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
22 #include <stddef.h>
23 #include <stdint.h>
24
25 /*
26 * All UTF-8 in art is actually modified UTF-8. Mostly, this distinction
27 * doesn't matter.
28 *
29 * See http://en.wikipedia.org/wiki/UTF-8#Modified_UTF-8 for the details.
30 */
31 namespace art {
32
33 /*
34 * Returns the number of UTF-16 characters in the given modified UTF-8 string.
35 */
36 size_t CountModifiedUtf8Chars(const char* utf8);
37 size_t CountModifiedUtf8Chars(const char* utf8, size_t byte_count);
38
39 /*
40 * Returns the number of modified UTF-8 bytes needed to represent the given
41 * UTF-16 string.
42 */
43 size_t CountUtf8Bytes(const uint16_t* chars, size_t char_count);
44
45 /*
46 * Convert from Modified UTF-8 to UTF-16.
47 */
48 void ConvertModifiedUtf8ToUtf16(uint16_t* utf16_out, const char* utf8_in);
49 void ConvertModifiedUtf8ToUtf16(uint16_t* utf16_out, size_t out_chars,
50 const char* utf8_in, size_t in_bytes);
51
52 /*
53 * Compare two modified UTF-8 strings as UTF-16 code point values in a non-locale sensitive manner
54 */
55 ALWAYS_INLINE int CompareModifiedUtf8ToModifiedUtf8AsUtf16CodePointValues(const char* utf8_1,
56 const char* utf8_2);
57
58 /*
59 * Compare a null-terminated modified UTF-8 string with a UTF-16 string (not null-terminated)
60 * as code point values in a non-locale sensitive manner.
61 */
62 int CompareModifiedUtf8ToUtf16AsCodePointValues(const char* utf8, const uint16_t* utf16,
63 size_t utf16_length);
64
65 /*
66 * Convert from UTF-16 to Modified UTF-8. Note that the output is _not_
67 * NUL-terminated. You probably need to call CountUtf8Bytes before calling
68 * this anyway, so if you want a NUL-terminated string, you know where to
69 * put the NUL byte.
70 */
71 void ConvertUtf16ToModifiedUtf8(char* utf8_out, size_t byte_count,
72 const uint16_t* utf16_in, size_t char_count);
73
74 /*
75 * The java.lang.String hashCode() algorithm.
76 */
77 template<typename MemoryType>
ComputeUtf16Hash(const MemoryType * chars,size_t char_count)78 int32_t ComputeUtf16Hash(const MemoryType* chars, size_t char_count) {
79 uint32_t hash = 0;
80 while (char_count--) {
81 hash = hash * 31 + *chars++;
82 }
83 return static_cast<int32_t>(hash);
84 }
85
86 int32_t ComputeUtf16HashFromModifiedUtf8(const char* utf8, size_t utf16_length);
87
88 // Compute a hash code of a modified UTF-8 string. Not the standard java hash since it returns a
89 // uint32_t and hashes individual chars instead of codepoint words.
90 uint32_t ComputeModifiedUtf8Hash(const char* chars);
91
92 /*
93 * Retrieve the next UTF-16 character or surrogate pair from a UTF-8 string.
94 * single byte, 2-byte and 3-byte UTF-8 sequences result in a single UTF-16
95 * character (possibly one half of a surrogate) whereas 4-byte UTF-8 sequences
96 * result in a surrogate pair. Use GetLeadingUtf16Char and GetTrailingUtf16Char
97 * to process the return value of this function.
98 *
99 * Advances "*utf8_data_in" to the start of the next character.
100 *
101 * WARNING: If a string is corrupted by dropping a '\0' in the middle
102 * of a multi byte sequence, you can end up overrunning the buffer with
103 * reads (and possibly with the writes if the length was computed and
104 * cached before the damage). For performance reasons, this function
105 * assumes that the string being parsed is known to be valid (e.g., by
106 * already being verified). Most strings we process here are coming
107 * out of dex files or other internal translations, so the only real
108 * risk comes from the JNI NewStringUTF call.
109 */
110 uint32_t GetUtf16FromUtf8(const char** utf8_data_in);
111
112 /**
113 * Gets the leading UTF-16 character from a surrogate pair, or the sole
114 * UTF-16 character from the return value of GetUtf16FromUtf8.
115 */
116 ALWAYS_INLINE uint16_t GetLeadingUtf16Char(uint32_t maybe_pair);
117
118 /**
119 * Gets the trailing UTF-16 character from a surrogate pair, or 0 otherwise
120 * from the return value of GetUtf16FromUtf8.
121 */
122 ALWAYS_INLINE uint16_t GetTrailingUtf16Char(uint32_t maybe_pair);
123
124 } // namespace art
125
126 #endif // ART_RUNTIME_UTF_H_
127