OLD | NEW |
1 // Copyright (c) 2013, the Dart project authors. Please see the AUTHORS file | 1 // Copyright (c) 2013, the Dart project authors. Please see the AUTHORS file |
2 // for details. All rights reserved. Use of this source code is governed by a | 2 // for details. All rights reserved. Use of this source code is governed by a |
3 // BSD-style license that can be found in the LICENSE file. | 3 // BSD-style license that can be found in the LICENSE file. |
4 | 4 |
5 library utf8_test; | |
6 import "package:expect/expect.dart"; | 5 import "package:expect/expect.dart"; |
7 import 'dart:convert'; | 6 import 'dart:convert'; |
8 | 7 import 'unicode_tests.dart'; |
9 // Google favorite: "Îñţérñåţîöñåļîžåţîờñ". | |
10 const INTER_BYTES = const [0xc3, 0x8e, 0xc3, 0xb1, 0xc5, 0xa3, 0xc3, 0xa9, 0x72, | |
11 0xc3, 0xb1, 0xc3, 0xa5, 0xc5, 0xa3, 0xc3, 0xae, 0xc3, | |
12 0xb6, 0xc3, 0xb1, 0xc3, 0xa5, 0xc4, 0xbc, 0xc3, 0xae, | |
13 0xc5, 0xbe, 0xc3, 0xa5, 0xc5, 0xa3, 0xc3, 0xae, 0xe1, | |
14 0xbb, 0x9d, 0xc3, 0xb1]; | |
15 const INTER_STRING = "Îñţérñåţîöñåļîžåţîờñ"; | |
16 | |
17 // Blueberry porridge in Danish: "blåbærgrød". | |
18 const BLUEBERRY_BYTES = const [0x62, 0x6c, 0xc3, 0xa5, 0x62, 0xc3, 0xa6, 0x72, | |
19 0x67, 0x72, 0xc3, 0xb8, 0x64]; | |
20 const BLUEBERRY_STRING = "blåbærgrød"; | |
21 | |
22 // "சிவா அணாமாைல", that is "Siva Annamalai" in Tamil. | |
23 const SIVA_BYTES1 = const [0xe0, 0xae, 0x9a, 0xe0, 0xae, 0xbf, 0xe0, 0xae, 0xb5, | |
24 0xe0, 0xae, 0xbe, 0x20, 0xe0, 0xae, 0x85, 0xe0, 0xae, | |
25 0xa3, 0xe0, 0xae, 0xbe, 0xe0, 0xae, 0xae, 0xe0, 0xae, | |
26 0xbe, 0xe0, 0xaf, 0x88, 0xe0, 0xae, 0xb2]; | |
27 const SIVA_STRING1 = "சிவா அணாமாைல"; | |
28 | |
29 // "िसवा अणामालै", that is "Siva Annamalai" in Devanagari. | |
30 const SIVA_BYTES2 = const [0xe0, 0xa4, 0xbf, 0xe0, 0xa4, 0xb8, 0xe0, 0xa4, 0xb5, | |
31 0xe0, 0xa4, 0xbe, 0x20, 0xe0, 0xa4, 0x85, 0xe0, 0xa4, | |
32 0xa3, 0xe0, 0xa4, 0xbe, 0xe0, 0xa4, 0xae, 0xe0, 0xa4, | |
33 0xbe, 0xe0, 0xa4, 0xb2, 0xe0, 0xa5, 0x88]; | |
34 const SIVA_STRING2 = "िसवा अणामालै"; | |
35 | |
36 // DESERET CAPITAL LETTER BEE, unicode 0x10412(0xD801+0xDC12) | |
37 // UTF-8: F0 90 90 92 | |
38 const BEE_BYTES = const [0xf0, 0x90, 0x90, 0x92]; | |
39 const BEE_STRING = "𐐒"; | |
40 | |
41 const DIGIT_BYTES = const [ 0x35 ]; | |
42 const DIGIT_STRING = "5"; | |
43 | |
44 const ASCII_BYTES = const [ 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, | |
45 0x69, 0x6A, 0x6B, 0x6C, 0x6D, 0x6E, 0x6F, 0x70, | |
46 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, | |
47 0x79, 0x7A ]; | |
48 const ASCII_STRING = "abcdefghijklmnopqrstuvwxyz"; | |
49 | |
50 const TESTS = const [ | |
51 const [ const [], "" ], | |
52 const [ INTER_BYTES, INTER_STRING ], | |
53 const [ BLUEBERRY_BYTES, BLUEBERRY_STRING ], | |
54 const [ SIVA_BYTES1, SIVA_STRING1 ], | |
55 const [ SIVA_BYTES2, SIVA_STRING2 ], | |
56 const [ BEE_BYTES, BEE_STRING ], | |
57 const [ DIGIT_BYTES, DIGIT_STRING ], | |
58 const [ ASCII_BYTES, ASCII_STRING ]]; | |
59 | 8 |
60 List<int> encode(String str) => new Utf8Encoder().convert(str); | 9 List<int> encode(String str) => new Utf8Encoder().convert(str); |
61 List<int> encode2(String str) => UTF8.encode(str); | 10 List<int> encode2(String str) => UTF8.encode(str); |
62 | 11 |
63 main() { | 12 main() { |
64 Expect.equals(2, BEE_STRING.length); | 13 for (var test in UNICODE_TESTS) { |
65 var tests = TESTS.expand((test) { | |
66 var bytes = test[0]; | |
67 var string = test[1]; | |
68 var longBytes = []; | |
69 var longString = ""; | |
70 for (int i = 0; i < 100; i++) { | |
71 longBytes.addAll(bytes); | |
72 longString += string; | |
73 } | |
74 return [test, [longBytes, longString]]; | |
75 }); | |
76 for (var test in tests) { | |
77 List<int> bytes = test[0]; | 14 List<int> bytes = test[0]; |
78 String string = test[1]; | 15 String string = test[1]; |
79 Expect.listEquals(bytes, encode(string)); | 16 Expect.listEquals(bytes, encode(string)); |
80 Expect.listEquals(bytes, encode2(string)); | 17 Expect.listEquals(bytes, encode2(string)); |
81 } | 18 } |
82 } | 19 } |
OLD | NEW |