Email Regular Expression update for exclusion of invalid domain parts

content/renderer/android/email_detector.cc

 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "content/renderer/android/email_detector.h"
 
 #include "base/logging.h"
 #include "base/memory/scoped_ptr.h"
 #include "base/strings/utf_string_conversions.h"
 #include "content/public/renderer/android_content_detection_prefixes.h"
 #include "net/base/escape.h"
 #include "third_party/icu/source/i18n/unicode/regex.h"
 
 namespace {
 
 // Maximum length of an email address.
 const size_t kMaximumEmailLength = 254;
 
 // Regex to match email addresses.
 // This is more specific than RFC 2822 (uncommon special characters are
 // disallowed) in order to avoid false positives.
 // Delimiters are word boundaries to allow punctuation, quote marks etc. around
 // the address.
-const char kEmailRegex[] = "\\b[A-Z0-9._%+-]+@[A-Z0-9.-]+\\.[A-Z]{2,6}\\b";
+const char kEmailRegex[] =
+    "\\b[A-Z0-9._%+-]+@[A-Z0-9-]+(\\.[A-Z0-9-]+)*(\\.[A-Z]{2,6})\\b";
 
 }  // anonymous namespace
 
 namespace content {
 
 EmailDetector::EmailDetector() {
 }
 
 size_t EmailDetector::GetMaximumContentLength() {
   return kMaximumEmailLength;
@@ skipping 30 matching lines @@
     DCHECK(U_SUCCESS(status));
     base::UTF16ToUTF8(content_ustr.getBuffer(), content_ustr.length(),
         content_text);
     return true;
   }
 
   return false;
 }
 
 }  // namespace content