1 // Copyright (c) 2009 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/file_util.h"
8 #include "base/path_service.h"
9 #include "base/utf_string_conversions.h"
10 #include "testing/gtest/include/gtest/gtest.h"
11 #include "testing/platform_test.h"
12
13 // file_util winds up using autoreleased objects on the Mac, so this needs
14 // to be a PlatformTest
15 class FileUtilICUTest : public PlatformTest {
16 protected:
SetUp()17 virtual void SetUp() {
18 PlatformTest::SetUp();
19 // Name a subdirectory of the temp directory.
20 ASSERT_TRUE(PathService::Get(base::DIR_TEMP, &test_dir_));
21 test_dir_ = test_dir_.Append(FILE_PATH_LITERAL("FileUtilTest"));
22
23 // Create a fresh, empty copy of this directory.
24 file_util::Delete(test_dir_, true);
25 file_util::CreateDirectory(test_dir_);
26 }
TearDown()27 virtual void TearDown() {
28 PlatformTest::TearDown();
29 // Clean up test directory
30 ASSERT_TRUE(file_util::Delete(test_dir_, true));
31 ASSERT_FALSE(file_util::PathExists(test_dir_));
32 }
33
34 // the path to temporary directory used to contain the test operations
35 FilePath test_dir_;
36 };
37
38 #if defined(OS_POSIX) && !defined(OS_MACOSX)
39
40 // Linux disallows some evil ASCII characters, but passes all non-ASCII.
41 static const struct goodbad_pair {
42 const char* bad_name;
43 const char* good_name;
44 } kIllegalCharacterCases[] = {
45 {"bad*file:name?.jpg", "bad-file-name-.jpg"},
46 {"**********::::.txt", "--------------.txt"},
47 {"\xe9\xf0zzzz.\xff", "\xe9\xf0zzzz.\xff"},
48 };
49
TEST_F(FileUtilICUTest,ReplaceIllegalCharacersInPathLinuxTest)50 TEST_F(FileUtilICUTest, ReplaceIllegalCharacersInPathLinuxTest) {
51 for (size_t i = 0; i < arraysize(kIllegalCharacterCases); ++i) {
52 std::string bad_name(kIllegalCharacterCases[i].bad_name);
53 file_util::ReplaceIllegalCharactersInPath(&bad_name, '-');
54 EXPECT_EQ(kIllegalCharacterCases[i].good_name, bad_name);
55 }
56 }
57
58 #else
59
60 // For Mac & Windows, which both do Unicode validation on filenames. These
61 // characters are given as wide strings since its more convenient to specify
62 // unicode characters. For Mac they should be converted to UTF-8.
63 static const struct goodbad_pair {
64 const wchar_t* bad_name;
65 const wchar_t* good_name;
66 } kIllegalCharacterCases[] = {
67 {L"bad*file:name?.jpg", L"bad-file-name-.jpg"},
68 {L"**********::::.txt", L"--------------.txt"},
69 // We can't use UCNs (universal character names) for C0/C1 characters and
70 // U+007F, but \x escape is interpreted by MSVC and gcc as we intend.
71 {L"bad\x0003\x0091 file\u200E\u200Fname.png", L"bad-- file--name.png"},
72 #if defined(OS_WIN)
73 {L"bad*file\\name.jpg", L"bad-file-name.jpg"},
74 {L"\t bad*file\\name/.jpg ", L"bad-file-name-.jpg"},
75 #elif defined(OS_MACOSX)
76 {L"bad*file?name.jpg", L"bad-file-name.jpg"},
77 {L"\t bad*file?name/.jpg ", L"bad-file-name-.jpg"},
78 #endif
79 {L"this_file_name is okay!.mp3", L"this_file_name is okay!.mp3"},
80 {L"\u4E00\uAC00.mp3", L"\u4E00\uAC00.mp3"},
81 {L"\u0635\u200C\u0644.mp3", L"\u0635\u200C\u0644.mp3"},
82 {L"\U00010330\U00010331.mp3", L"\U00010330\U00010331.mp3"},
83 // Unassigned codepoints are ok.
84 {L"\u0378\U00040001.mp3", L"\u0378\U00040001.mp3"},
85 // Non-characters are not allowed.
86 {L"bad\uFFFFfile\U0010FFFEname.jpg ", L"bad-file-name.jpg"},
87 {L"bad\uFDD0file\uFDEFname.jpg ", L"bad-file-name.jpg"},
88 };
89
TEST_F(FileUtilICUTest,ReplaceIllegalCharactersInPathTest)90 TEST_F(FileUtilICUTest, ReplaceIllegalCharactersInPathTest) {
91 for (size_t i = 0; i < arraysize(kIllegalCharacterCases); ++i) {
92 #if defined(OS_WIN)
93 std::wstring bad_name(kIllegalCharacterCases[i].bad_name);
94 file_util::ReplaceIllegalCharactersInPath(&bad_name, '-');
95 EXPECT_EQ(kIllegalCharacterCases[i].good_name, bad_name);
96 #elif defined(OS_MACOSX)
97 std::string bad_name(WideToUTF8(kIllegalCharacterCases[i].bad_name));
98 file_util::ReplaceIllegalCharactersInPath(&bad_name, '-');
99 EXPECT_EQ(WideToUTF8(kIllegalCharacterCases[i].good_name), bad_name);
100 #endif
101 }
102 }
103
104 #endif
105
106