多线程应用一

1.图片下载

  1  public class Program
  2     {
  3         static void Main(string[] args)
  4         {
  5             string[] str = { "model", "sexy", "belle", "stars" };
  6             for (int url = 0; url < str.Length; url++)
  7             {
  8                 Thread thread = new Thread(DownLoad);
  9
 10                 thread.Start(str[url]);
 11             }
 12             Console.Read();
 13         }
 14
 15         public static void DownLoad(object category)
 16         {
 17             string url = string.Empty;
 18
 19             for (int purl = 9014; purl > 10; purl--)
 20             {
 21                 for (int pageSize = 0; pageSize < 20; pageSize++)
 22                 {
 23                     try
 24                     {
 25                         if (pageSize == 0)
 26                             url = "http://www.mm8mm8.com/" + category + "/" + purl + ".html";
 27                         else
 28                             url = "http://www.mm8mm8.com/" + category + "/" + purl + "_" + pageSize + ".html";
 29
 30                         //创建http链接
 31                         var request = (HttpWebRequest)WebRequest.Create(url);
 32
 33                         request.Timeout = 1000 * 5;    //5s过期
 34
 35                         var response = (HttpWebResponse)request.GetResponse();
 36
 37                         Stream stream = response.GetResponseStream();
 38
 39                         StreamReader sr = new StreamReader(stream);
 40
 41                         string content = sr.ReadToEnd();
 42
 43                         var list = GetHtmlImageUrlList(content);
 44
 45                         WebClient client = new WebClient();
 46
 47
 48                         var directoryName = @"C:\down\";
 49
 50                         if (!Directory.Exists(directoryName))
 51                             Directory.CreateDirectory(directoryName);
 52
 53                         var fileName = string.Empty;
 54
 55                         if (list.Count == 0)
 56                         {
 57                             Console.WriteLine("时间：" + DateTime.Now + " 当前网址：" + url + "  未发现图片");
 58                             break;
 59                         }
 60
 61                         try
 62                         {
 63
 64                             fileName = category + "_" + purl + "_" + (pageSize + 1) + ".jpg";
 65
 66                             var localFile = directoryName + fileName;
 67
 68                             var imageRequest = (HttpWebRequest)WebRequest.Create(list[0]);
 69
 70                             imageRequest.Timeout = 1000 * 5;  //5s 超时
 71
 72                             var imageResponse = (HttpWebResponse)imageRequest.GetResponse();
 73
 74                             var s = imageResponse.GetResponseStream();
 75
 76                             Image image = Image.FromStream(s);
 77
 78                             image.Save(localFile);
 79
 80                             image.Dispose();
 81
 82                             Console.WriteLine("时间：" + DateTime.Now + "  图片：" + fileName + " 已经下载   存入磁盘位置：" + localFile);
 83
 84                         }
 85                         catch (Exception e)
 86                         {
 87                             Console.WriteLine("时间：" + DateTime.Now + " 当前图片：" + fileName + " 错误信息：" + e.Message);
 88                             continue;
 89                         }
 90                     }
 91                     catch (Exception ex)
 92                     {
 93                         Console.WriteLine("时间：" + DateTime.Now + " 当前网址：" + url + " 错误信息：" + ex.Message);
 94                     }
 95                 }
 96             }
 97         }
 98
 99         /// <summary>
100         /// 取得HTML中所有图片的 URL。
101         /// </summary>
102         /// <param name="sHtmlText">HTML代码</param>
103         /// <returns>图片的URL列表</returns>
104         public static List<string> GetHtmlImageUrlList(string sHtmlText)
105         {
106             // 定义正则表达式用来匹配 img 标签
107             Regex regImg = new Regex(@"<img\b[^<>]*?\bsrc[\s\t\r\n]*=[\s\t\r\n]*[""‘]?[\s\t\r\n]*(?<imgUrl>[^\s\t\r\n""‘<>]*)[^<>]*?/?[\s\t\r\n]*>", RegexOptions.IgnoreCase);
108
109             // 搜索匹配的字符串
110             MatchCollection matches = regImg.Matches(sHtmlText);
111
112             List<string> sUrlList = new List<string>();
113
114             // 取得匹配项列表
115             foreach (Match match in matches)
116                 sUrlList.Add(match.Groups["imgUrl"].Value);
117             return sUrlList;
118         }
119     }