OLD | NEW |
1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. | 1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. |
2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
4 | 4 |
5 #include "content/renderer/android/email_detector.h" | 5 #include "content/renderer/android/email_detector.h" |
6 | 6 |
7 #include "base/logging.h" | 7 #include "base/logging.h" |
8 #include "base/memory/scoped_ptr.h" | 8 #include "base/memory/scoped_ptr.h" |
9 #include "base/strings/utf_string_conversions.h" | 9 #include "base/strings/utf_string_conversions.h" |
10 #include "content/public/renderer/android_content_detection_prefixes.h" | 10 #include "content/public/renderer/android_content_detection_prefixes.h" |
11 #include "net/base/escape.h" | 11 #include "net/base/escape.h" |
12 #include "third_party/icu/source/i18n/unicode/regex.h" | 12 #include "third_party/icu/source/i18n/unicode/regex.h" |
13 | 13 |
14 namespace { | 14 namespace { |
15 | 15 |
16 // Maximum length of an email address. | 16 // Maximum length of an email address. |
17 const size_t kMaximumEmailLength = 254; | 17 const size_t kMaximumEmailLength = 254; |
18 | 18 |
19 // Regex to match email addresses. | 19 // Regex to match email addresses. |
20 // This is more specific than RFC 2822 (uncommon special characters are | 20 // This is more specific than RFC 2822 (uncommon special characters are |
21 // disallowed) in order to avoid false positives. | 21 // disallowed) in order to avoid false positives. |
22 // Delimiters are word boundaries to allow punctuation, quote marks etc. around | 22 // Delimiters are word boundaries to allow punctuation, quote marks etc. around |
23 // the address. | 23 // the address. |
24 const char kEmailRegex[] = "\\b[A-Z0-9._%+-]+@[A-Z0-9.-]+\\.[A-Z]{2,6}\\b"; | 24 const char kEmailRegex[] = |
| 25 "\\b[A-Z0-9._%+-]+@[A-Z0-9-]+(\\.[A-Z0-9-]+)*(\\.[A-Z]{2,6})\\b"; |
25 | 26 |
26 } // anonymous namespace | 27 } // anonymous namespace |
27 | 28 |
28 namespace content { | 29 namespace content { |
29 | 30 |
30 EmailDetector::EmailDetector() { | 31 EmailDetector::EmailDetector() { |
31 } | 32 } |
32 | 33 |
33 size_t EmailDetector::GetMaximumContentLength() { | 34 size_t EmailDetector::GetMaximumContentLength() { |
34 return kMaximumEmailLength; | 35 return kMaximumEmailLength; |
(...skipping 30 matching lines...) Expand all Loading... |
65 DCHECK(U_SUCCESS(status)); | 66 DCHECK(U_SUCCESS(status)); |
66 base::UTF16ToUTF8(content_ustr.getBuffer(), content_ustr.length(), | 67 base::UTF16ToUTF8(content_ustr.getBuffer(), content_ustr.length(), |
67 content_text); | 68 content_text); |
68 return true; | 69 return true; |
69 } | 70 } |
70 | 71 |
71 return false; | 72 return false; |
72 } | 73 } |
73 | 74 |
74 } // namespace content | 75 } // namespace content |
OLD | NEW |