1 /*
2  * Copyright (C) 2011 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 ART_RUNTIME_UTF_H_
18 #define ART_RUNTIME_UTF_H_
19 
20 #include "base/macros.h"
21 #include "base/mutex.h"
22 
23 #include <stddef.h>
24 #include <stdint.h>
25 
26 /*
27  * All UTF-8 in art is actually modified UTF-8. Mostly, this distinction
28  * doesn't matter.
29  *
30  * See http://en.wikipedia.org/wiki/UTF-8#Modified_UTF-8 for the details.
31  */
32 namespace art {
33 
34 namespace mirror {
35   template<class T> class PrimitiveArray;
36   typedef PrimitiveArray<uint16_t> CharArray;
37 }  // namespace mirror
38 
39 /*
40  * Returns the number of UTF-16 characters in the given modified UTF-8 string.
41  */
42 size_t CountModifiedUtf8Chars(const char* utf8);
43 
44 /*
45  * Returns the number of modified UTF-8 bytes needed to represent the given
46  * UTF-16 string.
47  */
48 size_t CountUtf8Bytes(const uint16_t* chars, size_t char_count);
49 
50 /*
51  * Convert from Modified UTF-8 to UTF-16.
52  */
53 void ConvertModifiedUtf8ToUtf16(uint16_t* utf16_out, const char* utf8_in);
54 
55 /*
56  * Compare two modified UTF-8 strings as UTF-16 code point values in a non-locale sensitive manner
57  */
58 ALWAYS_INLINE int CompareModifiedUtf8ToModifiedUtf8AsUtf16CodePointValues(const char* utf8_1,
59                                                                           const char* utf8_2);
60 
61 /*
62  * Compare a modified UTF-8 string with a UTF-16 string as code point values in a non-locale
63  * sensitive manner.
64  */
65 int CompareModifiedUtf8ToUtf16AsCodePointValues(const char* utf8_1, const uint16_t* utf8_2);
66 
67 /*
68  * Convert from UTF-16 to Modified UTF-8. Note that the output is _not_
69  * NUL-terminated. You probably need to call CountUtf8Bytes before calling
70  * this anyway, so if you want a NUL-terminated string, you know where to
71  * put the NUL byte.
72  */
73 void ConvertUtf16ToModifiedUtf8(char* utf8_out, const uint16_t* utf16_in, size_t char_count);
74 
75 /*
76  * The java.lang.String hashCode() algorithm.
77  */
78 int32_t ComputeUtf16Hash(mirror::CharArray* chars, int32_t offset, size_t char_count)
79     SHARED_LOCKS_REQUIRED(Locks::mutator_lock_);
80 int32_t ComputeUtf16Hash(const uint16_t* chars, size_t char_count);
81 
82 // Compute a hash code of a modified UTF-8 string. Not the standard java hash since it returns a
83 // size_t and hashes individual chars instead of codepoint words.
84 size_t ComputeModifiedUtf8Hash(const char* chars);
85 
86 /*
87  * Retrieve the next UTF-16 character from a UTF-8 string.
88  *
89  * Advances "*utf8_data_in" to the start of the next character.
90  *
91  * WARNING: If a string is corrupted by dropping a '\0' in the middle
92  * of a 3-byte sequence, you can end up overrunning the buffer with
93  * reads (and possibly with the writes if the length was computed and
94  * cached before the damage). For performance reasons, this function
95  * assumes that the string being parsed is known to be valid (e.g., by
96  * already being verified). Most strings we process here are coming
97  * out of dex files or other internal translations, so the only real
98  * risk comes from the JNI NewStringUTF call.
99  */
100 uint16_t GetUtf16FromUtf8(const char** utf8_data_in);
101 
102 }  // namespace art
103 
104 #endif  // ART_RUNTIME_UTF_H_
105