1 /* Normalization forms (composition and decomposition) of Unicode strings.
2 Copyright (C) 2001-2002, 2009-2010 Free Software Foundation, Inc.
3 Written by Bruno Haible <bruno@clisp.org>, 2009.
5 This program is free software: you can redistribute it and/or modify it
6 under the terms of the GNU Lesser General Public License as published
7 by the Free Software Foundation; either version 3 of the License, or
8 (at your option) any later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 Lesser General Public License for more details.
15 You should have received a copy of the GNU Lesser General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>. */
34 All functions prefixed with u8_ operate on UTF-8 encoded strings.
35 Their unit is an uint8_t (1 byte).
37 All functions prefixed with u16_ operate on UTF-16 encoded strings.
38 Their unit is an uint16_t (a 2-byte word).
40 All functions prefixed with u32_ operate on UCS-4 encoded strings.
41 Their unit is an uint32_t (a 4-byte word).
43 All argument pairs (s, n) denote a Unicode string s[0..n-1] with exactly
46 Functions returning a string result take a (resultbuf, lengthp) argument
47 pair. If resultbuf is not NULL and the result fits into *lengthp units,
48 it is put in resultbuf, and resultbuf is returned. Otherwise, a freshly
49 allocated string is returned. In both cases, *lengthp is set to the
50 length (number of units) of the returned string. In case of error,
51 NULL is returned and errno is set. */
56 UC_DECOMP_CANONICAL,/* Canonical decomposition. */
57 UC_DECOMP_FONT, /* <font> A font variant (e.g. a blackletter form). */
58 UC_DECOMP_NOBREAK, /* <noBreak> A no-break version of a space or hyphen. */
59 UC_DECOMP_INITIAL, /* <initial> An initial presentation form (Arabic). */
60 UC_DECOMP_MEDIAL, /* <medial> A medial presentation form (Arabic). */
61 UC_DECOMP_FINAL, /* <final> A final presentation form (Arabic). */
62 UC_DECOMP_ISOLATED,/* <isolated> An isolated presentation form (Arabic). */
63 UC_DECOMP_CIRCLE, /* <circle> An encircled form. */
64 UC_DECOMP_SUPER, /* <super> A superscript form. */
65 UC_DECOMP_SUB, /* <sub> A subscript form. */
66 UC_DECOMP_VERTICAL,/* <vertical> A vertical layout presentation form. */
67 UC_DECOMP_WIDE, /* <wide> A wide (or zenkaku) compatibility character. */
68 UC_DECOMP_NARROW, /* <narrow> A narrow (or hankaku) compatibility character. */
69 UC_DECOMP_SMALL, /* <small> A small variant form (CNS compatibility). */
70 UC_DECOMP_SQUARE, /* <square> A CJK squared font variant. */
71 UC_DECOMP_FRACTION,/* <fraction> A vulgar fraction form. */
72 UC_DECOMP_COMPAT /* <compat> Otherwise unspecified compatibility character. */
75 /* Maximum size of decomposition of a single Unicode character. */
76 #define UC_DECOMPOSITION_MAX_LENGTH 32
78 /* Return the character decomposition mapping of a Unicode character.
79 DECOMPOSITION must point to an array of at least UC_DECOMPOSITION_MAX_LENGTH
81 When a decomposition exists, DECOMPOSITION[0..N-1] and *DECOMP_TAG are
82 filled and N is returned. Otherwise -1 is returned. */
84 uc_decomposition (ucs4_t uc, int *decomp_tag, ucs4_t *decomposition);
86 /* Return the canonical character decomposition mapping of a Unicode character.
87 DECOMPOSITION must point to an array of at least UC_DECOMPOSITION_MAX_LENGTH
89 When a decomposition exists, DECOMPOSITION[0..N-1] is filled and N is
90 returned. Otherwise -1 is returned. */
92 uc_canonical_decomposition (ucs4_t uc, ucs4_t *decomposition);
95 /* Attempt to combine the Unicode characters uc1, uc2.
96 uc1 is known to have canonical combining class 0.
97 Return the combination of uc1 and uc2, if it exists.
99 Not all decompositions can be recombined using this function. See the
100 Unicode file CompositionExclusions.txt for details. */
102 uc_composition (ucs4_t uc1, ucs4_t uc2);
105 /* An object of type uninorm_t denotes a Unicode normalization form. */
106 struct unicode_normalization_form;
107 typedef const struct unicode_normalization_form *uninorm_t;
109 /* UNINORM_NFD: Normalization form D: canonical decomposition. */
110 extern const struct unicode_normalization_form uninorm_nfd;
111 #define UNINORM_NFD (&uninorm_nfd)
113 /* UNINORM_NFC: Normalization form C: canonical decomposition, then
114 canonical composition. */
115 extern const struct unicode_normalization_form uninorm_nfc;
116 #define UNINORM_NFC (&uninorm_nfc)
118 /* UNINORM_NFKD: Normalization form KD: compatibility decomposition. */
119 extern const struct unicode_normalization_form uninorm_nfkd;
120 #define UNINORM_NFKD (&uninorm_nfkd)
122 /* UNINORM_NFKC: Normalization form KC: compatibility decomposition, then
123 canonical composition. */
124 extern const struct unicode_normalization_form uninorm_nfkc;
125 #define UNINORM_NFKC (&uninorm_nfkc)
127 /* Test whether a normalization form does compatibility decomposition. */
128 #define uninorm_is_compat_decomposing(nf) \
129 ((* (const unsigned int *) (nf) >> 0) & 1)
131 /* Test whether a normalization form includes canonical composition. */
132 #define uninorm_is_composing(nf) \
133 ((* (const unsigned int *) (nf) >> 1) & 1)
135 /* Return the decomposing variant of a normalization form.
136 This maps NFC,NFD -> NFD and NFKC,NFKD -> NFKD. */
137 extern uninorm_t uninorm_decomposing_form (uninorm_t nf);
140 /* Return the specified normalization form of a string. */
142 u8_normalize (uninorm_t nf, const uint8_t *s, size_t n,
143 uint8_t *resultbuf, size_t *lengthp);
145 u16_normalize (uninorm_t nf, const uint16_t *s, size_t n,
146 uint16_t *resultbuf, size_t *lengthp);
148 u32_normalize (uninorm_t nf, const uint32_t *s, size_t n,
149 uint32_t *resultbuf, size_t *lengthp);
152 /* Compare S1 and S2, ignoring differences in normalization.
153 NF must be either UNINORM_NFD or UNINORM_NFKD.
154 If successful, set *RESULTP to -1 if S1 < S2, 0 if S1 = S2, 1 if S1 > S2, and
155 return 0. Upon failure, return -1 with errno set. */
157 u8_normcmp (const uint8_t *s1, size_t n1, const uint8_t *s2, size_t n2,
158 uninorm_t nf, int *resultp);
160 u16_normcmp (const uint16_t *s1, size_t n1, const uint16_t *s2, size_t n2,
161 uninorm_t nf, int *resultp);
163 u32_normcmp (const uint32_t *s1, size_t n1, const uint32_t *s2, size_t n2,
164 uninorm_t nf, int *resultp);
167 /* Converts the string S of length N to a NUL-terminated byte sequence, in such
168 a way that comparing uN_normxfrm (S1) and uN_normxfrm (S2) with uN_cmp2() is
169 equivalent to comparing S1 and S2 with uN_normcoll().
170 NF must be either UNINORM_NFC or UNINORM_NFKC. */
172 u8_normxfrm (const uint8_t *s, size_t n, uninorm_t nf,
173 char *resultbuf, size_t *lengthp);
175 u16_normxfrm (const uint16_t *s, size_t n, uninorm_t nf,
176 char *resultbuf, size_t *lengthp);
178 u32_normxfrm (const uint32_t *s, size_t n, uninorm_t nf,
179 char *resultbuf, size_t *lengthp);
182 /* Compare S1 and S2, ignoring differences in normalization, using the
183 collation rules of the current locale.
184 NF must be either UNINORM_NFC or UNINORM_NFKC.
185 If successful, set *RESULTP to -1 if S1 < S2, 0 if S1 = S2, 1 if S1 > S2, and
186 return 0. Upon failure, return -1 with errno set. */
188 u8_normcoll (const uint8_t *s1, size_t n1, const uint8_t *s2, size_t n2,
189 uninorm_t nf, int *resultp);
191 u16_normcoll (const uint16_t *s1, size_t n1, const uint16_t *s2, size_t n2,
192 uninorm_t nf, int *resultp);
194 u32_normcoll (const uint32_t *s1, size_t n1, const uint32_t *s2, size_t n2,
195 uninorm_t nf, int *resultp);
198 /* Normalization of a stream of Unicode characters.
200 A "stream of Unicode characters" is essentially a function that accepts an
201 ucs4_t argument repeatedly, optionally combined with a function that
202 "flushes" the stream. */
204 /* Data type of a stream of Unicode characters that normalizes its input
205 according to a given normalization form and passes the normalized character
206 sequence to the encapsulated stream of Unicode characters. */
207 struct uninorm_filter;
209 /* Create and return a normalization filter for Unicode characters.
210 The pair (stream_func, stream_data) is the encapsulated stream.
211 stream_func (stream_data, uc) receives the Unicode character uc
212 and returns 0 if successful, or -1 with errno set upon failure.
213 Return the new filter, or NULL with errno set upon failure. */
214 extern struct uninorm_filter *
215 uninorm_filter_create (uninorm_t nf,
216 int (*stream_func) (void *stream_data, ucs4_t uc),
219 /* Stuff a Unicode character into a normalizing filter.
220 Return 0 if successful, or -1 with errno set upon failure. */
222 uninorm_filter_write (struct uninorm_filter *filter, ucs4_t uc);
224 /* Bring data buffered in the filter to its destination, the encapsulated
226 Return 0 if successful, or -1 with errno set upon failure.
227 Note! If after calling this function, additional characters are written
228 into the filter, the resulting character sequence in the encapsulated stream
229 will not necessarily be normalized. */
231 uninorm_filter_flush (struct uninorm_filter *filter);
233 /* Bring data buffered in the filter to its destination, the encapsulated
234 stream, then close and free the filter.
235 Return 0 if successful, or -1 with errno set upon failure. */
237 uninorm_filter_free (struct uninorm_filter *filter);
245 #endif /* _UNINORM_H */