base/strings/utf_string_conversions_unittest.cc

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