blob: 1a0db8b07a850ec383b75e0be6dc209e16e43fbf [file] [log] [blame]
// 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 <memory>
#include "base/logging.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[] =
} // anonymous namespace
namespace content {
EmailDetector::EmailDetector() {
size_t EmailDetector::GetMaximumContentLength() {
return kMaximumEmailLength;
GURL EmailDetector::GetIntentURL(const std::string& content_text) {
if (content_text.empty())
return GURL();
return GURL(kEmailPrefix +
net::EscapeQueryParamValue(content_text, true));
bool EmailDetector::FindContent(const base::string16::const_iterator& begin,
const base::string16::const_iterator& end,
size_t* start_pos,
size_t* end_pos,
std::string* content_text) {
base::string16 utf16_input = base::string16(begin, end);
icu::UnicodeString pattern(kEmailRegex);
icu::UnicodeString input(, utf16_input.length());
UErrorCode status = U_ZERO_ERROR;
std::unique_ptr<icu::RegexMatcher> matcher(
new icu::RegexMatcher(pattern, input, UREGEX_CASE_INSENSITIVE, status));
if (matcher->find()) {
*start_pos = matcher->start(status);
*end_pos = matcher->end(status);
icu::UnicodeString content_ustr(matcher->group(status));
return true;
return false;
} // namespace content