• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
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 
23 extern "C" {
24 
25 // Definitions exist in C++11
26 #if defined __cplusplus && __cplusplus < 201103L
27 typedef uint32_t char32_t;
28 typedef uint16_t char16_t;
29 #endif
30 
31 // Standard string functions on char16_t strings.
32 int strcmp16(const char16_t *, const char16_t *);
33 int strncmp16(const char16_t *s1, const char16_t *s2, size_t n);
34 size_t strlen16(const char16_t *);
35 size_t strnlen16(const char16_t *, size_t);
36 char16_t *strcpy16(char16_t *, const char16_t *);
37 char16_t *strncpy16(char16_t *, const char16_t *, size_t);
38 
39 // Version of comparison that supports embedded nulls.
40 // This is different than strncmp() because we don't stop
41 // at a nul character and consider the strings to be different
42 // if the lengths are different (thus we need to supply the
43 // lengths of both strings).  This can also be used when
44 // your string is not nul-terminated as it will have the
45 // equivalent result as strcmp16 (unlike strncmp16).
46 int strzcmp16(const char16_t *s1, size_t n1, const char16_t *s2, size_t n2);
47 
48 // Version of strzcmp16 for comparing strings in different endianness.
49 int strzcmp16_h_n(const char16_t *s1H, size_t n1, const char16_t *s2N, size_t n2);
50 
51 // Standard string functions on char32_t strings.
52 size_t strlen32(const char32_t *);
53 size_t strnlen32(const char32_t *, size_t);
54 
55 /**
56  * Measure the length of a UTF-32 string in UTF-8. If the string is invalid
57  * such as containing a surrogate character, -1 will be returned.
58  */
59 ssize_t utf32_to_utf8_length(const char32_t *src, size_t src_len);
60 
61 /**
62  * Stores a UTF-8 string converted from "src" in "dst", if "dst_length" is not
63  * large enough to store the string, the part of the "src" string is stored
64  * into "dst" as much as possible. See the examples for more detail.
65  * Returns the size actually used for storing the string.
66  * dst" is not null-terminated when dst_len is fully used (like strncpy).
67  *
68  * Example 1
69  * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
70  * "src_len" == 2
71  * "dst_len" >= 7
72  * ->
73  * Returned value == 6
74  * "dst" becomes \xE3\x81\x82\xE3\x81\x84\0
75  * (note that "dst" is null-terminated)
76  *
77  * Example 2
78  * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
79  * "src_len" == 2
80  * "dst_len" == 5
81  * ->
82  * Returned value == 3
83  * "dst" becomes \xE3\x81\x82\0
84  * (note that "dst" is null-terminated, but \u3044 is not stored in "dst"
85  * since "dst" does not have enough size to store the character)
86  *
87  * Example 3
88  * "src" == \u3042\u3044 (\xE3\x81\x82\xE3\x81\x84)
89  * "src_len" == 2
90  * "dst_len" == 6
91  * ->
92  * Returned value == 6
93  * "dst" becomes \xE3\x81\x82\xE3\x81\x84
94  * (note that "dst" is NOT null-terminated, like strncpy)
95  */
96 void utf32_to_utf8(const char32_t* src, size_t src_len, char* dst);
97 
98 /**
99  * Returns the unicode value at "index".
100  * Returns -1 when the index is invalid (equals to or more than "src_len").
101  * If returned value is positive, it is able to be converted to char32_t, which
102  * is unsigned. Then, if "next_index" is not NULL, the next index to be used is
103  * stored in "next_index". "next_index" can be NULL.
104  */
105 int32_t utf32_from_utf8_at(const char *src, size_t src_len, size_t index, size_t *next_index);
106 
107 
108 /**
109  * Returns the UTF-8 length of UTF-16 string "src".
110  */
111 ssize_t utf16_to_utf8_length(const char16_t *src, size_t src_len);
112 
113 /**
114  * Converts a UTF-16 string to UTF-8. The destination buffer must be large
115  * enough to fit the UTF-16 as measured by utf16_to_utf8_length with an added
116  * NULL terminator.
117  */
118 void utf16_to_utf8(const char16_t* src, size_t src_len, char* dst);
119 
120 /**
121  * Returns the length of "src" when "src" is valid UTF-8 string.
122  * Returns 0 if src is NULL or 0-length string. Returns -1 when the source
123  * is an invalid string.
124  *
125  * This function should be used to determine whether "src" is valid UTF-8
126  * characters with valid unicode codepoints. "src" must be null-terminated.
127  *
128  * If you are going to use other utf8_to_... functions defined in this header
129  * with string which may not be valid UTF-8 with valid codepoint (form 0 to
130  * 0x10FFFF), you should use this function before calling others, since the
131  * other functions do not check whether the string is valid UTF-8 or not.
132  *
133  * If you do not care whether "src" is valid UTF-8 or not, you should use
134  * strlen() as usual, which should be much faster.
135  */
136 ssize_t utf8_length(const char *src);
137 
138 /**
139  * Measure the length of a UTF-32 string.
140  */
141 size_t utf8_to_utf32_length(const char *src, size_t src_len);
142 
143 /**
144  * Stores a UTF-32 string converted from "src" in "dst". "dst" must be large
145  * enough to store the entire converted string as measured by
146  * utf8_to_utf32_length plus space for a NULL terminator.
147  */
148 void utf8_to_utf32(const char* src, size_t src_len, char32_t* dst);
149 
150 /**
151  * Returns the UTF-16 length of UTF-8 string "src".
152  */
153 ssize_t utf8_to_utf16_length(const uint8_t* src, size_t srcLen);
154 
155 /**
156  * Convert UTF-8 to UTF-16 including surrogate pairs.
157  * Returns a pointer to the end of the string (where a null terminator might go
158  * if you wanted to add one).
159  */
160 char16_t* utf8_to_utf16_no_null_terminator(const uint8_t* src, size_t srcLen, char16_t* dst);
161 
162 /**
163  * Convert UTF-8 to UTF-16 including surrogate pairs. The destination buffer
164  * must be large enough to hold the result as measured by utf8_to_utf16_length
165  * plus an added NULL terminator.
166  */
167 void utf8_to_utf16(const uint8_t* src, size_t srcLen, char16_t* dst);
168 
169 /**
170  * Like utf8_to_utf16_no_null_terminator, but you can supply a maximum length of the
171  * decoded string.  The decoded string will fill up to that length; if it is longer
172  * the returned pointer will be to the character after dstLen.
173  */
174 char16_t* utf8_to_utf16_n(const uint8_t* src, size_t srcLen, char16_t* dst, size_t dstLen);
175 
176 }
177 
178 #endif
179