utf-inl.h revision cb6b0f31ede2275e79e6199ec391147585a37a2a 
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_INL_H_
18#define ART_RUNTIME_UTF_INL_H_
19
20#include "utf.h"
21
22namespace art {
23
24inline uint16_t GetUtf16FromUtf8(const char** utf8_data_in) {
25  uint8_t one = *(*utf8_data_in)++;
26  if ((one & 0x80) == 0) {
27    // one-byte encoding
28    return one;
29  }
30  // two- or three-byte encoding
31  uint8_t two = *(*utf8_data_in)++;
32  if ((one & 0x20) == 0) {
33    // two-byte encoding
34    return ((one & 0x1f) << 6) | (two & 0x3f);
35  }
36  // three-byte encoding
37  uint8_t three = *(*utf8_data_in)++;
38  return ((one & 0x0f) << 12) | ((two & 0x3f) << 6) | (three & 0x3f);
39}
40
41inline int CompareModifiedUtf8ToModifiedUtf8AsUtf16CodePointValues(const char* utf8_1,
42                                                                   const char* utf8_2) {
43  uint16_t c1, c2;
44  do {
45    c1 = *utf8_1;
46    c2 = *utf8_2;
47    // Did we reach a terminating character?
48    if (c1 == 0) {
49      return (c2 == 0) ? 0 : -1;
50    } else if (c2 == 0) {
51      return 1;
52    }
53    // Assume 1-byte value and handle all cases first.
54    utf8_1++;
55    utf8_2++;
56    if ((c1 & 0x80) == 0) {
57      if (c1 == c2) {
58        // Matching 1-byte values.
59        continue;
60      } else {
61        // Non-matching values.
62        if ((c2 & 0x80) == 0) {
63          // 1-byte value, do nothing.
64        } else if ((c2 & 0x20) == 0) {
65          // 2-byte value.
66          c2 = ((c2 & 0x1f) << 6) | (*utf8_2 & 0x3f);
67        } else {
68          // 3-byte value.
69          c2 = ((c2 & 0x0f) << 12) | ((utf8_2[0] & 0x3f) << 6) | (utf8_2[1] & 0x3f);
70        }
71        return static_cast<int>(c1) - static_cast<int>(c2);
72      }
73    }
74    // Non-matching or multi-byte values.
75    if ((c1 & 0x20) == 0) {
76      // 2-byte value.
77      c1 = ((c1 & 0x1f) << 6) | (*utf8_1 & 0x3f);
78      utf8_1++;
79    } else {
80      // 3-byte value.
81      c1 = ((c1 & 0x0f) << 12) | ((utf8_1[0] & 0x3f) << 6) | (utf8_1[1] & 0x3f);
82      utf8_1 += 2;
83    }
84    if ((c2 & 0x80) == 0) {
85      // 1-byte value, do nothing.
86    } else if ((c2 & 0x20) == 0) {
87      // 2-byte value.
88      c2 = ((c2 & 0x1f) << 6) | (*utf8_2 & 0x3f);
89      utf8_2++;
90    } else {
91      // 3-byte value.
92      c2 = ((c2 & 0x0f) << 12) | ((utf8_2[0] & 0x3f) << 6) | (utf8_2[1] & 0x3f);
93      utf8_2 += 2;
94    }
95  } while (c1 == c2);
96  return static_cast<int>(c1) - static_cast<int>(c2);
97}
98
99}  // namespace art
100
101#endif  // ART_RUNTIME_UTF_INL_H_
102