1 // Copyright (c) 2011 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 <algorithm>
6
7 #include "base/logging.h"
8 #include "base/string_piece.h"
9 #include "base/utf_offset_string_conversions.h"
10 #include "testing/gtest/include/gtest/gtest.h"
11
12 namespace base {
13
14 namespace {
15
16 static const size_t kNpos = std::wstring::npos;
17
18 // Given a null-terminated string of wchar_t with each wchar_t representing
19 // a UTF-16 code unit, returns a string16 made up of wchar_t's in the input.
20 // Each wchar_t should be <= 0xFFFF and a non-BMP character (> U+FFFF)
21 // should be represented as a surrogate pair (two UTF-16 units)
22 // *even* where wchar_t is 32-bit (Linux and Mac).
23 //
24 // This is to help write tests for functions with string16 params until
25 // the C++ 0x UTF-16 literal is well-supported by compilers.
BuildString16(const wchar_t * s)26 string16 BuildString16(const wchar_t* s) {
27 #if defined(WCHAR_T_IS_UTF16)
28 return string16(s);
29 #elif defined(WCHAR_T_IS_UTF32)
30 string16 u16;
31 while (*s != 0) {
32 DCHECK(static_cast<unsigned int>(*s) <= 0xFFFFu);
33 u16.push_back(*s++);
34 }
35 return u16;
36 #endif
37 }
38
39 } // namespace
40
TEST(UTFOffsetStringConversionsTest,AdjustOffset)41 TEST(UTFOffsetStringConversionsTest, AdjustOffset) {
42 struct UTF8ToWideCase {
43 const char* utf8;
44 size_t input_offset;
45 size_t output_offset;
46 } utf8_to_wide_cases[] = {
47 {"", 0, kNpos},
48 {"\xe4\xbd\xa0\xe5\xa5\xbd", 1, kNpos},
49 {"\xe4\xbd\xa0\xe5\xa5\xbd", 3, 1},
50 {"\xed\xb0\x80z", 3, 1},
51 {"A\xF0\x90\x8C\x80z", 1, 1},
52 {"A\xF0\x90\x8C\x80z", 2, kNpos},
53 #if defined(WCHAR_T_IS_UTF16)
54 {"A\xF0\x90\x8C\x80z", 5, 3},
55 #elif defined(WCHAR_T_IS_UTF32)
56 {"A\xF0\x90\x8C\x80z", 5, 2},
57 #endif
58 };
59 for (size_t i = 0; i < ARRAYSIZE_UNSAFE(utf8_to_wide_cases); ++i) {
60 size_t offset = utf8_to_wide_cases[i].input_offset;
61 UTF8ToWideAndAdjustOffset(utf8_to_wide_cases[i].utf8, &offset);
62 EXPECT_EQ(utf8_to_wide_cases[i].output_offset, offset);
63 }
64
65 #if defined(WCHAR_T_IS_UTF32)
66 struct UTF16ToWideCase {
67 const wchar_t* wide;
68 size_t input_offset;
69 size_t output_offset;
70 } utf16_to_wide_cases[] = {
71 {L"\xD840\xDC00\x4E00", 0, 0},
72 {L"\xD840\xDC00\x4E00", 1, kNpos},
73 {L"\xD840\xDC00\x4E00", 2, 1},
74 };
75 for (size_t i = 0; i < ARRAYSIZE_UNSAFE(utf16_to_wide_cases); ++i) {
76 size_t offset = utf16_to_wide_cases[i].input_offset;
77 UTF16ToWideAndAdjustOffset(BuildString16(utf16_to_wide_cases[i].wide),
78 &offset);
79 EXPECT_EQ(utf16_to_wide_cases[i].output_offset, offset);
80 }
81 #endif
82 }
83
TEST(UTFOffsetStringConversionsTest,LimitOffsets)84 TEST(UTFOffsetStringConversionsTest, LimitOffsets) {
85 const size_t kLimit = 10;
86 const size_t kItems = 20;
87 std::vector<size_t> size_ts;
88 for (size_t t = 0; t < kItems; ++t)
89 size_ts.push_back(t);
90 std::for_each(size_ts.begin(), size_ts.end(),
91 LimitOffset<std::wstring>(kLimit));
92 size_t unlimited_count = 0;
93 for (std::vector<size_t>::iterator ti = size_ts.begin(); ti != size_ts.end();
94 ++ti) {
95 if (*ti < kLimit && *ti != kNpos)
96 ++unlimited_count;
97 }
98 EXPECT_EQ(10U, unlimited_count);
99
100 // Reverse the values in the vector and try again.
101 size_ts.clear();
102 for (size_t t = kItems; t > 0; --t)
103 size_ts.push_back(t - 1);
104 std::for_each(size_ts.begin(), size_ts.end(),
105 LimitOffset<std::wstring>(kLimit));
106 unlimited_count = 0;
107 for (std::vector<size_t>::iterator ti = size_ts.begin(); ti != size_ts.end();
108 ++ti) {
109 if (*ti < kLimit && *ti != kNpos)
110 ++unlimited_count;
111 }
112 EXPECT_EQ(10U, unlimited_count);
113 }
114
TEST(UTFOffsetStringConversionsTest,AdjustOffsets)115 TEST(UTFOffsetStringConversionsTest, AdjustOffsets) {
116 // Imagine we have strings as shown in the following cases where the
117 // X's represent encoded characters.
118 // 1: abcXXXdef ==> abcXdef
119 std::vector<size_t> offsets;
120 for (size_t t = 0; t < 9; ++t)
121 offsets.push_back(t);
122 AdjustOffset::Adjustments adjustments;
123 adjustments.push_back(AdjustOffset::Adjustment(3, 3, 1));
124 std::for_each(offsets.begin(), offsets.end(), AdjustOffset(adjustments));
125 size_t expected_1[] = {0, 1, 2, 3, kNpos, kNpos, 4, 5, 6};
126 EXPECT_EQ(offsets.size(), arraysize(expected_1));
127 for (size_t i = 0; i < arraysize(expected_1); ++i)
128 EXPECT_EQ(expected_1[i], offsets[i]);
129
130 // 2: XXXaXXXXbcXXXXXXXdefXXX ==> XaXXbcXXXXdefX
131 offsets.clear();
132 for (size_t t = 0; t < 23; ++t)
133 offsets.push_back(t);
134 adjustments.clear();
135 adjustments.push_back(AdjustOffset::Adjustment(0, 3, 1));
136 adjustments.push_back(AdjustOffset::Adjustment(4, 4, 2));
137 adjustments.push_back(AdjustOffset::Adjustment(10, 7, 4));
138 adjustments.push_back(AdjustOffset::Adjustment(20, 3, 1));
139 std::for_each(offsets.begin(), offsets.end(), AdjustOffset(adjustments));
140 size_t expected_2[] = {0, kNpos, kNpos, 1, 2, kNpos, kNpos, kNpos, 4, 5, 6,
141 kNpos, kNpos, kNpos, kNpos, kNpos, kNpos, 10, 11, 12,
142 13, kNpos, kNpos};
143 EXPECT_EQ(offsets.size(), arraysize(expected_2));
144 for (size_t i = 0; i < arraysize(expected_2); ++i)
145 EXPECT_EQ(expected_2[i], offsets[i]);
146
147 // 3: XXXaXXXXbcdXXXeXX ==> aXXXXbcdXXXe
148 offsets.clear();
149 for (size_t t = 0; t < 17; ++t)
150 offsets.push_back(t);
151 adjustments.clear();
152 adjustments.push_back(AdjustOffset::Adjustment(0, 3, 0));
153 adjustments.push_back(AdjustOffset::Adjustment(4, 4, 4));
154 adjustments.push_back(AdjustOffset::Adjustment(11, 3, 3));
155 adjustments.push_back(AdjustOffset::Adjustment(15, 2, 0));
156 std::for_each(offsets.begin(), offsets.end(), AdjustOffset(adjustments));
157 size_t expected_3[] = {kNpos, kNpos, kNpos, 0, 1, kNpos, kNpos, kNpos, 5, 6,
158 7, 8, kNpos, kNpos, 11, kNpos, kNpos};
159 EXPECT_EQ(offsets.size(), arraysize(expected_3));
160 for (size_t i = 0; i < arraysize(expected_3); ++i)
161 EXPECT_EQ(expected_3[i], offsets[i]);
162 }
163
164 } // namaspace base
165