1/*
2 * Copyright (C) 2005 The Android Open Source Project
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 *      http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17#ifndef ANDROID_UNICODE_H
18#define ANDROID_UNICODE_H
19
20#include <sys/types.h>
21#include <stdint.h>
22
23extern "C" {
24
25typedef uint32_t char32_t;
26typedef uint16_t char16_t;
27
28// Standard string functions on char16_t strings.
29int strcmp16(const char16_t *, const char16_t *);
30int strncmp16(const char16_t *s1, const char16_t *s2, size_t n);
31size_t strlen16(const char16_t *);
32size_t strnlen16(const char16_t *, size_t);
33char16_t *strcpy16(char16_t *, const char16_t *);
34char16_t *strncpy16(char16_t *, const char16_t *, size_t);
35
36// Version of comparison that supports embedded nulls.
37// This is different than strncmp() because we don't stop
38// at a nul character and consider the strings to be different
39// if the lengths are different (thus we need to supply the
40// lengths of both strings).  This can also be used when
41// your string is not nul-terminated as it will have the
42// equivalent result as strcmp16 (unlike strncmp16).
43int strzcmp16(const char16_t *s1, size_t n1, const char16_t *s2, size_t n2);
44
45// Version of strzcmp16 for comparing strings in different endianness.
46int strzcmp16_h_n(const char16_t *s1H, size_t n1, const char16_t *s2N, size_t n2);
47
48// Standard string functions on char32_t strings.
49size_t strlen32(const char32_t *);
50size_t strnlen32(const char32_t *, size_t);
51
52/**
53 * Measure the length of a UTF-32 string in UTF-8. If the string is invalid
54 * such as containing a surrogate character, -1 will be returned.
55 */
56ssize_t utf32_to_utf8_length(const char32_t *src, size_t src_len);
57
58/**
59 * Stores a UTF-8 string converted from "src" in "dst", if "dst_length" is not
60 * large enough to store the string, the part of the "src" string is stored
61 * into "dst" as much as possible. See the examples for more detail.
62 * Returns the size actually used for storing the string.
63 * dst" is not null-terminated when dst_len is fully used (like strncpy).
64 *
65 * Example 1
66 * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
67 * "src_len" == 2
68 * "dst_len" >= 7
69 * ->
70 * Returned value == 6
71 * "dst" becomes \xE3\x81\x82\xE3\x81\x84\0
72 * (note that "dst" is null-terminated)
73 *
74 * Example 2
75 * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
76 * "src_len" == 2
77 * "dst_len" == 5
78 * ->
79 * Returned value == 3
80 * "dst" becomes \xE3\x81\x82\0
81 * (note that "dst" is null-terminated, but \u3044 is not stored in "dst"
82 * since "dst" does not have enough size to store the character)
83 *
84 * Example 3
85 * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
86 * "src_len" == 2
87 * "dst_len" == 6
88 * ->
89 * Returned value == 6
90 * "dst" becomes \xE3\x81\x82\xE3\x81\x84
91 * (note that "dst" is NOT null-terminated, like strncpy)
92 */
93void utf32_to_utf8(const char32_t* src, size_t src_len, char* dst);
94
95/**
96 * Returns the unicode value at "index".
97 * Returns -1 when the index is invalid (equals to or more than "src_len").
98 * If returned value is positive, it is able to be converted to char32_t, which
99 * is unsigned. Then, if "next_index" is not NULL, the next index to be used is
100 * stored in "next_index". "next_index" can be NULL.
101 */
102int32_t utf32_from_utf8_at(const char *src, size_t src_len, size_t index, size_t *next_index);
103
104
105/**
106 * Returns the UTF-8 length of UTF-16 string "src".
107 */
108ssize_t utf16_to_utf8_length(const char16_t *src, size_t src_len);
109
110/**
111 * Converts a UTF-16 string to UTF-8. The destination buffer must be large
112 * enough to fit the UTF-16 as measured by utf16_to_utf8_length with an added
113 * NULL terminator.
114 */
115void utf16_to_utf8(const char16_t* src, size_t src_len, char* dst);
116
117/**
118 * Returns the length of "src" when "src" is valid UTF-8 string.
119 * Returns 0 if src is NULL or 0-length string. Returns -1 when the source
120 * is an invalid string.
121 *
122 * This function should be used to determine whether "src" is valid UTF-8
123 * characters with valid unicode codepoints. "src" must be null-terminated.
124 *
125 * If you are going to use other utf8_to_... functions defined in this header
126 * with string which may not be valid UTF-8 with valid codepoint (form 0 to
127 * 0x10FFFF), you should use this function before calling others, since the
128 * other functions do not check whether the string is valid UTF-8 or not.
129 *
130 * If you do not care whether "src" is valid UTF-8 or not, you should use
131 * strlen() as usual, which should be much faster.
132 */
133ssize_t utf8_length(const char *src);
134
135/**
136 * Measure the length of a UTF-32 string.
137 */
138size_t utf8_to_utf32_length(const char *src, size_t src_len);
139
140/**
141 * Stores a UTF-32 string converted from "src" in "dst". "dst" must be large
142 * enough to store the entire converted string as measured by
143 * utf8_to_utf32_length plus space for a NULL terminator.
144 */
145void utf8_to_utf32(const char* src, size_t src_len, char32_t* dst);
146
147/**
148 * Returns the UTF-16 length of UTF-8 string "src".
149 */
150ssize_t utf8_to_utf16_length(const uint8_t* src, size_t srcLen);
151
152/**
153 * Convert UTF-8 to UTF-16 including surrogate pairs.
154 * Returns a pointer to the end of the string (where a null terminator might go
155 * if you wanted to add one).
156 */
157char16_t* utf8_to_utf16_no_null_terminator(const uint8_t* src, size_t srcLen, char16_t* dst);
158
159/**
160 * Convert UTF-8 to UTF-16 including surrogate pairs. The destination buffer
161 * must be large enough to hold the result as measured by utf8_to_utf16_length
162 * plus an added NULL terminator.
163 */
164void utf8_to_utf16(const uint8_t* src, size_t srcLen, char16_t* dst);
165
166/**
167 * Like utf8_to_utf16_no_null_terminator, but you can supply a maximum length of the
168 * decoded string.  The decoded string will fill up to that length; if it is longer
169 * the returned pointer will be to the character after dstLen.
170 */
171char16_t* utf8_to_utf16_n(const uint8_t* src, size_t srcLen, char16_t* dst, size_t dstLen);
172
173}
174
175#endif
176