1 // Copyright (c) 2010 The Chromium Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
4
5 #include <stddef.h>
6
7 #include "base/logging.h"
8 #include "base/macros.h"
9 #include "base/strings/string_piece.h"
10 #include "base/strings/string_util.h"
11 #include "base/strings/utf_string_conversions.h"
12 #include "build/build_config.h"
13 #include "testing/gtest/include/gtest/gtest.h"
14
15 namespace base {
16
17 namespace {
18
19 const wchar_t* const kConvertRoundtripCases[] = {
20 L"Google Video",
21 // "网页 图片 资讯更多 »"
22 L"\x7f51\x9875\x0020\x56fe\x7247\x0020\x8d44\x8baf\x66f4\x591a\x0020\x00bb",
23 // "Παγκόσμιος Ιστός"
24 L"\x03a0\x03b1\x03b3\x03ba\x03cc\x03c3\x03bc\x03b9"
25 L"\x03bf\x03c2\x0020\x0399\x03c3\x03c4\x03cc\x03c2",
26 // "Поиск страниц на русском"
27 L"\x041f\x043e\x0438\x0441\x043a\x0020\x0441\x0442"
28 L"\x0440\x0430\x043d\x0438\x0446\x0020\x043d\x0430"
29 L"\x0020\x0440\x0443\x0441\x0441\x043a\x043e\x043c",
30 // "전체서비스"
31 L"\xc804\xccb4\xc11c\xbe44\xc2a4",
32
33 // Test characters that take more than 16 bits. This will depend on whether
34 // wchar_t is 16 or 32 bits.
35 #if defined(WCHAR_T_IS_UTF16)
36 L"\xd800\xdf00",
37 // ????? (Mathematical Alphanumeric Symbols (U+011d40 - U+011d44 : A,B,C,D,E)
38 L"\xd807\xdd40\xd807\xdd41\xd807\xdd42\xd807\xdd43\xd807\xdd44",
39 #elif defined(WCHAR_T_IS_UTF32)
40 L"\x10300",
41 // ????? (Mathematical Alphanumeric Symbols (U+011d40 - U+011d44 : A,B,C,D,E)
42 L"\x11d40\x11d41\x11d42\x11d43\x11d44",
43 #endif
44 };
45
46 } // namespace
47
TEST(UTFStringConversionsTest,ConvertUTF8AndWide)48 TEST(UTFStringConversionsTest, ConvertUTF8AndWide) {
49 // we round-trip all the wide strings through UTF-8 to make sure everything
50 // agrees on the conversion. This uses the stream operators to test them
51 // simultaneously.
52 for (size_t i = 0; i < arraysize(kConvertRoundtripCases); ++i) {
53 std::ostringstream utf8;
54 utf8 << WideToUTF8(kConvertRoundtripCases[i]);
55 std::wostringstream wide;
56 wide << UTF8ToWide(utf8.str());
57
58 EXPECT_EQ(kConvertRoundtripCases[i], wide.str());
59 }
60 }
61
TEST(UTFStringConversionsTest,ConvertUTF8AndWideEmptyString)62 TEST(UTFStringConversionsTest, ConvertUTF8AndWideEmptyString) {
63 // An empty std::wstring should be converted to an empty std::string,
64 // and vice versa.
65 std::wstring wempty;
66 std::string empty;
67 EXPECT_EQ(empty, WideToUTF8(wempty));
68 EXPECT_EQ(wempty, UTF8ToWide(empty));
69 }
70
TEST(UTFStringConversionsTest,ConvertUTF8ToWide)71 TEST(UTFStringConversionsTest, ConvertUTF8ToWide) {
72 struct UTF8ToWideCase {
73 const char* utf8;
74 const wchar_t* wide;
75 bool success;
76 } convert_cases[] = {
77 // Regular UTF-8 input.
78 {"\xe4\xbd\xa0\xe5\xa5\xbd", L"\x4f60\x597d", true},
79 // Non-character is passed through.
80 {"\xef\xbf\xbfHello", L"\xffffHello", true},
81 // Truncated UTF-8 sequence.
82 {"\xe4\xa0\xe5\xa5\xbd", L"\xfffd\x597d", false},
83 // Truncated off the end.
84 {"\xe5\xa5\xbd\xe4\xa0", L"\x597d\xfffd", false},
85 // Non-shortest-form UTF-8.
86 {"\xf0\x84\xbd\xa0\xe5\xa5\xbd", L"\xfffd\x597d", false},
87 // This UTF-8 character decodes to a UTF-16 surrogate, which is illegal.
88 {"\xed\xb0\x80", L"\xfffd", false},
89 // Non-BMP characters. The second is a non-character regarded as valid.
90 // The result will either be in UTF-16 or UTF-32.
91 #if defined(WCHAR_T_IS_UTF16)
92 {"A\xF0\x90\x8C\x80z", L"A\xd800\xdf00z", true},
93 {"A\xF4\x8F\xBF\xBEz", L"A\xdbff\xdffez", true},
94 #elif defined(WCHAR_T_IS_UTF32)
95 {"A\xF0\x90\x8C\x80z", L"A\x10300z", true},
96 {"A\xF4\x8F\xBF\xBEz", L"A\x10fffez", true},
97 #endif
98 };
99
100 for (size_t i = 0; i < arraysize(convert_cases); i++) {
101 std::wstring converted;
102 EXPECT_EQ(convert_cases[i].success,
103 UTF8ToWide(convert_cases[i].utf8,
104 strlen(convert_cases[i].utf8),
105 &converted));
106 std::wstring expected(convert_cases[i].wide);
107 EXPECT_EQ(expected, converted);
108 }
109
110 // Manually test an embedded NULL.
111 std::wstring converted;
112 EXPECT_TRUE(UTF8ToWide("\00Z\t", 3, &converted));
113 ASSERT_EQ(3U, converted.length());
114 EXPECT_EQ(static_cast<wchar_t>(0), converted[0]);
115 EXPECT_EQ('Z', converted[1]);
116 EXPECT_EQ('\t', converted[2]);
117
118 // Make sure that conversion replaces, not appends.
119 EXPECT_TRUE(UTF8ToWide("B", 1, &converted));
120 ASSERT_EQ(1U, converted.length());
121 EXPECT_EQ('B', converted[0]);
122 }
123
124 #if defined(WCHAR_T_IS_UTF16)
125 // This test is only valid when wchar_t == UTF-16.
TEST(UTFStringConversionsTest,ConvertUTF16ToUTF8)126 TEST(UTFStringConversionsTest, ConvertUTF16ToUTF8) {
127 struct WideToUTF8Case {
128 const wchar_t* utf16;
129 const char* utf8;
130 bool success;
131 } convert_cases[] = {
132 // Regular UTF-16 input.
133 {L"\x4f60\x597d", "\xe4\xbd\xa0\xe5\xa5\xbd", true},
134 // Test a non-BMP character.
135 {L"\xd800\xdf00", "\xF0\x90\x8C\x80", true},
136 // Non-characters are passed through.
137 {L"\xffffHello", "\xEF\xBF\xBFHello", true},
138 {L"\xdbff\xdffeHello", "\xF4\x8F\xBF\xBEHello", true},
139 // The first character is a truncated UTF-16 character.
140 {L"\xd800\x597d", "\xef\xbf\xbd\xe5\xa5\xbd", false},
141 // Truncated at the end.
142 {L"\x597d\xd800", "\xe5\xa5\xbd\xef\xbf\xbd", false},
143 };
144
145 for (const auto& test : convert_cases) {
146 std::string converted;
147 EXPECT_EQ(test.success,
148 WideToUTF8(test.utf16, wcslen(test.utf16), &converted));
149 std::string expected(test.utf8);
150 EXPECT_EQ(expected, converted);
151 }
152 }
153
154 #elif defined(WCHAR_T_IS_UTF32)
155 // This test is only valid when wchar_t == UTF-32.
TEST(UTFStringConversionsTest,ConvertUTF32ToUTF8)156 TEST(UTFStringConversionsTest, ConvertUTF32ToUTF8) {
157 struct WideToUTF8Case {
158 const wchar_t* utf32;
159 const char* utf8;
160 bool success;
161 } convert_cases[] = {
162 // Regular 16-bit input.
163 {L"\x4f60\x597d", "\xe4\xbd\xa0\xe5\xa5\xbd", true},
164 // Test a non-BMP character.
165 {L"A\x10300z", "A\xF0\x90\x8C\x80z", true},
166 // Non-characters are passed through.
167 {L"\xffffHello", "\xEF\xBF\xBFHello", true},
168 {L"\x10fffeHello", "\xF4\x8F\xBF\xBEHello", true},
169 // Invalid Unicode code points.
170 {L"\xfffffffHello", "\xEF\xBF\xBDHello", false},
171 // The first character is a truncated UTF-16 character.
172 {L"\xd800\x597d", "\xef\xbf\xbd\xe5\xa5\xbd", false},
173 {L"\xdc01Hello", "\xef\xbf\xbdHello", false},
174 };
175
176 for (const auto& test : convert_cases) {
177 std::string converted;
178 EXPECT_EQ(test.success,
179 WideToUTF8(test.utf32, wcslen(test.utf32), &converted));
180 std::string expected(test.utf8);
181 EXPECT_EQ(expected, converted);
182 }
183 }
184 #endif // defined(WCHAR_T_IS_UTF32)
185
TEST(UTFStringConversionsTest,ConvertMultiString)186 TEST(UTFStringConversionsTest, ConvertMultiString) {
187 static char16 multi16[] = {
188 'f', 'o', 'o', '\0',
189 'b', 'a', 'r', '\0',
190 'b', 'a', 'z', '\0',
191 '\0'
192 };
193 static char multi[] = {
194 'f', 'o', 'o', '\0',
195 'b', 'a', 'r', '\0',
196 'b', 'a', 'z', '\0',
197 '\0'
198 };
199 string16 multistring16;
200 memcpy(WriteInto(&multistring16, arraysize(multi16)), multi16,
201 sizeof(multi16));
202 EXPECT_EQ(arraysize(multi16) - 1, multistring16.length());
203 std::string expected;
204 memcpy(WriteInto(&expected, arraysize(multi)), multi, sizeof(multi));
205 EXPECT_EQ(arraysize(multi) - 1, expected.length());
206 const std::string& converted = UTF16ToUTF8(multistring16);
207 EXPECT_EQ(arraysize(multi) - 1, converted.length());
208 EXPECT_EQ(expected, converted);
209 }
210
211 } // namespace base
212