OLD | NEW |
| (Empty) |
1 // Copyright 2015 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 <stddef.h> | |
6 | |
7 #include <algorithm> | |
8 #include <memory> | |
9 #include <utility> | |
10 | |
11 #include "base/macros.h" | |
12 #include "base/path_service.h" | |
13 #include "base/strings/string_split.h" | |
14 #include "base/strings/string_util.h" | |
15 #include "base/strings/utf_string_conversions.h" | |
16 #include "chrome/renderer/spellchecker/spellcheck.h" | |
17 #include "chrome/renderer/spellchecker/spellcheck_provider_test.h" | |
18 #include "components/spellcheck/common/spellcheck_common.h" | |
19 #include "components/spellcheck/common/spellcheck_result.h" | |
20 #include "testing/gtest/include/gtest/gtest.h" | |
21 #include "third_party/WebKit/public/platform/WebString.h" | |
22 #include "third_party/WebKit/public/platform/WebVector.h" | |
23 #include "third_party/WebKit/public/web/WebTextCheckingResult.h" | |
24 | |
25 namespace { | |
26 | |
27 struct SpellcheckTestCase { | |
28 // A string of text for checking. | |
29 const wchar_t* input; | |
30 // The position and the length of the first misspelled word, if any. | |
31 int expected_misspelling_start; | |
32 int expected_misspelling_length; | |
33 }; | |
34 | |
35 base::FilePath GetHunspellDirectory() { | |
36 base::FilePath hunspell_directory; | |
37 if (!PathService::Get(base::DIR_SOURCE_ROOT, &hunspell_directory)) | |
38 return base::FilePath(); | |
39 | |
40 hunspell_directory = hunspell_directory.AppendASCII("third_party"); | |
41 hunspell_directory = hunspell_directory.AppendASCII("hunspell_dictionaries"); | |
42 return hunspell_directory; | |
43 } | |
44 | |
45 } // namespace | |
46 | |
47 class MultilingualSpellCheckTest : public testing::Test { | |
48 public: | |
49 MultilingualSpellCheckTest() {} | |
50 | |
51 void ReinitializeSpellCheck(const std::string& unsplit_languages) { | |
52 spellcheck_ = new SpellCheck(); | |
53 provider_.reset(new TestingSpellCheckProvider(spellcheck_)); | |
54 InitializeSpellCheck(unsplit_languages); | |
55 } | |
56 | |
57 void InitializeSpellCheck(const std::string& unsplit_languages) { | |
58 base::FilePath hunspell_directory = GetHunspellDirectory(); | |
59 EXPECT_FALSE(hunspell_directory.empty()); | |
60 std::vector<std::string> languages = base::SplitString( | |
61 unsplit_languages, ",", base::TRIM_WHITESPACE, base::SPLIT_WANT_ALL); | |
62 | |
63 for (const auto& language : languages) { | |
64 base::File file( | |
65 spellcheck::GetVersionedFileName(language, hunspell_directory), | |
66 base::File::FLAG_OPEN | base::File::FLAG_READ); | |
67 spellcheck_->AddSpellcheckLanguage(std::move(file), language); | |
68 } | |
69 } | |
70 | |
71 ~MultilingualSpellCheckTest() override {} | |
72 TestingSpellCheckProvider* provider() { return provider_.get(); } | |
73 | |
74 protected: | |
75 void ExpectSpellCheckWordResults(const std::string& languages, | |
76 const SpellcheckTestCase* test_cases, | |
77 size_t num_test_cases) { | |
78 ReinitializeSpellCheck(languages); | |
79 | |
80 for (size_t i = 0; i < num_test_cases; ++i) { | |
81 int misspelling_start = 0; | |
82 int misspelling_length = 0; | |
83 static_cast<blink::WebSpellCheckClient*>(provider()) | |
84 ->spellCheck(blink::WebString(base::WideToUTF16(test_cases[i].input)), | |
85 misspelling_start, misspelling_length, nullptr); | |
86 | |
87 EXPECT_EQ(test_cases[i].expected_misspelling_start, misspelling_start) | |
88 << "Improper misspelling location found with the languages " | |
89 << languages << " when checking \"" << test_cases[i].input << "\"."; | |
90 EXPECT_EQ(test_cases[i].expected_misspelling_length, misspelling_length) | |
91 << "Improper misspelling length found with the languages " | |
92 << languages << " when checking \"" << test_cases[i].input << "\"."; | |
93 } | |
94 } | |
95 | |
96 void ExpectSpellCheckParagraphResults( | |
97 const base::string16& input, | |
98 const std::vector<SpellCheckResult>& expected) { | |
99 blink::WebVector<blink::WebTextCheckingResult> results; | |
100 spellcheck_->SpellCheckParagraph(blink::WebString(input), &results); | |
101 | |
102 EXPECT_EQ(expected.size(), results.size()); | |
103 size_t size = std::min(results.size(), expected.size()); | |
104 for (size_t i = 0; i < size; ++i) { | |
105 EXPECT_EQ(blink::WebTextDecorationTypeSpelling, results[i].decoration); | |
106 EXPECT_EQ(expected[i].location, results[i].location); | |
107 EXPECT_EQ(expected[i].length, results[i].length); | |
108 } | |
109 } | |
110 | |
111 private: | |
112 // Owned by |provider_|. | |
113 SpellCheck* spellcheck_; | |
114 std::unique_ptr<TestingSpellCheckProvider> provider_; | |
115 }; | |
116 | |
117 // Check that a string of different words is properly spellchecked for different | |
118 // combinations of different languages. | |
119 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckWord) { | |
120 static const SpellcheckTestCase kTestCases[] = { | |
121 // An English, Spanish, Russian, and Greek word, all spelled correctly. | |
122 {L"rocket destruyan \x0432\x0441\x0435\x0445 \x03C4\x03B9\x03C2", 0, 0}, | |
123 // A misspelled English word. | |
124 {L"rocktt destruyan \x0432\x0441\x0435\x0445 \x03C4\x03B9\x03C2", 0, 6}, | |
125 // A misspelled Spanish word. | |
126 {L"rocket destruynn \x0432\x0441\x0435\x0445 \x03C4\x03B9\x03C2", 7, 9}, | |
127 // A misspelled Russian word. | |
128 {L"rocket destruyan \x0430\x0430\x0430\x0430 \x03C4\x03B9\x03C2", 17, 4}, | |
129 // A misspelled Greek word. | |
130 {L"rocket destruyan \x0432\x0441\x0435\x0445 \x03B1\x03B1\x03B1\x03B1", | |
131 22, 4}, | |
132 // An English word, then Russian, and then a misspelled English word. | |
133 {L"rocket \x0432\x0441\x0435\x0445 rocktt", 12, 6}, | |
134 }; | |
135 | |
136 // A sorted list of languages. This must start sorted to get all possible | |
137 // permutations. | |
138 std::string languages = "el-GR,en-US,es-ES,ru-RU"; | |
139 std::vector<std::string> permuted_languages = base::SplitString( | |
140 languages, ",", base::TRIM_WHITESPACE, base::SPLIT_WANT_ALL); | |
141 | |
142 do { | |
143 languages = base::JoinString(permuted_languages, ","); | |
144 ExpectSpellCheckWordResults(languages, kTestCases, arraysize(kTestCases)); | |
145 } while (std::next_permutation(permuted_languages.begin(), | |
146 permuted_languages.end())); | |
147 } | |
148 | |
149 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckWordEnglishSpanish) { | |
150 static const SpellcheckTestCase kTestCases[] = { | |
151 {L"", 0, 0}, | |
152 {L"head hand foot legs arms", 0, 0}, | |
153 {L"head hand foot legs arms zzzz", 25, 4}, | |
154 {L"head hand zzzz foot legs arms", 10, 4}, | |
155 {L"zzzz head hand foot legs arms", 0, 4}, | |
156 {L"zzzz head zzzz foot zzzz arms", 0, 4}, | |
157 {L"head hand foot arms zzzz zzzz", 20, 4}, | |
158 {L"I do not want a monstrous snake near me.", 0, 0}, | |
159 {L"zz do not want a monstrous snake near me.", 0, 2}, | |
160 {L"I do not want zz monstrous snake near me.", 14, 2}, | |
161 {L"I do not want a monstrous zz near me.", 26, 2}, | |
162 {L"I do not want a monstrou snake near me.", 16, 8}, | |
163 {L"I do not want a monstrous snake near zz.", 37, 2}, | |
164 {L"Partially Spanish is very bueno.", 0, 0}, | |
165 {L"Sleeping in the biblioteca is good.", 0, 0}, | |
166 {L"Hermano is my favorite name.", 0, 0}, | |
167 {L"hola hola hola hola hola hola", 0, 0}, | |
168 {L"sand hola hola hola hola hola", 0, 0}, | |
169 {L"hola sand sand sand sand sand", 0, 0}, | |
170 {L"sand sand sand sand sand hola", 0, 0}, | |
171 {L"sand hola sand hola sand hola", 0, 0}, | |
172 {L"hola sand hola sand hola sand", 0, 0}, | |
173 {L"hola:legs", 0, 9}, | |
174 {L"legs:hola", 0, 9}}; | |
175 ExpectSpellCheckWordResults("en-US,es-ES", kTestCases, arraysize(kTestCases)); | |
176 } | |
177 | |
178 // If there are no spellcheck languages, no text should be marked as misspelled. | |
179 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckParagraphBlank) { | |
180 ReinitializeSpellCheck(std::string()); | |
181 | |
182 ExpectSpellCheckParagraphResults( | |
183 // English, German, Spanish, and a misspelled word. | |
184 base::UTF8ToUTF16("rocket Schwarzkommando destruyan pcnyhon"), | |
185 std::vector<SpellCheckResult>()); | |
186 } | |
187 | |
188 // Make sure nothing is considered misspelled when at least one of the selected | |
189 // languages determines that a word is correctly spelled. | |
190 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckParagraphCorrect) { | |
191 ReinitializeSpellCheck("en-US,es-ES,de-DE"); | |
192 | |
193 ExpectSpellCheckParagraphResults( | |
194 // English, German, and Spanish words, all spelled correctly. | |
195 base::UTF8ToUTF16("rocket Schwarzkommando destruyan"), | |
196 std::vector<SpellCheckResult>()); | |
197 } | |
198 | |
199 // Make sure that all the misspellings in the text are found. | |
200 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckParagraph) { | |
201 ReinitializeSpellCheck("en-US,es-ES"); | |
202 std::vector<SpellCheckResult> expected; | |
203 expected.push_back(SpellCheckResult(SpellCheckResult::SPELLING, 7, 15)); | |
204 expected.push_back(SpellCheckResult(SpellCheckResult::SPELLING, 33, 7)); | |
205 | |
206 ExpectSpellCheckParagraphResults( | |
207 // English, German, Spanish, and a misspelled word. | |
208 base::UTF8ToUTF16("rocket Schwarzkommando destruyan pcnyhon"), expected); | |
209 } | |
210 | |
211 // Ensure that suggestions are handled properly for multiple languages. | |
212 TEST_F(MultilingualSpellCheckTest, MultilingualSpellCheckSuggestions) { | |
213 ReinitializeSpellCheck("en-US,es-ES"); | |
214 static const struct { | |
215 // A string of text for checking. | |
216 const wchar_t* input; | |
217 // The position and the length of the first invalid word. | |
218 int expected_misspelling_start; | |
219 int expected_misspelling_length; | |
220 // A comma separated string of suggested words that should occur, in their | |
221 // expected order. | |
222 const wchar_t* expected_suggestions; | |
223 } kTestCases[] = { | |
224 {L"rocket", 0, 0}, | |
225 {L"destruyan", 0, 0}, | |
226 {L"rocet", 0, 5, L"rocket,roce,crochet,troce,rocen"}, | |
227 {L"jum", 0, 3, L"hum,jun,ju,um,juma"}, | |
228 {L"asdne", 0, 5, L"sadness,desasne"}, | |
229 }; | |
230 | |
231 for (size_t i = 0; i < arraysize(kTestCases); ++i) { | |
232 blink::WebVector<blink::WebString> suggestions; | |
233 int misspelling_start; | |
234 int misspelling_length; | |
235 static_cast<blink::WebSpellCheckClient*>(provider()) | |
236 ->spellCheck(blink::WebString(base::WideToUTF16(kTestCases[i].input)), | |
237 misspelling_start, misspelling_length, &suggestions); | |
238 | |
239 EXPECT_EQ(kTestCases[i].expected_misspelling_start, misspelling_start); | |
240 EXPECT_EQ(kTestCases[i].expected_misspelling_length, misspelling_length); | |
241 if (!kTestCases[i].expected_suggestions) { | |
242 EXPECT_EQ(0UL, suggestions.size()); | |
243 continue; | |
244 } | |
245 | |
246 std::vector<base::string16> expected_suggestions = base::SplitString( | |
247 base::WideToUTF16(kTestCases[i].expected_suggestions), | |
248 base::string16(1, ','), base::TRIM_WHITESPACE, base::SPLIT_WANT_ALL); | |
249 | |
250 EXPECT_EQ(expected_suggestions.size(), suggestions.size()); | |
251 for (size_t j = 0; | |
252 j < std::min(expected_suggestions.size(), suggestions.size()); j++) { | |
253 EXPECT_EQ(expected_suggestions[j], base::string16(suggestions[j])); | |
254 } | |
255 } | |
256 } | |
OLD | NEW |