[爬蟲]抓取知乎百萬用戶信息之爬蟲模塊

             點擊我前往Github查看源代碼   別忘記starhtml

本項目github地址:https://github.com/wangqifan/ZhiHu     node

UserManage是獲取用戶信息的爬蟲模塊git

public   class UserManage
    {
        private string html;

        private string url_token;

     }

構造函數github

 

用戶主頁的uRL格式爲"https://www.zhihu.com/people/"+url_token+"/following";json

 public UserManage(string urltoken)
         {
             url_token = urltoken;
         }

先封裝一個獲取html頁面的方法api

 

 private bool GetHtml()

        {                

            string url="https://www.zhihu.com/people/"+url_token+"/following";

            html = HttpHelp.DownLoadString(url);

            return  !string.IsNullOrEmpty(html);

        }

拿到了html頁面,接下來是剝取頁面中的JSON,藉助HtmlAgilityPack數組

public  void  analyse()
        {
                if (GetHtml())
                {
                    try
                    {
                        Stopwatch watch = new Stopwatch();
                        watch.Start();
                        HtmlDocument doc = new HtmlAgilityPack.HtmlDocument();
                        doc.LoadHtml(html);
                        HtmlNode node = doc.GetElementbyId("data");
                        StringBuilder stringbuilder =new StringBuilder(node.GetAttributeValue("data-state", ""));
                        stringbuilder.Replace(""", "'");           
                        stringbuilder.Replace("&lt;", "<");
                        stringbuilder.Replace("&gt;", ">");
                     
                        watch.Stop();
                       Console.WriteLine("分析Html用了{0}毫秒", watch.ElapsedMilliseconds.ToString());
                       
                    }
                    catch (Exception ex)
                    {
                        Console.WriteLine(ex.ToString());
                    }
                }
            
            }    

添加用戶的關注列表的連接服務器

 private void  GetUserFlowerandNext(string json)

        {

                 string foollowed = "https://www.zhihu.com/api/v4/members/" + url_token + "/followers?include=data%5B*%5D.answer_count%2Carticles_count%2Cfollower_count%2Cis_followed%2Cis_following%2Cbadge%5B%3F(type%3Dbest_answerer)%5D.topics&offset=0&limit=20";

                 string following = "https://www.zhihu.com/api/v4/members/" + url_token + "/followees?include=data%5B%2A%5D.answer_count%2Carticles_count%2Cfollower_count%2Cis_followed%2Cis_following%2Cbadge%5B%3F%28type%3Dbest_answerer%29%5D.topics&limit=20&offset=0";

                 RedisCore.PushIntoList(1, "nexturl", following);

                 RedisCore.PushIntoList(1, "nexturl", foollowed);

        }

 

對json數據進一步剝取,只要用戶的信息,藉助JSON解析工具Newtonsoft.Json多線程

 

private void  GetUserInformation(string json)
        {  
                JObject obj = JObject.Parse(json);
                string xpath = "['" + url_token + "']";
                JToken tocken = obj.SelectToken("['entities']").SelectToken("['users']").SelectToken(xpath);
                RedisCore.PushIntoList(2, "User", tocken.ToString());
              
        }  
 

 

如今來完成下analyse函數
 public  void  analyse()
        {
                if (GetHtml())
                {
                    try
                    {
                        Stopwatch watch = new Stopwatch();
                        watch.Start();
                        HtmlDocument doc = new HtmlAgilityPack.HtmlDocument();
                        doc.LoadHtml(html);
                        HtmlNode node = doc.GetElementbyId("data");
                        StringBuilder stringbuilder =new StringBuilder(node.GetAttributeValue("data-state", ""));
                        stringbuilder.Replace(""", "'");           
                        stringbuilder.Replace("<", "<");
                        stringbuilder.Replace(">", ">");
                        GetUserInformation(stringbuilder.ToString());
                        GetUserFlowerandNext(stringbuilder.ToString());
                        watch.Stop();
                        Console.WriteLine("分析Html用了{0}毫秒", watch.ElapsedMilliseconds.ToString());
                       
                    }
                    catch (Exception ex)
                    {
                        Console.WriteLine(ex.ToString());
                    }
                }
            
            }    
        }
    

  

UrlTask是從nexturl隊列獲取用戶的關注列表的url,獲取關注列表。服務器返回的Json的數據函數

封裝一個對象的序列化和反序列化的類

public   class SerializeHelper
    {
        /// <summary>
        /// 對數據進行序列化
        /// </summary>
        /// <param name="value"></param>
        /// <returns></returns>
        public static string SerializeToString(object value)
        {
            return JsonConvert.SerializeObject(value);
        }
        /// <summary>
        /// 反序列化操做
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="str"></param>
        /// <returns></returns>
        public static T DeserializeToObject<T>(string str)
        {
          
            return JsonConvert.DeserializeObject<T>(str);
        }
}

封裝UrlTask類

 
 
 public class UrlTask { private string url { get; set; } private string JSONstring { get; set; } public UrlTask(string _url) { url = _url; } }
 
 

添加一個獲取資源的方法

 
 
 private bool GetHtml() { JSONstring= HttpHelp.DownLoadString(url); Console.WriteLine("Json下載完成"); return !string.IsNullOrEmpty(JSONstring); }
解析json方法
 
 
 public  void Analyse() { try { if (GetHtml()) { Stopwatch watch = new Stopwatch(); watch.Start(); followerResult result = SerializeHelper.DeserializeToObject<followerResult>(JSONstring); if (!result.paging.is_end) { RedisCore.PushIntoList(1, "nexturl", result.paging.next); } foreach (var item in result.data) { int type=Math.Abs(item.GetHashCode())% 3 + 3; if (RedisCore.InsetIntoHash(type, "urltokenhash", item.url_token, "存在")) { RedisCore.PushIntoList(1, "urltoken", item.url_token); } } watch.Stop(); Console.WriteLine("解析json用了{0}毫秒",watch.ElapsedMilliseconds.ToString()); } } catch (Exception ex) { Console.WriteLine(ex.ToString()); } }
 
 

解析:若是result.paging.is_end爲true,那麼這個是用戶關注列表的最後一頁,那麼它的nexturl應該加入隊列,負責不要加入,對於後面的用戶數組,由於信息不去全,不要了,有了Id前往主頁獲取詳細信息。

 

 模塊組合

封裝一個一個方法,從隊列拿到nextutl,前往用戶的關注列表,拿到更多用戶ID

  private static void GetNexturl() { string nexturl = RedisCore.PopFromList(1, "nexturl"); if (!string.IsNullOrEmpty(nexturl)) { UrlTask task = new UrlTask(nexturl); task.Analyse(); } }

封裝一個方法,循環從隊列獲取用戶的urltoken(若是隊列空了,執行GetNexturl),前往用戶主頁,獲取信息

private static void GetUser(object data) { while (true) { string url_token = RedisCore.PopFromList(1, "urltoken"); Console.WriteLine(url_token); if (!string.IsNullOrEmpty(url_token)) { UserManage manage = new UserManage(url_token); manage.analyse(); } else { GetNexturl(); } } }

在main函數裏面執行這些方法,因爲任務量大,採用多線程,線程數視狀況而定

 

for (int i = 0; i < 10; i++) { ThreadPool.QueueUserWorkItem(GetUser); }

 

添加種子數據,用於剛開始時候隊列都是空的,須要添加種子數據

  1. 手動添加,在redile-cl.exe敲命令
  2. 在main函數中加入

    

 UserTask task=new UserTask(「某個用戶的uRLtoken」); task.analyse();

 

執行一次以後要註釋掉,避免重複

相關文章
相關標籤/搜索