jcs.org
Serenity Operating System
1/*
2 * Copyright (c) 2019-2020, Sergey Bugaev <bugaevc@serenityos.org>
3 *
4 * SPDX-License-Identifier: BSD-2-Clause
5 */
6
7#include <LibTest/TestCase.h>
8
9#include <AK/ByteBuffer.h>
10#include <AK/Utf8View.h>
11
12TEST_CASE(decode_ascii)
13{
14 Utf8View utf8 { "Hello World!11"sv };
15 EXPECT(utf8.validate());
16
17 u32 expected[] = { 72, 101, 108, 108, 111, 32, 87, 111, 114, 108, 100, 33, 49, 49 };
18 size_t expected_size = sizeof(expected) / sizeof(expected[0]);
19
20 size_t i = 0;
21 for (u32 code_point : utf8) {
22 VERIFY(i < expected_size);
23 EXPECT_EQ(code_point, expected[i]);
24 i++;
25 }
26 EXPECT_EQ(i, expected_size);
27}
28
29TEST_CASE(decode_utf8)
30{
31 Utf8View utf8 { "Привет, мир! 😀 γειά σου κόσμος こんにちは世界"sv };
32 size_t valid_bytes;
33 EXPECT(utf8.validate(valid_bytes));
34 EXPECT(valid_bytes == (size_t)utf8.byte_length());
35
36 u32 expected[] = { 1055, 1088, 1080, 1074, 1077, 1090, 44, 32, 1084, 1080, 1088, 33, 32, 128512, 32, 947, 949, 953, 940, 32, 963, 959, 965, 32, 954, 972, 963, 956, 959, 962, 32, 12371, 12435, 12395, 12385, 12399, 19990, 30028 };
37 DeprecatedString expected_underlying_bytes[] = { "П", "р", "и", "в", "е", "т", ",", " ", "м", "и", "р", "!", " ", "😀", " ", "γ", "ε", "ι", "ά", " ", "σ", "ο", "υ", " ", "κ", "ό", "σ", "μ", "ο", "ς", " ", "こ", "ん", "に", "ち", "は", "世", "界" };
38 size_t expected_size = sizeof(expected) / sizeof(expected[0]);
39
40 size_t i = 0;
41 for (auto it = utf8.begin(); it != utf8.end(); ++it) {
42 u32 code_point = *it;
43 VERIFY(i < expected_size);
44 EXPECT_EQ(code_point, expected[i]);
45 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
46 i++;
47 }
48 EXPECT_EQ(i, expected_size);
49}
50
51TEST_CASE(validate_invalid_ut8)
52{
53 size_t valid_bytes;
54 char invalid_utf8_1[] = { 42, 35, (char)182, 9 };
55 Utf8View utf8_1 { StringView { invalid_utf8_1, 4 } };
56 EXPECT(!utf8_1.validate(valid_bytes));
57 EXPECT(valid_bytes == 2);
58
59 char invalid_utf8_2[] = { 42, 35, (char)208, (char)208 };
60 Utf8View utf8_2 { StringView { invalid_utf8_2, 4 } };
61 EXPECT(!utf8_2.validate(valid_bytes));
62 EXPECT(valid_bytes == 2);
63
64 char invalid_utf8_3[] = { (char)208 };
65 Utf8View utf8_3 { StringView { invalid_utf8_3, 1 } };
66 EXPECT(!utf8_3.validate(valid_bytes));
67 EXPECT(valid_bytes == 0);
68
69 char invalid_utf8_4[] = { (char)208, 35 };
70 Utf8View utf8_4 { StringView { invalid_utf8_4, 2 } };
71 EXPECT(!utf8_4.validate(valid_bytes));
72 EXPECT(valid_bytes == 0);
73
74 char invalid_utf8_5[] = { (char)0xf4, (char)0x8f, (char)0xbf, (char)0xc0 }; // U+110000
75 Utf8View utf8_5 { StringView { invalid_utf8_5, 4 } };
76 EXPECT(!utf8_5.validate(valid_bytes));
77 EXPECT(valid_bytes == 0);
78
79 char invalid_utf8_6[] = { (char)0xf4, (char)0xa1, (char)0xb0, (char)0xbd }; // U+121c3d
80 Utf8View utf8_6 { StringView { invalid_utf8_6, 4 } };
81 EXPECT(!utf8_6.validate(valid_bytes));
82 EXPECT(valid_bytes == 0);
83}
84
85TEST_CASE(validate_overlong_utf8)
86{
87 size_t valid_bytes = 0;
88
89 // Overlong 2-byte encoding of U+002F
90 char invalid_utf8_1[] = { 42, 35, static_cast<char>(0xc0), static_cast<char>(0xaf) };
91 Utf8View utf8_1 { StringView { invalid_utf8_1, sizeof(invalid_utf8_1) } };
92 EXPECT(!utf8_1.validate(valid_bytes));
93 EXPECT(valid_bytes == 2);
94
95 // Overlong 3-byte encoding of U+002F
96 char invalid_utf8_2[] = { 42, 35, static_cast<char>(0xe0), static_cast<char>(0x80), static_cast<char>(0xaf) };
97 Utf8View utf8_2 { StringView { invalid_utf8_2, sizeof(invalid_utf8_2) } };
98 EXPECT(!utf8_2.validate(valid_bytes));
99 EXPECT(valid_bytes == 2);
100
101 // Overlong 4-byte encoding of U+002F
102 char invalid_utf8_3[] = { 42, 35, static_cast<char>(0xf0), static_cast<char>(0x80), static_cast<char>(0x80), static_cast<char>(0xaf) };
103 Utf8View utf8_3 { StringView { invalid_utf8_3, sizeof(invalid_utf8_3) } };
104 EXPECT(!utf8_3.validate(valid_bytes));
105 EXPECT(valid_bytes == 2);
106
107 // Overlong 3-byte encoding of U+00FF
108 char invalid_utf8_4[] = { 42, 35, static_cast<char>(0xe0), static_cast<char>(0x83), static_cast<char>(0xbf) };
109 Utf8View utf8_4 { StringView { invalid_utf8_4, sizeof(invalid_utf8_4) } };
110 EXPECT(!utf8_4.validate(valid_bytes));
111 EXPECT(valid_bytes == 2);
112
113 // Overlong 4-byte encoding of U+00FF
114 char invalid_utf8_5[] = { 42, 35, static_cast<char>(0xf0), static_cast<char>(0x80), static_cast<char>(0x83), static_cast<char>(0xbf) };
115 Utf8View utf8_5 { StringView { invalid_utf8_5, sizeof(invalid_utf8_5) } };
116 EXPECT(!utf8_5.validate(valid_bytes));
117 EXPECT(valid_bytes == 2);
118
119 // Overlong 4-byte encoding of U+0FFF
120 char invalid_utf8_6[] = { 42, 35, static_cast<char>(0xf0), static_cast<char>(0x8f), static_cast<char>(0xbf), static_cast<char>(0xbf) };
121 Utf8View utf8_6 { StringView { invalid_utf8_6, sizeof(invalid_utf8_6) } };
122 EXPECT(!utf8_6.validate(valid_bytes));
123 EXPECT(valid_bytes == 2);
124}
125
126TEST_CASE(iterate_utf8)
127{
128 Utf8View view("Some weird characters \u00A9\u266A\uA755"sv);
129 Utf8CodePointIterator iterator = view.begin();
130
131 EXPECT(*iterator == 'S');
132 EXPECT(iterator.peek().has_value() && iterator.peek().value() == 'S');
133 EXPECT(iterator.peek(0).has_value() && iterator.peek(0).value() == 'S');
134 EXPECT(iterator.peek(1).has_value() && iterator.peek(1).value() == 'o');
135 EXPECT(iterator.peek(22).has_value() && iterator.peek(22).value() == 0x00A9);
136 EXPECT(iterator.peek(24).has_value() && iterator.peek(24).value() == 0xA755);
137 EXPECT(!iterator.peek(25).has_value());
138
139 ++iterator;
140
141 EXPECT(*iterator == 'o');
142 EXPECT(iterator.peek(23).has_value() && iterator.peek(23).value() == 0xA755);
143
144 for (size_t i = 0; i < 23; ++i)
145 ++iterator;
146
147 EXPECT(!iterator.done());
148 EXPECT(*iterator == 0xA755);
149 EXPECT(iterator.peek().has_value() && iterator.peek().value() == 0xA755);
150 EXPECT(!iterator.peek(1).has_value());
151
152 ++iterator;
153
154 EXPECT(iterator.done());
155 EXPECT(!iterator.peek(0).has_value());
156 EXPECT_CRASH("Dereferencing Utf8CodePointIterator which is already done.", [&iterator] {
157 *iterator;
158 return Test::Crash::Failure::DidNotCrash;
159 });
160}
161
162TEST_CASE(decode_invalid_ut8)
163{
164 // Test case 1 : Getting an extension byte as first byte of the code point
165 {
166 char raw_data[] = { 'a', 'b', (char)0xA0, 'd' };
167 Utf8View view { StringView { raw_data, 4 } };
168 u32 expected_characters[] = { 'a', 'b', 0xFFFD, 'd' };
169 DeprecatedString expected_underlying_bytes[] = { "a", "b", "\xA0", "d" };
170 size_t expected_size = sizeof(expected_characters) / sizeof(expected_characters[0]);
171 size_t i = 0;
172 for (auto it = view.begin(); it != view.end(); ++it) {
173 u32 code_point = *it;
174 VERIFY(i < expected_size);
175 EXPECT_EQ(code_point, expected_characters[i]);
176 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
177 i++;
178 }
179 VERIFY(i == expected_size);
180 }
181
182 // Test case 2 : Getting a non-extension byte when an extension byte is expected
183 {
184 char raw_data[] = { 'a', 'b', (char)0xC0, 'd', 'e' };
185 Utf8View view { StringView { raw_data, 5 } };
186 u32 expected_characters[] = { 'a', 'b', 0xFFFD, 'd', 'e' };
187 DeprecatedString expected_underlying_bytes[] = { "a", "b", "\xC0", "d", "e" };
188 size_t expected_size = sizeof(expected_characters) / sizeof(expected_characters[0]);
189 size_t i = 0;
190 for (auto it = view.begin(); it != view.end(); ++it) {
191 u32 code_point = *it;
192 VERIFY(i < expected_size);
193 EXPECT_EQ(code_point, expected_characters[i]);
194 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
195 i++;
196 }
197 VERIFY(i == expected_size);
198 }
199
200 // Test case 3 : Not enough bytes before the end of the string
201 {
202 char raw_data[] = { 'a', 'b', (char)0x90, 'd' };
203 Utf8View view { StringView { raw_data, 4 } };
204 u32 expected_characters[] = { 'a', 'b', 0xFFFD, 'd' };
205 DeprecatedString expected_underlying_bytes[] = { "a", "b", "\x90", "d" };
206 size_t expected_size = sizeof(expected_characters) / sizeof(expected_characters[0]);
207 size_t i = 0;
208 for (auto it = view.begin(); it != view.end(); ++it) {
209 u32 code_point = *it;
210 VERIFY(i < expected_size);
211 EXPECT_EQ(code_point, expected_characters[i]);
212 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
213 i++;
214 }
215 VERIFY(i == expected_size);
216 }
217
218 // Test case 4 : Not enough bytes at the end of the string
219 {
220 char raw_data[] = { 'a', 'b', 'c', (char)0x90 };
221 Utf8View view { StringView { raw_data, 4 } };
222 u32 expected_characters[] = { 'a', 'b', 'c', 0xFFFD };
223 DeprecatedString expected_underlying_bytes[] = { "a", "b", "c", "\x90" };
224 size_t expected_size = sizeof(expected_characters) / sizeof(expected_characters[0]);
225 size_t i = 0;
226 for (auto it = view.begin(); it != view.end(); ++it) {
227 u32 code_point = *it;
228 VERIFY(i < expected_size);
229 EXPECT_EQ(code_point, expected_characters[i]);
230 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
231 i++;
232 }
233 VERIFY(i == expected_size);
234 }
235
236 // Test case 5 : Oversized four-byte sequence (e.g. U+123456)
237 {
238 // Want to encode: (000)1 0010 0011 0100 0101 0110
239 // Into mask: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
240 // Shifted: 100 100011 010001 010110
241 // Result: 11110100 10100011 10010001 10010110
242 char raw_data[] = { 'a', (char)0xF4, (char)0xA3, (char)0x91, (char)0x96, 'b' };
243 Utf8View view { StringView { raw_data, 6 } };
244 // This definition seems to suggest that we should instead output multiple replacement characters:
245 // https://encoding.spec.whatwg.org/#ref-for-concept-stream-prepend②
246 // This is supported by the plaintext description and example collection, which annoyingly does not give an example of how to deal with this:
247 // https://www.unicode.org/versions/Unicode14.0.0/ch03.pdf , section "U+FFFD Substitution of Maximal Subparts"
248 // However, that would go against how we deal with several other kinds of errors, so we stick to emitting only one U+FFFD.
249 u32 expected_characters[] = { 'a', 0xFFFD, 'b' };
250 DeprecatedString expected_underlying_bytes[] = { "a", "\xF4\xA3\x91\x96", "b" };
251 size_t expected_size = sizeof(expected_characters) / sizeof(expected_characters[0]);
252 size_t i = 0;
253 for (auto it = view.begin(); it != view.end(); ++it) {
254 u32 code_point = *it;
255 VERIFY(i < expected_size);
256 EXPECT_EQ(code_point, expected_characters[i]);
257 EXPECT_EQ(it.underlying_code_point_bytes(), expected_underlying_bytes[i].bytes());
258 i++;
259 }
260 VERIFY(i == expected_size);
261 }
262}
263
264TEST_CASE(trim)
265{
266 Utf8View whitespace { " "sv };
267 {
268 Utf8View view { "word"sv };
269 EXPECT_EQ(view.trim(whitespace, TrimMode::Both).as_string(), "word");
270 EXPECT_EQ(view.trim(whitespace, TrimMode::Left).as_string(), "word");
271 EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), "word");
272 }
273 {
274 Utf8View view { " word"sv };
275 EXPECT_EQ(view.trim(whitespace, TrimMode::Both).as_string(), "word");
276 EXPECT_EQ(view.trim(whitespace, TrimMode::Left).as_string(), "word");
277 EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), " word");
278 }
279 {
280 Utf8View view { "word "sv };
281 EXPECT_EQ(view.trim(whitespace, TrimMode::Both).as_string(), "word");
282 EXPECT_EQ(view.trim(whitespace, TrimMode::Left).as_string(), "word ");
283 EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), "word");
284 }
285 {
286 Utf8View view { " word "sv };
287 EXPECT_EQ(view.trim(whitespace, TrimMode::Both).as_string(), "word");
288 EXPECT_EQ(view.trim(whitespace, TrimMode::Left).as_string(), "word ");
289 EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), " word");
290 }
291 {
292 Utf8View view { "\u180E"sv };
293 EXPECT_EQ(view.trim(whitespace, TrimMode::Both).as_string(), "\u180E");
294 EXPECT_EQ(view.trim(whitespace, TrimMode::Left).as_string(), "\u180E");
295 EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), "\u180E");
296 }
297}