blob: 39c8d153d55d1413ead5fefd17f4c5f1a0c3e426 [file] [log] [blame]
Elliott Hughes2faa5f12012-01-30 14:42:07 -08001/*
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 */
Elliott Hughes814e4032011-08-23 12:07:56 -070016
17#include "utf.h"
18
Elliott Hughes07ed66b2012-12-12 18:34:25 -080019#include "base/logging.h"
Ian Rogers2dd0e2c2013-01-24 12:42:14 -080020#include "mirror/array.h"
Ian Rogers4f6ad8a2013-03-18 15:27:28 -070021#include "mirror/object-inl.h"
Ian Rogersa6724902013-09-23 09:23:37 -070022#include "utf-inl.h"
Elliott Hughesb465ab02011-08-24 11:21:21 -070023
Elliott Hughes814e4032011-08-23 12:07:56 -070024namespace art {
25
26size_t CountModifiedUtf8Chars(const char* utf8) {
27 size_t len = 0;
28 int ic;
29 while ((ic = *utf8++) != '\0') {
30 len++;
31 if ((ic & 0x80) == 0) {
32 // one-byte encoding
33 continue;
34 }
35 // two- or three-byte encoding
36 utf8++;
37 if ((ic & 0x20) == 0) {
38 // two-byte encoding
39 continue;
40 }
Elliott Hughes814e4032011-08-23 12:07:56 -070041 utf8++;
Narayan Kamatha5afcfc2015-01-29 20:06:46 +000042 if ((ic & 0x10) == 0) {
43 // three-byte encoding
44 continue;
45 }
46
47 // four-byte encoding: needs to be converted into a surrogate
48 // pair.
49 utf8++;
50 len++;
Elliott Hughes814e4032011-08-23 12:07:56 -070051 }
52 return len;
53}
54
55void ConvertModifiedUtf8ToUtf16(uint16_t* utf16_data_out, const char* utf8_data_in) {
56 while (*utf8_data_in != '\0') {
Narayan Kamatha5afcfc2015-01-29 20:06:46 +000057 const uint32_t ch = GetUtf16FromUtf8(&utf8_data_in);
58 const uint16_t leading = GetLeadingUtf16Char(ch);
59 const uint16_t trailing = GetTrailingUtf16Char(ch);
60
61 *utf16_data_out++ = leading;
62 if (trailing != 0) {
63 *utf16_data_out++ = trailing;
64 }
Elliott Hughes814e4032011-08-23 12:07:56 -070065 }
66}
67
Elliott Hughesb465ab02011-08-24 11:21:21 -070068void ConvertUtf16ToModifiedUtf8(char* utf8_out, const uint16_t* utf16_in, size_t char_count) {
69 while (char_count--) {
70 uint16_t ch = *utf16_in++;
71 if (ch > 0 && ch <= 0x7f) {
72 *utf8_out++ = ch;
73 } else {
74 if (ch > 0x07ff) {
75 *utf8_out++ = (ch >> 12) | 0xe0;
76 *utf8_out++ = ((ch >> 6) & 0x3f) | 0x80;
77 *utf8_out++ = (ch & 0x3f) | 0x80;
78 } else /*(ch > 0x7f || ch == 0)*/ {
79 *utf8_out++ = (ch >> 6) | 0xc0;
80 *utf8_out++ = (ch & 0x3f) | 0x80;
81 }
82 }
83 }
84}
85
Ian Rogersef7d42f2014-01-06 12:55:46 -080086int32_t ComputeUtf16Hash(mirror::CharArray* chars, int32_t offset,
Ian Rogers0cfe1fb2011-08-26 03:29:44 -070087 size_t char_count) {
Ian Rogers8f41dc32014-10-30 15:16:16 -070088 uint32_t hash = 0;
Ian Rogers0cfe1fb2011-08-26 03:29:44 -070089 for (size_t i = 0; i < char_count; i++) {
90 hash = hash * 31 + chars->Get(offset + i);
91 }
Ian Rogers8f41dc32014-10-30 15:16:16 -070092 return static_cast<int32_t>(hash);
Ian Rogers0cfe1fb2011-08-26 03:29:44 -070093}
94
Elliott Hughes814e4032011-08-23 12:07:56 -070095int32_t ComputeUtf16Hash(const uint16_t* chars, size_t char_count) {
Ian Rogers8f41dc32014-10-30 15:16:16 -070096 uint32_t hash = 0;
Elliott Hughes814e4032011-08-23 12:07:56 -070097 while (char_count--) {
98 hash = hash * 31 + *chars++;
99 }
Ian Rogers8f41dc32014-10-30 15:16:16 -0700100 return static_cast<int32_t>(hash);
Elliott Hughes814e4032011-08-23 12:07:56 -0700101}
102
Mathieu Chartiere7c9a8c2014-11-06 16:35:45 -0800103size_t ComputeModifiedUtf8Hash(const char* chars) {
104 size_t hash = 0;
Ian Rogers68b56852014-08-29 20:19:11 -0700105 while (*chars != '\0') {
Mathieu Chartiere7c9a8c2014-11-06 16:35:45 -0800106 hash = hash * 31 + *chars++;
Ian Rogers68b56852014-08-29 20:19:11 -0700107 }
Ian Rogers8f41dc32014-10-30 15:16:16 -0700108 return static_cast<int32_t>(hash);
Ian Rogers68b56852014-08-29 20:19:11 -0700109}
110
Vladimir Markoa48aef42014-12-03 17:53:53 +0000111int CompareModifiedUtf8ToUtf16AsCodePointValues(const char* utf8, const uint16_t* utf16,
112 size_t utf16_length) {
Ian Rogers637c65b2013-05-31 11:46:00 -0700113 for (;;) {
Vladimir Markoa48aef42014-12-03 17:53:53 +0000114 if (*utf8 == '\0') {
115 return (utf16_length == 0) ? 0 : -1;
116 } else if (utf16_length == 0) {
Ian Rogers637c65b2013-05-31 11:46:00 -0700117 return 1;
118 }
119
Narayan Kamatha5afcfc2015-01-29 20:06:46 +0000120 const uint32_t pair = GetUtf16FromUtf8(&utf8);
Ian Rogers637c65b2013-05-31 11:46:00 -0700121
Narayan Kamatha5afcfc2015-01-29 20:06:46 +0000122 // First compare the leading utf16 char.
123 const uint16_t lhs = GetLeadingUtf16Char(pair);
124 const uint16_t rhs = *utf16++;
125 --utf16_length;
126 if (lhs != rhs) {
127 return lhs > rhs ? 1 : -1;
128 }
129
130 // Then compare the trailing utf16 char. First check if there
131 // are any characters left to consume.
132 const uint16_t lhs2 = GetTrailingUtf16Char(pair);
133 if (lhs2 != 0) {
134 if (utf16_length == 0) {
135 return 1;
136 }
137
138 const uint16_t rhs2 = *utf16++;
139 --utf16_length;
140 if (lhs2 != rhs2) {
141 return lhs2 > rhs2 ? 1 : -1;
142 }
Ian Rogers637c65b2013-05-31 11:46:00 -0700143 }
144 }
145}
146
Elliott Hughes814e4032011-08-23 12:07:56 -0700147size_t CountUtf8Bytes(const uint16_t* chars, size_t char_count) {
148 size_t result = 0;
149 while (char_count--) {
150 uint16_t ch = *chars++;
151 if (ch > 0 && ch <= 0x7f) {
152 ++result;
153 } else {
154 if (ch > 0x7ff) {
155 result += 3;
156 } else {
157 result += 2;
158 }
159 }
160 }
161 return result;
162}
163
164} // namespace art