OLD | NEW |
| (Empty) |
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 "third_party/libphonenumber/cpp/src/regexp_adapter.h" | |
6 | |
7 // Setup all of the Chromium and WebKit defines | |
8 #include "base/logging.h" | |
9 #include "base/scoped_ptr.h" | |
10 #include "build/build_config.h" | |
11 #include "unicode/regex.h" | |
12 #include "unicode/stringpiece.h" | |
13 #include "unicode/unistr.h" | |
14 | |
15 namespace { | |
16 | |
17 // Converts |source| to UTF-8 string, returns it starting at position |pos|. | |
18 std::string UnicodeStringToUtf8String(icu::UnicodeString const& source, | |
19 int pos) { | |
20 std::string data; | |
21 source.toUTF8String<std::string>(data); | |
22 return data.substr(pos); | |
23 } | |
24 | |
25 } // namespace | |
26 | |
27 // Implementation of the abstract classes RegularExpressionInput and | |
28 // RegularExpression using ICU regular expression capabilities. | |
29 | |
30 // The Regular Expression input class. | |
31 class IcuRegularExpressionInput : public reg_exp::RegularExpressionInput { | |
32 public: | |
33 explicit IcuRegularExpressionInput(const char* utf8_input); | |
34 | |
35 // RegularExpressionInput implementation: | |
36 // Matches string to regular expression, returns true if expression was | |
37 // matched, false otherwise, advances position in the match. | |
38 // |reg_exp| - expression to be matched. | |
39 // |beginning_only| - if true match would be successfull only if appears at | |
40 // the beginning of the tested region of the string. | |
41 // |matched_string1| - successfully matched first string. Can be NULL. | |
42 // |matched_string2| - successfully matched second string. Can be NULL. | |
43 virtual bool ConsumeRegExp(std::string const& reg_exp, | |
44 bool beginning_only, | |
45 std::string* matched_string1, | |
46 std::string* matched_string2); | |
47 | |
48 // Convert unmatched input to a string. | |
49 virtual std::string ToString() const; | |
50 | |
51 icu::UnicodeString* Data() { return &utf8_input_; } | |
52 | |
53 // Position in the input. For the newly created input position is 0, | |
54 // each call to ConsumeRegExp() or RegularExpression::Consume() advances | |
55 // position in the case of the successful match to be after the match. | |
56 int pos() const { return pos_; } | |
57 void set_pos(int pos) { pos_ = pos; } | |
58 | |
59 private: | |
60 icu::UnicodeString utf8_input_; | |
61 int pos_; | |
62 | |
63 DISALLOW_COPY_AND_ASSIGN(IcuRegularExpressionInput); | |
64 }; | |
65 | |
66 // The regular expression class. | |
67 class IcuRegularExpression : public reg_exp::RegularExpression { | |
68 public: | |
69 explicit IcuRegularExpression(const char* utf8_regexp); | |
70 | |
71 // RegularExpression implementation: | |
72 // Matches string to regular expression, returns true if expression was | |
73 // matched, false otherwise, advances position in the match. | |
74 // |input_string| - string to be searched. | |
75 // |beginning_only| - if true match would be successfull only if appears at | |
76 // the beginning of the tested region of the string. | |
77 // |matched_string1| - successfully matched first string. Can be NULL. | |
78 // |matched_string2| - successfully matched second string. Can be NULL. | |
79 // |matched_string3| - successfully matched third string. Can be NULL. | |
80 virtual bool Consume(reg_exp::RegularExpressionInput* input_string, | |
81 bool beginning_only, | |
82 std::string* matched_string1, | |
83 std::string* matched_string2, | |
84 std::string* matched_string3) const; | |
85 | |
86 // Matches string to regular expression, returns true if expression was | |
87 // matched, false otherwise. | |
88 // |input_string| - string to be searched. | |
89 // |full_match| - if true match would be successfull only if it matches the | |
90 // complete string. | |
91 // |matched_string| - successfully matched string. Can be NULL. | |
92 virtual bool Match(const char* input_string, | |
93 bool full_match, | |
94 std::string* matched_string) const; | |
95 | |
96 // Replaces match(es) in the |string_to_process|. if |global| is true, | |
97 // replaces all the matches, only the first match otherwise. | |
98 // |replacement_string| - text the matches are replaced with. | |
99 // Returns true if expression successfully processed through the string, | |
100 // even if no actual replacements were made. Returns false in case of an | |
101 // error. | |
102 virtual bool Replace(std::string* string_to_process, | |
103 bool global, | |
104 const char* replacement_string) const; | |
105 private: | |
106 scoped_ptr<icu::RegexPattern> utf8_regexp_; | |
107 | |
108 DISALLOW_COPY_AND_ASSIGN(IcuRegularExpression); | |
109 }; | |
110 | |
111 IcuRegularExpressionInput::IcuRegularExpressionInput(const char* utf8_input) | |
112 : pos_(0) { | |
113 DCHECK(utf8_input); | |
114 utf8_input_ = icu::UnicodeString::fromUTF8(utf8_input); | |
115 } | |
116 | |
117 bool IcuRegularExpressionInput::ConsumeRegExp(std::string const& reg_exp, | |
118 bool beginning_only, | |
119 std::string* matched_string1, | |
120 std::string* matched_string2) { | |
121 IcuRegularExpression re(reg_exp.c_str()); | |
122 | |
123 return re.Consume(this, beginning_only, matched_string1, matched_string2, | |
124 NULL); | |
125 } | |
126 | |
127 std::string IcuRegularExpressionInput::ToString() const { | |
128 if (pos_ < 0 || pos_ > utf8_input_.length()) | |
129 return std::string(); | |
130 return UnicodeStringToUtf8String(utf8_input_, pos_); | |
131 } | |
132 | |
133 IcuRegularExpression::IcuRegularExpression(const char* utf8_regexp) { | |
134 DCHECK(utf8_regexp); | |
135 UParseError pe; | |
136 UErrorCode status = U_ZERO_ERROR; | |
137 utf8_regexp_.reset(icu::RegexPattern::compile( | |
138 icu::UnicodeString::fromUTF8(utf8_regexp), 0, pe, status)); | |
139 if (U_FAILURE(status)) { | |
140 // All of the passed regular expressions should compile correctly. | |
141 utf8_regexp_.reset(NULL); | |
142 NOTREACHED(); | |
143 } | |
144 } | |
145 | |
146 bool IcuRegularExpression::Consume( | |
147 reg_exp::RegularExpressionInput* input_string, | |
148 bool beginning_only, | |
149 std::string* matched_string1, | |
150 std::string* matched_string2, | |
151 std::string* matched_string3) const { | |
152 DCHECK(input_string); | |
153 // matched_string1 may be NULL | |
154 // matched_string2 may be NULL | |
155 // matched_string3 may be NULL | |
156 if (!utf8_regexp_.get()) | |
157 return false; | |
158 | |
159 IcuRegularExpressionInput* input = | |
160 reinterpret_cast<IcuRegularExpressionInput *>(input_string); | |
161 UErrorCode status = U_ZERO_ERROR; | |
162 scoped_ptr<icu::RegexMatcher> matcher(utf8_regexp_->matcher(*(input->Data()), | |
163 status)); | |
164 | |
165 if (U_FAILURE(status)) | |
166 return false; | |
167 | |
168 if (beginning_only) { | |
169 if (!matcher->lookingAt(input->pos(), status)) | |
170 return false; | |
171 } else { | |
172 if (!matcher->find(input->pos(), status)) | |
173 return false; | |
174 } | |
175 if (U_FAILURE(status)) | |
176 return false; | |
177 // If less matches than expected - fail. | |
178 if ((matched_string3 && matcher->groupCount() < 3) || | |
179 (matched_string2 && matcher->groupCount() < 2) || | |
180 (matched_string1 && matcher->groupCount() < 1)) { | |
181 return false; | |
182 } | |
183 if (matcher->groupCount() > 0 && matched_string1) { | |
184 *matched_string1 = UnicodeStringToUtf8String(matcher->group(1, status), 0); | |
185 } | |
186 if (matcher->groupCount() > 1 && matched_string2) { | |
187 *matched_string2 = UnicodeStringToUtf8String(matcher->group(2, status), 0); | |
188 } | |
189 if (matcher->groupCount() > 2 && matched_string3) { | |
190 *matched_string3 = UnicodeStringToUtf8String(matcher->group(3, status), 0); | |
191 } | |
192 input->set_pos(matcher->end(status)); | |
193 return true; | |
194 } | |
195 | |
196 bool IcuRegularExpression::Match(const char* input_string, | |
197 bool full_match, | |
198 std::string* matched_string) const { | |
199 DCHECK(input_string); | |
200 // matched_string may be NULL | |
201 if (!utf8_regexp_.get()) | |
202 return false; | |
203 | |
204 IcuRegularExpressionInput input(input_string); | |
205 UErrorCode status = U_ZERO_ERROR; | |
206 scoped_ptr<icu::RegexMatcher> matcher(utf8_regexp_->matcher(*(input.Data()), | |
207 status)); | |
208 | |
209 if (U_FAILURE(status)) | |
210 return false; | |
211 | |
212 if (full_match) { | |
213 if (!matcher->matches(input.pos(), status)) | |
214 return false; | |
215 } else { | |
216 if (!matcher->find(input.pos(), status)) | |
217 return false; | |
218 } | |
219 if (U_FAILURE(status)) | |
220 return false; | |
221 if (matcher->groupCount() > 0 && matched_string) { | |
222 *matched_string = UnicodeStringToUtf8String(matcher->group(1, status), 0); | |
223 } | |
224 return true; | |
225 } | |
226 | |
227 bool IcuRegularExpression::Replace(std::string* string_to_process, | |
228 bool global, | |
229 const char* replacement_string) const { | |
230 DCHECK(string_to_process); | |
231 DCHECK(replacement_string); | |
232 | |
233 std::string adapted_replacement(replacement_string); | |
234 // Adapt replacement string from RE2 (\0-9 for matches) format to ICU format | |
235 // ($0-9 for matches). All '$' should be prepended with '\' as well. | |
236 size_t backslash_pos = adapted_replacement.find('\\'); | |
237 size_t dollar_pos = adapted_replacement.find('$'); | |
238 while (backslash_pos != std::string::npos || | |
239 dollar_pos != std::string::npos) { | |
240 bool process_dollar = false; | |
241 if (backslash_pos == std::string::npos || | |
242 (dollar_pos != std::string::npos && dollar_pos < backslash_pos)) { | |
243 process_dollar = true; | |
244 } | |
245 if (process_dollar) { | |
246 adapted_replacement.insert(dollar_pos, "\\"); | |
247 dollar_pos = adapted_replacement.find('$', dollar_pos + 2); | |
248 if (backslash_pos != std::string::npos) | |
249 ++backslash_pos; | |
250 } else { | |
251 if (adapted_replacement.length() > backslash_pos + 1) { | |
252 if (adapted_replacement[backslash_pos + 1] >= '0' && | |
253 adapted_replacement[backslash_pos + 1] <= '9') { | |
254 adapted_replacement[backslash_pos] = '$'; | |
255 } | |
256 if (adapted_replacement[backslash_pos + 1] == '\\') { | |
257 // Skip two characters instead of one. | |
258 ++backslash_pos; | |
259 } | |
260 } | |
261 backslash_pos = adapted_replacement.find('\\', backslash_pos + 1); | |
262 } | |
263 } | |
264 | |
265 IcuRegularExpressionInput input(string_to_process->c_str()); | |
266 UErrorCode status = U_ZERO_ERROR; | |
267 scoped_ptr<icu::RegexMatcher> matcher(utf8_regexp_->matcher(*(input.Data()), | |
268 status)); | |
269 if (U_FAILURE(status)) | |
270 return false; | |
271 | |
272 icu::UnicodeString result; | |
273 | |
274 if (global) { | |
275 result = matcher->replaceAll( | |
276 icu::UnicodeString::fromUTF8(adapted_replacement), | |
277 status); | |
278 } else { | |
279 result = matcher->replaceFirst( | |
280 icu::UnicodeString::fromUTF8(adapted_replacement), | |
281 status); | |
282 } | |
283 if (U_FAILURE(status)) | |
284 return false; | |
285 *string_to_process = UnicodeStringToUtf8String(result, 0); | |
286 return true; | |
287 } | |
288 | |
289 namespace reg_exp { | |
290 | |
291 RegularExpressionInput* CreateRegularExpressionInput(const char* utf8_input) { | |
292 return new IcuRegularExpressionInput(utf8_input); | |
293 } | |
294 | |
295 RegularExpression* CreateRegularExpression(const char* utf8_regexp) { | |
296 return new IcuRegularExpression(utf8_regexp); | |
297 } | |
298 | |
299 } // namespace reg_exp | |
OLD | NEW |