base/i18n/file_util_icu_unittest.cc

   1 // Copyright (c) 2012 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/i18n/file_util_icu.h"
   6
   7 #include "base/files/file_util.h"
   8 #include "base/strings/utf_string_conversions.h"
   9 #include "testing/gtest/include/gtest/gtest.h"
  10 #include "testing/platform_test.h"
  11
  12 namespace base {
  13 namespace i18n {
  14
  15 // file_util winds up using autoreleased objects on the Mac, so this needs
  16 // to be a PlatformTest
  17 class FileUtilICUTest : public PlatformTest {
  18 };
  19
  20 #if defined(OS_POSIX) && !defined(OS_MACOSX)
  21
  22 // Linux disallows some evil ASCII characters, but passes all non-ASCII.
  23 static const struct GoodBadPairLinux {
  24   const char* bad_name;
  25   const char* good_name;
  26 } kLinuxIllegalCharacterCases[] = {
  27   {"bad*\\/file:name?.jpg", "bad---file-name-.jpg"},
  28   {"**********::::.txt", "--------------.txt"},
  29   {"\xe9\xf0zzzz.\xff", "\xe9\xf0zzzz.\xff"},
  30   {" _ ", "-_-"},
  31   {".", "-"},
  32   {" .( ). ", "-.( ).-"},
  33   {"     ", "-   -"},
  34 };
  35
  36 TEST_F(FileUtilICUTest, ReplaceIllegalCharacersInPathLinuxTest) {
  37   for (size_t i = 0; i < arraysize(kLinuxIllegalCharacterCases); ++i) {
  38     std::string bad_name(kLinuxIllegalCharacterCases[i].bad_name);
  39     ReplaceIllegalCharactersInPath(&bad_name, '-');
  40     EXPECT_EQ(kLinuxIllegalCharacterCases[i].good_name, bad_name);
  41   }
  42 }
  43
  44 #endif
  45
  46 // For Mac & Windows, which both do Unicode validation on filenames. These
  47 // characters are given as wide strings since its more convenient to specify
  48 // unicode characters. For Mac they should be converted to UTF-8.
  49 static const struct goodbad_pair {
  50   const wchar_t* bad_name;
  51   const wchar_t* good_name;
  52 } kIllegalCharacterCases[] = {
  53     {L"bad*file:name?.jpg", L"bad-file-name-.jpg"},
  54     {L"**********::::.txt", L"--------------.txt"},
  55     // We can't use UCNs (universal character names) for C0/C1 characters and
  56     // U+007F, but \x escape is interpreted by MSVC and gcc as we intend.
  57     {L"bad\x0003\x0091 file\u200E\u200Fname.png", L"bad-- file--name.png"},
  58     {L"bad*file\\?name.jpg", L"bad-file--name.jpg"},
  59     {L"\t  bad*file\\name/.jpg", L"-  bad-file-name-.jpg"},
  60     {L"this_file_name is okay!.mp3", L"this_file_name is okay!.mp3"},
  61     {L"\u4E00\uAC00.mp3", L"\u4E00\uAC00.mp3"},
  62     {L"\u0635\u200C\u0644.mp3", L"\u0635-\u0644.mp3"},
  63     {L"\U00010330\U00010331.mp3", L"\U00010330\U00010331.mp3"},
  64     // Unassigned codepoints are ok.
  65     {L"\u0378\U00040001.mp3", L"\u0378\U00040001.mp3"},
  66     // Non-characters are not allowed.
  67     {L"bad\uFFFFfile\U0010FFFEname.jpg", L"bad-file-name.jpg"},
  68     {L"bad\uFDD0file\uFDEFname.jpg", L"bad-file-name.jpg"},
  69     // CVE-2014-9390
  70     {L"(\u200C.\u200D.\u200E.\u200F.\u202A.\u202B.\u202C.\u202D.\u202E.\u206A."
  71      L"\u206B.\u206C.\u206D.\u206F.\uFEFF)",
  72      L"(-.-.-.-.-.-.-.-.-.-.-.-.-.-.-)"},
  73     {L"config~1", L"config-1"},
  74     {L" _ ", L"-_-"},
  75     {L" ", L"-"},
  76     {L"\u2008.(\u2007).\u3000", L"-.(\u2007).-"},
  77     {L"     ", L"-   -"},
  78     {L".    ", L"-   -"}
  79 };
  80
  81 #if defined(OS_WIN) || defined(OS_MACOSX)
  82
  83 TEST_F(FileUtilICUTest, ReplaceIllegalCharactersInPathTest) {
  84   for (size_t i = 0; i < arraysize(kIllegalCharacterCases); ++i) {
  85 #if defined(OS_WIN)
  86     std::wstring bad_name(kIllegalCharacterCases[i].bad_name);
  87     ReplaceIllegalCharactersInPath(&bad_name, '-');
  88     EXPECT_EQ(kIllegalCharacterCases[i].good_name, bad_name);
  89 #elif defined(OS_MACOSX)
  90     std::string bad_name(WideToUTF8(kIllegalCharacterCases[i].bad_name));
  91     ReplaceIllegalCharactersInPath(&bad_name, '-');
  92     EXPECT_EQ(WideToUTF8(kIllegalCharacterCases[i].good_name), bad_name);
  93 #endif
  94   }
  95 }
  96
  97 #endif
  98
  99 TEST_F(FileUtilICUTest, IsFilenameLegalTest) {
 100   EXPECT_TRUE(IsFilenameLegal(string16()));
 101
 102   for (const auto& test_case : kIllegalCharacterCases) {
 103     string16 bad_name = WideToUTF16(test_case.bad_name);
 104     string16 good_name = WideToUTF16(test_case.good_name);
 105
 106     EXPECT_TRUE(IsFilenameLegal(good_name)) << good_name;
 107     if (good_name != bad_name)
 108       EXPECT_FALSE(IsFilenameLegal(bad_name)) << bad_name;
 109   }
 110 }
 111
 112 #if defined(OS_CHROMEOS)
 113 static const struct normalize_name_encoding_test_cases {
 114   const char* original_path;
 115   const char* normalized_path;
 116 } kNormalizeFileNameEncodingTestCases[] = {
 117   { "foo_na\xcc\x88me.foo", "foo_n\xc3\xa4me.foo"},
 118   { "foo_dir_na\xcc\x88me/foo_na\xcc\x88me.foo",
 119     "foo_dir_na\xcc\x88me/foo_n\xc3\xa4me.foo"},
 120   { "", ""},
 121   { "foo_dir_na\xcc\x88me/", "foo_dir_n\xc3\xa4me"}
 122 };
 123
 124 TEST_F(FileUtilICUTest, NormalizeFileNameEncoding) {
 125   for (size_t i = 0; i < arraysize(kNormalizeFileNameEncodingTestCases); i++) {
 126     FilePath path(kNormalizeFileNameEncodingTestCases[i].original_path);
 127     NormalizeFileNameEncoding(&path);
 128     EXPECT_EQ(FilePath(kNormalizeFileNameEncodingTestCases[i].normalized_path),
 129               path);
 130   }
 131 }
 132
 133 #endif
 134
 135 }  // namespace i18n
 136 }  // namespace base