1 // Copyright 2014 The Chromium Authors
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
4
5 #ifdef UNSAFE_BUFFERS_BUILD
6 // TODO(crbug.com/350788890): Remove this and spanify to fix the errors.
7 #pragma allow_unsafe_buffers
8 #endif
9
10 #include "url/url_canon_icu.h"
11
12 #include <stddef.h>
13
14 #include "base/logging.h"
15 #include "base/memory/raw_ptr.h"
16 #include "testing/gtest/include/gtest/gtest.h"
17 #include "third_party/icu/source/common/unicode/ucnv.h"
18 #include "url/url_canon.h"
19 #include "url/url_canon_icu_test_helpers.h"
20 #include "url/url_canon_stdstring.h"
21 #include "url/url_test_utils.h"
22
23 namespace url {
24
25 namespace {
26
TEST(URLCanonIcuTest,ICUCharsetConverter)27 TEST(URLCanonIcuTest, ICUCharsetConverter) {
28 struct ICUCase {
29 const wchar_t* input;
30 const char* encoding;
31 const char* expected;
32 } icu_cases[] = {
33 // UTF-8.
34 {L"Hello, world", "utf-8", "Hello, world"},
35 {L"\x4f60\x597d", "utf-8", "\xe4\xbd\xa0\xe5\xa5\xbd"},
36 // Non-BMP UTF-8.
37 {L"!\xd800\xdf00!", "utf-8", "!\xf0\x90\x8c\x80!"},
38 // Big5
39 {L"\x4f60\x597d", "big5", "\xa7\x41\xa6\x6e"},
40 // Unrepresentable character in the destination set.
41 {L"hello\x4f60\x06de\x597dworld", "big5",
42 "hello\xa7\x41%26%231758%3B\xa6\x6eworld"},
43 };
44
45 for (size_t i = 0; i < std::size(icu_cases); i++) {
46 test::UConvScoper conv(icu_cases[i].encoding);
47 ASSERT_TRUE(conv.converter() != NULL);
48 ICUCharsetConverter converter(conv.converter());
49
50 std::string str;
51 StdStringCanonOutput output(&str);
52
53 std::u16string input_str(
54 test_utils::TruncateWStringToUTF16(icu_cases[i].input));
55 converter.ConvertFromUTF16(input_str, &output);
56 output.Complete();
57
58 EXPECT_STREQ(icu_cases[i].expected, str.c_str());
59 }
60
61 // Test string sizes around the resize boundary for the output to make sure
62 // the converter resizes as needed.
63 const int static_size = 16;
64 test::UConvScoper conv("utf-8");
65 ASSERT_TRUE(conv.converter());
66 ICUCharsetConverter converter(conv.converter());
67 for (int i = static_size - 2; i <= static_size + 2; i++) {
68 // Make a string with the appropriate length.
69 std::u16string input;
70 for (int ch = 0; ch < i; ch++)
71 input.push_back('a');
72
73 RawCanonOutput<static_size> output;
74 converter.ConvertFromUTF16(input, &output);
75 EXPECT_EQ(input.length(), output.length());
76 }
77 }
78
TEST(URLCanonIcuTest,QueryWithConverter)79 TEST(URLCanonIcuTest, QueryWithConverter) {
80 struct QueryCase {
81 const char* input8;
82 const wchar_t* input16;
83 const char* encoding;
84 const char* expected;
85 } query_cases[] = {
86 // Regular ASCII case in some different encodings.
87 {"foo=bar", L"foo=bar", "utf-8", "?foo=bar"},
88 {"foo=bar", L"foo=bar", "shift_jis", "?foo=bar"},
89 {"foo=bar", L"foo=bar", "gb2312", "?foo=bar"},
90 // Chinese input/output
91 {"q=\xe4\xbd\xa0\xe5\xa5\xbd", L"q=\x4f60\x597d", "gb2312",
92 "?q=%C4%E3%BA%C3"},
93 {"q=\xe4\xbd\xa0\xe5\xa5\xbd", L"q=\x4f60\x597d", "big5", "?q=%A7A%A6n"},
94 // Unencodable character in the destination character set should be
95 // escaped. The escape sequence unescapes to be the entity name:
96 // "?q=你"
97 {"q=Chinese\xef\xbc\xa7", L"q=Chinese\xff27", "iso-8859-1",
98 "?q=Chinese%26%2365319%3B"},
99 };
100
101 for (size_t i = 0; i < std::size(query_cases); i++) {
102 Component out_comp;
103
104 test::UConvScoper conv(query_cases[i].encoding);
105 ASSERT_TRUE(!query_cases[i].encoding || conv.converter());
106 ICUCharsetConverter converter(conv.converter());
107
108 if (query_cases[i].input8) {
109 int len = static_cast<int>(strlen(query_cases[i].input8));
110 Component in_comp(0, len);
111 std::string out_str;
112
113 StdStringCanonOutput output(&out_str);
114 CanonicalizeQuery(query_cases[i].input8, in_comp, &converter, &output,
115 &out_comp);
116 output.Complete();
117
118 EXPECT_EQ(query_cases[i].expected, out_str);
119 }
120
121 if (query_cases[i].input16) {
122 std::u16string input16(
123 test_utils::TruncateWStringToUTF16(query_cases[i].input16));
124 int len = static_cast<int>(input16.length());
125 Component in_comp(0, len);
126 std::string out_str;
127
128 StdStringCanonOutput output(&out_str);
129 CanonicalizeQuery(input16.c_str(), in_comp, &converter, &output,
130 &out_comp);
131 output.Complete();
132
133 EXPECT_EQ(query_cases[i].expected, out_str);
134 }
135 }
136
137 // Extra test for input with embedded NULL;
138 std::string out_str;
139 StdStringCanonOutput output(&out_str);
140 Component out_comp;
141 CanonicalizeQuery("a \x00z\x01", Component(0, 5), NULL, &output, &out_comp);
142 output.Complete();
143 EXPECT_EQ("?a%20%00z%01", out_str);
144 }
145
146 } // namespace
147
148 } // namespace url
149