private void ExtractCitys(OnCompletedEventArgs e)
{
var matches = Regex.Matches(e.pageSoure, @"<a[^>]+href=""*(?<href>/hotel/[^>\s]+)""\s*[^>]*>(?<text>(?!.*img).*?)</a>");
cityList.Clear();
foreach (Match m in matches)
{
var city = new City
{
CityName = m.Groups["text"].Value,
Uri = new Uri("http://hotels.ctrip.com" + m.Groups["href"].Value)
};
if (!cityList.Contains(city))
{
cityList.Add(city);
}
}
Console.WriteLine("城市抓取完成！");
Console.WriteLine("耗时：" + e.milliseconds + "毫秒");
Console.WriteLine("线程:" + e.ThreadId);
Console.WriteLine("地址:" + e.uri.ToString());
this.Invoke(new Action(() => {
this.cityListBox.DataSource = cityList;
cityListBox.SelectedIndex = 0;
cityListBox_DoubleClick(this, null);
}));
}
void hotelListCrawler_OnCompleted(object sender, OnCompletedEventArgs e)
{
Console.WriteLine("酒店搜索完成");
var matches = Regex.Matches(e.pageSoure, "><a[^>]+href=\"*(?<href>/hotel/[^>\\s]+)\"\\s*data-dopost[^>]*><span[^>]+>.*?</span>(?<text>.*?)</a>");
hotelList.Clear();
foreach (Match m in matches)
{
var hotel = new Hotel
{
HotelName = m.Groups["text"].Value,
Uri = new Uri("http://hotels.ctrip.com" + m.Groups["href"].Value)
};
if (!hotelList.Contains(hotel))
{
hotelList.Add(hotel);
}
}
HotelListBox.Invoke(new Action(() =>
{
BindingSource bs = new BindingSource();
bs.DataSource = hotelList;
this.HotelListBox.DataSource = bs;
}));
}
private void HotelCrawler(OnCompletedEventArgs e)
{
hd = ConvertHelper.ConvertWebeleToHotel(e.driver);
loadHotelInfo();
CheckBtnStatus();
}
void hotelCrawler_OnCompleted(object sender, OnCompletedEventArgs e)
{
HotelCrawler(e);
}
void hotelsCrawler_OnCompleted(object sender, OnCompletedEventArgs e)
{
var s = e.pageSoure;
ExtractCitys(e);
}
