コード例 #1
0
        /// <summary>
        /// 爬取数据
        /// </summary>
        public static void CrawlHotMovie()
        {
            Task.Factory.StartNew(() =>
            {
                try
                {
                    LogHelper.Info("CrawlHotMovie Start...");
                    var htmlDoc    = HTTPHelper.GetHTMLByURL("http://www.dy2018.com/");
                    var dom        = htmlParser.Parse(htmlDoc);
                    var lstDivInfo = dom.QuerySelectorAll("div.co_content222");
                    if (lstDivInfo != null)
                    {
                        //前三个DIV为新电影
                        foreach (var divInfo in lstDivInfo.Take(3))
                        {
                            divInfo.QuerySelectorAll("a").Where(a => a.GetAttribute("href").Contains("/i/")).ToList().ForEach(
                                a =>
                            {
                                var onlineURL = "http://www.dy2018.com" + a.GetAttribute("href");
                                if (!hotMovieList.IsContainsMoive(onlineURL))
                                {
                                    MovieInfo movieInfo = Dy2018MoviceInfoHelper.GetMovieInfoFromOnlineURL(onlineURL);
                                    if (movieInfo != null && movieInfo.XunLeiDownLoadURLList != null && movieInfo.XunLeiDownLoadURLList.Count != 0)
                                    {
                                        hotMovieList.AddToMovieDic(movieInfo);
                                    }
                                }
                            });
                        }
                    }

                    LogHelper.Info("CrawlHotMovie Finish...");
                }
                catch (Exception ex)
                {
                    LogHelper.Error("CrawlHotMovie Exception", ex);
                }
            });
        }
コード例 #2
0
        /// <summary>
        /// 从在线网页提取数据
        /// </summary>
        /// <param name="i"></param>
        private static void FillMovieFromOnline(string indexURL)
        {
            var htmlDoc    = HTTPHelper.GetHTMLByURL(indexURL);
            var dom        = htmlParser.Parse(htmlDoc);
            var lstDivInfo = dom.QuerySelectorAll("div.co_content8");

            if (lstDivInfo != null)
            {
                lstDivInfo.FirstOrDefault().QuerySelectorAll("a").Where(a => a.GetAttribute("href").Contains("/i/")).ToList()
                .ForEach(a =>
                {
                    var onlineURL = "http://www.dy2018.com" + a.GetAttribute("href");
                    if (!latestMovieList.IsContainsMoive(onlineURL))
                    {
                        MovieInfo movieInfo = Dy2018MoviceInfoHelper.GetMovieInfoFromOnlineURL(onlineURL);
                        if (movieInfo != null && movieInfo.XunLeiDownLoadURLList != null && movieInfo.XunLeiDownLoadURLList.Count != 0)
                        {
                            latestMovieList.AddToMovieDic(movieInfo);
                        }
                    }
                });
            }
        }