table[i][0] == c)
895 val = title_table[i][1];
901 /* Some lowercase letters, e.g., U+000AA, FEMININE ORDINAL INDICATOR,
902 * do not have an uppercase equivalent, in which case val will be
904 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
909 gsize char_len = g_utf8_skip[*(guchar *)last];
912 memcpy (out_buffer + len, last, char_len);
924 * @str: a UTF-8 encoded string
925 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
927 * Converts all Unicode characters in the string that have a case
928 * to uppercase. The exact manner that this is done depends
929 * on the current locale, and may result in the number of
930 * characters in the string increasing. (For instance, the
931 * German ess-zet will be changed to SS.)
933 * Return value: a newly allocated string, with all characters
934 * converted to uppercase.
937 g_utf8_strup (const gchar *str,
941 LocaleType locale_type;
944 g_return_val_if_fail (str != NULL, NULL);
946 locale_type = get_locale_type ();
949 * We use a two pass approach to keep memory management simple
951 result_len = real_toupper (str, len, NULL, locale_type);
952 result = g_malloc (result_len + 1);
953 real_toupper (str, len, result, locale_type);
954 result[result_len] = '\0';
959 /* traverses the string checking for characters with combining class == 230
960 * until a base character is found */
962 has_more_above (const gchar *str)
964 const gchar *p = str;
965 gint combining_class;
969 combining_class = _g_unichar_combining_class (g_utf8_get_char (p));
970 if (combining_class == 230)
972 else if (combining_class == 0)
975 p = g_utf8_next_char (p);
982 real_tolower (const gchar *str,
985 LocaleType locale_type)
987 const gchar *p = str;
988 const char *last = NULL;
991 while ((max_len < 0 || p < str + max_len) && *p)
993 gunichar c = g_utf8_get_char (p);
998 p = g_utf8_next_char (p);
1000 if (locale_type == LOCALE_TURKIC && c == 'I')
1002 if (g_utf8_get_char (p) == 0x0307)
1004 /* I + COMBINING DOT ABOVE => i (U+0069) */
1005 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
1006 p = g_utf8_next_char (p);
1010 /* I => LATIN SMALL LETTER DOTLESS I */
1011 len += g_unichar_to_utf8 (0x131, out_buffer ? out_buffer + len : NULL);
1014 /* Introduce an explicit dot above when lowercasing capital I's and J's
1015 * whenever there are more accents above. [SpecialCasing.txt] */
1016 else if (locale_type == LOCALE_LITHUANIAN &&
1017 (c == 0x00cc || c == 0x00cd || c == 0x0128))
1019 len += g_unichar_to_utf8 (0x0069, out_buffer ? out_buffer + len : NULL);
1020 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1025 len += g_unichar_to_utf8 (0x0300, out_buffer ? out_buffer + len : NULL);
1028 len += g_unichar_to_utf8 (0x0301, out_buffer ? out_buffer + len : NULL);
1031 len += g_unichar_to_utf8 (0x0303, out_buffer ? out_buffer + len : NULL);
1035 else if (locale_type == LOCALE_LITHUANIAN &&
1036 (c == 'I' || c == 'J' || c == 0x012e) &&
1039 len += g_unichar_to_utf8 (g_unichar_tolower (c), out_buffer ? out_buffer + len : NULL);
1040 len += g_unichar_to_utf8 (0x0307, out_buffer ? out_buffer + len : NULL);
1042 else if (c == 0x03A3) /* GREEK CAPITAL LETTER SIGMA */
1044 if ((max_len < 0 || p < str + max_len) && *p)
1046 gunichar next_c = g_utf8_get_char (p);
1047 int next_type = TYPE(next_c);
1049 /* SIGMA mapps differently depending on whether it is
1050 * final or not. The following simplified test would
1051 * fail in the case of combining marks following the
1052 * sigma, but I don't think that occurs in real text.
1053 * The test here matches that in ICU.
1055 if (ISALPHA (next_type)) /* Lu,Ll,Lt,Lm,Lo */
1056 val = 0x3c3; /* GREEK SMALL SIGMA */
1058 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1061 val = 0x3c2; /* GREEK SMALL FINAL SIGMA */
1063 len += g_unichar_to_utf8 (val, out_buffer ? out_buffer + len : NULL);
1066 OR (G_UNICODE_UPPERCASE_LETTER,
1067 OR (G_UNICODE_TITLECASE_LETTER,
1070 val = ATTTABLE (c >> 8, c & 0xff);
1072 if (val >= 0x1000000)
1074 len += output_special_case (out_buffer ? out_buffer + len : NULL, val - 0x1000000, t, 0);
1078 if (t == G_UNICODE_TITLECASE_LETTER)
1081 for (i = 0; i < G_N_ELEMENTS (title_table); ++i)
1083 if (title_table[i][0] == c)
1085 val = title_table[i][2];
1091 /* Not all uppercase letters are guaranteed to have a lowercase
1092 * equivalent. If this is the case, val will be zero. */
1093 len += g_unichar_to_utf8 (val ? val : c, out_buffer ? out_buffer + len : NULL);
1098 gsize char_len = g_utf8_skip[*(guchar *)last];
1101 memcpy (out_buffer + len, last, char_len);
1113 * @str: a UTF-8 encoded string
1114 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1116 * Converts all Unicode characters in the string that have a case
1117 * to lowercase. The exact manner that this is done depends
1118 * on the current locale, and may result in the number of
1119 * characters in the string changing.
1121 * Return value: a newly allocated string, with all characters
1122 * converted to lowercase.
1125 g_utf8_strdown (const gchar *str,
1129 LocaleType locale_type;
1132 g_return_val_if_fail (str != NULL, NULL);
1134 locale_type = get_locale_type ();
1137 * We use a two pass approach to keep memory management simple
1139 result_len = real_tolower (str, len, NULL, locale_type);
1140 result = g_malloc (result_len + 1);
1141 real_tolower (str, len, result, locale_type);
1142 result[result_len] = '\0';
1149 * @str: a UTF-8 encoded string
1150 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
1152 * Converts a string into a form that is independent of case. The
1153 * result will not correspond to any particular case, but can be
1154 * compared for equality or ordered with the results of calling
1155 * g_utf8_casefold() on other strings.
1157 * Note that calling g_utf8_casefold() followed by g_utf8_collate() is
1158 * only an approximation to the correct linguistic case insensitive
1159 * ordering, though it is a fairly good one. Getting this exactly
1160 * right would require a more sophisticated collation function that
1161 * takes case sensitivity into account. GLib does not currently
1162 * provide such a function.
1164 * Return value: a newly allocated string, that is a
1165 * case independent form of @str.
1168 g_utf8_casefold (const gchar *str,
1174 g_return_val_if_fail (str != NULL, NULL);
1176 result = g_string_new (NULL);
1178 while ((len < 0 || p < str + len) && *p)
1180 gunichar ch = g_utf8_get_char (p);
1183 int end = G_N_ELEMENTS (casefold_table);
1185 if (ch >= casefold_table[start].ch &&
1186 ch <= casefold_table[end - 1].ch)
1190 int half = (start + end) / 2;
1191 if (ch == casefold_table[half].ch)
1193 g_string_append (result, casefold_table[half].data);
1196 else if (half == start)
1198 else if (ch > casefold_table[half].ch)
1205 g_string_append_unichar (result, g_unichar_tolower (ch));
1208 p = g_utf8_next_char (p);
1211 return g_string_free (result, FALSE);
1215 * g_unichar_get_mirror_char:
1216 * @ch: a Unicode character
1217 * @mirrored_ch: location to store the mirrored character
1219 * In Unicode, some characters are <firstterm>mirrored</firstterm>. This
1220 * means that their images are mirrored horizontally in text that is laid
1221 * out from right to left. For instance, "(" would become its mirror image,
1222 * ")", in right-to-left text.
1224 * If @ch has the Unicode mirrored property and there is another unicode
1225 * character that typically has a glyph that is the mirror image of @ch's
1226 * glyph and @mirrored_ch is set, it puts that character in the address
1227 * pointed to by @mirrored_ch. Otherwise the original character is put.
1229 * Return value: %TRUE if @ch has a mirrored character, %FALSE otherwise
1234 g_unichar_get_mirror_char (gunichar ch,
1235 gunichar *mirrored_ch)
1240 mirrored = GLIB_GET_MIRRORING(ch);
1242 found = ch != mirrored;
1244 *mirrored_ch = mirrored;
1250 #define G_SCRIPT_TABLE_MIDPOINT (G_N_ELEMENTS (g_script_table) / 2)
1252 static inline GUnicodeScript
1253 g_unichar_get_script_bsearch (gunichar ch)
1256 int upper = G_N_ELEMENTS (g_script_table) - 1;
1257 static int saved_mid = G_SCRIPT_TABLE_MIDPOINT;
1258 int mid = saved_mid;
1263 if (ch < g_script_table[mid].start)
1265 else if (ch >= g_script_table[mid].start + g_script_table[mid].chars)
1268 return g_script_table[saved_mid = mid].script;
1270 mid = (lower + upper) / 2;
1272 while (lower <= upper);
1274 return G_UNICODE_SCRIPT_UNKNOWN;
1278 * g_unichar_get_script:
1279 * @ch: a Unicode character
1281 * Looks up the #GUnicodeScript for a particular character (as defined
1282 * by Unicode Standard Annex #24). No check is made for @ch being a
1283 * valid Unicode character; if you pass in invalid character, the
1284 * result is undefined.
1286 * Return value: the #GUnicodeScript for the character.
1291 g_unichar_get_script (gunichar ch)
1293 if (ch < G_EASY_SCRIPTS_RANGE)
1294 return g_script_easy_table[ch];
1296 return g_unichar_get_script_bsearch (ch);
1300 #define __G_UNIPROP_C__
1301 #include "galiasdef.c"