爬蟲系列一次採集.NET WebForm網站的坎坷歷程

2020 年 5 月 12 日
筆記
c#, 爬蟲, 採集

今天接到一個活，需要統計人員的工號資訊，由於種種原因不能直接連資料庫 [無奈]、[無奈]、[無奈]。採取迂迴方案，寫個工具自動登錄網站，採集用戶資訊。

這也不是第一次採集ASP.NET網站，以前採集的時候就知道，這種網站採集比較麻煩，尤其是WebForm的ASP.NET 網站，那叫一個費勁。

喜歡現在流行的Restful模式的網站，數據介面採集那才叫舒服。

閑話少說，開干

工作量不大，HTTP純手寫

先準備下一個GET/POST預備使用

　　　　public static string Get(string url, Action<string> SuccessCallback, Action<string> FailCallback) {
            HttpWebRequest req = WebRequest.Create(url) as HttpWebRequest;
            req.Method = "GET";
            req.UserAgent = "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36";
            req.Accept = "*/*";
            req.KeepAlive = true;
            req.ServicePoint.ConnectionLimit = int.MaxValue;
            req.ServicePoint.Expect100Continue = false;
            req.CookieContainer = sznyCookie; #靜態變數
            req.Credentials = System.Net.CredentialCache.DefaultCredentials;
            string msg = "";
            using (HttpWebResponse rsp = req.GetResponse() as HttpWebResponse)
            {
                using (StreamReader reader = new StreamReader(rsp.GetResponseStream()))
                {
                    msg = reader.ReadToEnd();
                }
            }
            return msg;
        }
 
　　　　public static string Post(string url, Dictionary<string, string> dicParms, Action<string> SuccessCallback, Action<string> FailCallback) {
            StringBuilder data = new StringBuilder();
            foreach (var kv in dicParms) {
                if (kv.Key.StartsWith("header"))
                    continue;
                data.Append($"&{Common.UrlEncode( kv.Key,Encoding.UTF8)}={ Common.UrlEncode( kv.Value,Encoding.UTF8)}");
            }
            if (data.Length > 0)
                data.Remove(0, 1);
            HttpWebRequest req = WebRequest.Create(url) as HttpWebRequest;
            req.Method = "POST";
            req.KeepAlive = true;
            req.CookieContainer = sznyCookie;
            req.Connection = "KeepAlive";
            req.KeepAlive = true;
            req.ContentType = "application/x-www-form-urlencoded";
            req.Accept = "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9";
            req.Referer = url;
            if (dicParms.ContainsKey("ScriptManager1"))
            {
                req.Headers.Add("X-MicrosoftAjax", "Delta=true");
                req.Headers.Add("X-Requested-With", "XMLHttpRequest");
                req.ContentType = "application/x-www-form-urlencoded; charset=UTF-8";
                req.Accept = "*/*";
            }
            req.Headers.Add("Cache-Control", "no-cache");
            
            req.UserAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.163 Safari/537.36";
            req.ServicePoint.ConnectionLimit = int.MaxValue;
            req.ServicePoint.Expect100Continue = false;
            req.AllowAutoRedirect = true;
            req.Credentials = System.Net.CredentialCache.DefaultCredentials;

            byte[] buffer = Encoding.UTF8.GetBytes(data.ToString());
            using (Stream reqStream = req.GetRequestStream())
            {
                reqStream.Write(buffer, 0, buffer.Length);
            }
            string msg = "";
            using (HttpWebResponse rsp = req.GetResponse() as HttpWebResponse)
            {
                using (StreamReader reader = new StreamReader(rsp.GetResponseStream()))
                {
                    
                    msg = reader.ReadToEnd();
                    if (msg.Contains("images/dl.jpg") || msg.Contains("pageRedirect||%2flogin.aspx"))
                    {
                        //登錄失敗
                        if (FailCallback != null)
                            FailCallback(msg);
                    }
                    else {
                        if (SuccessCallback!=null)
                            SuccessCallback(msg);
                    }
                }
            }
            return msg;
        }

整個過程分為登陸、用戶資訊列表、用戶資訊詳情，分三步走來完成這個項目

登陸

根據Chrome抓包結果編寫Login，帳號密碼沒有任何加密，直接明文顯示了，直接用了，根據是否跳轉頁面判斷是否登陸成功。調試查看結果登陸成功了。

根據上面的抓包數據，可以調用下面的程式碼確定是否登陸成功。

　　　　　　public static bool SznyLogin(string username, string password, Action<string> SuccessCallback, Action<string> FailCallback) {
            string url = "//127.0.0.1/login.aspx";
            string msg = Get(url, SuccessCallback, FailCallback);
            if (msg.Trim().Length > 0) {
                Dictionary<string, string> dicParms = new Dictionary<string, string>();
                dicParms.Add("__VIEWSTATE", "");
                dicParms.Add("__EVENTVALIDATION", "");
                dicParms.Add("Text_Name", "");
                dicParms.Add("Text_Pass", "");
                dicParms.Add("btn_Login.x", new Random().Next(100).ToString());
                dicParms.Add("btn_Login.y", new Random().Next(200).ToString());
                MatchCollection mc = Regex.Matches(msg, @"<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?value=""(?<val>[^""]*?)""[^<>]*?/?>|<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>|<select[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                foreach (Match mi in mc)
                {
                    if (dicParms.ContainsKey(mi.Groups["name"].Value.Trim()))
                        dicParms[mi.Groups["name"].Value.Trim()] = mi.Groups["val"].Value.Trim();
                }
                dicParms["Text_Name"] = username;
                dicParms["Text_Pass"] = password;

                msg=Post(url, dicParms, SuccessCallback, FailCallback);

                if (msg.Contains("images/dl.jpg") || msg.Contains("pageRedirect||%2flogin.aspx"))
                {
                    return false;
                }
                else
                    return true;
            }
            return false;
        }

抓取人員資訊

看到下面這個頁面，失望了，列表上沒有工號，如果列表上有工號設置一頁顯示全部資訊就可以把所有的數據都抓取到了。

換個思路：是不是我直接設置一頁顯示所有的數據後，然後根據員工ID可以獲取到所有的資訊呢？

接下來點擊任意一條資訊後，查看詳情，顯示下面的調用結果。Url上沒有ID，Get這條路走不通了，查看Post的數據，更失望，沒有ID,通過行資訊綁定。傳統的WebForm 提交模式…

把所有的數據顯示到一頁，把列表的數據先採集完，然後最後一個頁面一個頁面的採集工號資訊。

　　　　public static CookieContainer sznyCookie = new CookieContainer();
        /// <summary>
        /// 員工資訊
        /// </summary>
        public static Dictionary<int, Dictionary<string,string>> dicSznyEmployees = new Dictionary<int, Dictionary<string, string>>();

        public static Dictionary<string, string> dicSznyEmployeeParms = new Dictionary<string, string>();
        /// <summary>
        /// 人員順序號
        /// </summary>
        public static ConcurrentQueue<int> queueSznyEmployeeInfo = new ConcurrentQueue<int>();

        public static ConcurrentQueue<int> queueSuccessEmployeeInfo = new ConcurrentQueue<int>();
 
　　　　public static bool SznyEmployeeList(Action<string> SuccessCallback, Action<string> FailCallback)
        {
            string url = $"//127.0.0.1/HumanResources/EmployeeManage/EmployeeInfoList.aspx";
            string msg = Get(url, SuccessCallback, FailCallback);
            if (msg.Trim().Length > 100)
            {
                //統計參數
                //__doPostBack\('(?<name>[^']*?)'
                //new Regex(@"__doPostBack\('(?<name>[^']*?)'", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                string name = "";
                MatchCollection mc = Regex.Matches(msg, @"__doPostBack\('(?<name>[^']*?)'", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                foreach (Match mi in mc)
                {
                    name = mi.Groups["name"].Value.Trim();
                    break;
                }
                //(?<=<a[^<>]*?href="javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)"[^<>]*?>條/頁)
                //new Regex(@"(?<=<a[^<>]*?href=""javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)""[^<>]*?>條/頁)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                string smname = "";
                Match m = Regex.Match(msg, @"(?<=<a[^<>]*?href=""javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)""[^<>]*?>條/頁)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                if (m.Success)
                    smname = m.Value.Trim().Replace("'", "").Replace("&#39;", "");

                //<input[^<>]*?name="(?<name>[^"]*?)"[^<>]*?value="(?<val>[^"]*?)"[^<>]*?/?>|<input[^<>]*?name="(?<name>[^"]*?)"[^<>]*?"[^<>]*?/?>|<select[^<>]*?name="(?<name>[^"]*?)"[^<>]*?"[^<>]*?/?>
                //new Regex(@"<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?value=""(?<val>[^""]*?)""[^<>]*?/?>|<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>|<select[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                Dictionary<string, string> dicParms = new Dictionary<string, string>();
                dicParms.Add("ScriptManager1", $"UpdatePanel1|{smname}");
                dicParms.Add("__EVENTTARGET", smname);
                dicParms.Add("__EVENTARGUMENT", "");
                dicParms.Add("__VIEWSTATE", "");
                dicParms.Add("__EVENTVALIDATION", "");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_GridViewID", $"{name}");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentPage", "1");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalPage", "1");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalCount", "1");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageSize", "1");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageCount", "1");
                dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentNum", "0");
                dicParms.Add("XM", "ZXMCHECK");

                List<string> lstParms = new List<string>() { "XM", "MdGridView_t_unitemployees_dwyg_iCurrentPage", "MdGridView_t_unitemployees_dwyg_GridViewID", "MdGridView_t_unitemployees_dwyg_iCurrentNum", "MdGridView_t_unitemployees_dwyg_iPageCount", "MdGridView_t_unitemployees_dwyg_iPageSize", "Button_Query", "__EVENTTARGET", "__EVENTARGUMENT", "Button_SelQuery", "Button_view", "Button_edit", "Button_out", "ImageButton_Tx", "ImageButton_xx1", "Button_qd", "MdGridView_t_unitemployees_dwyg_GridViewID", "__ASYNCPOST", "MdGridView_t_unitemployees_dwyg__PageSetText" };
                mc = Regex.Matches(msg, @"<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?value=""(?<val>[^""]*?)""[^<>]*?/?>|<input[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>|<select[^<>]*?name=""(?<name>[^""]*?)""[^<>]*?""[^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                foreach (Match mi in mc)
                {

                    if (lstParms.Contains(mi.Groups["name"].Value.Trim()))
                        continue;

                    if (dicParms.ContainsKey(mi.Groups["name"].Value.Trim()))
                        dicParms[mi.Groups["name"].Value.Trim()] = mi.Groups["val"].Value.Trim();
                    else
                        dicParms.Add(mi.Groups["name"].Value.Trim(), mi.Groups["val"].Value.Trim());
                }
                if (dicParms.ContainsKey("MdGridView_t_unitemployees_dwyg$_PageSetText"))
                    dicParms["MdGridView_t_unitemployees_dwyg$_PageSetText"] = "1200";
                else
                    dicParms.Add("MdGridView_t_unitemployees_dwyg$_PageSetText", "100");//1200條 每頁

                msg = Post(url, dicParms, SuccessCallback, FailCallback);

                dicSznyEmployees.Clear();
                dicSznyEmployeeParms.Clear();
                dicSznyEmployeeParms.Clear();
                dicSznyEmployeeParms.Add("__EVENTTARGET", "");
                dicSznyEmployeeParms.Add("__EVENTARGUMENT", "");
                dicSznyEmployeeParms.Add("__VIEWSTATE", dicParms["__VIEWSTATE"]);
                dicSznyEmployeeParms.Add("__EVENTVALIDATION", dicParms["__EVENTVALIDATION"]);
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_GridViewID", $"{name}");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentPage", "1");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iTotalPage", "1");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iTotalCount", "1");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iPageSize", "1");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iPageCount", "1");
                dicSznyEmployeeParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentNum", "0");
                dicSznyEmployeeParms.Add("XM", "ZXMCHECK");
                
                lstParms.Clear();
                lstParms = new List<string>() { "XM", "__EVENTTARGET", "__EVENTARGUMENT", "Button_Query", "Button_SelQuery", };
                lstParms.Add("Button_edit");
                lstParms.Add("Button_out");
                lstParms.Add("ImageButton_Tx");
                lstParms.Add("ImageButton_xx1");
                lstParms.Add("Button_qd");
                lstParms.Add("MdGridView_t_unitemployees_dwyg_GridViewID");
                lstparms.add("mdgridview_t_unitemployees_dwyg_icurrentpage");
                lstparms.add("mdgridview_t_unitemployees_dwyg_itotalpage");
                lstparms.add("mdgridview_t_unitemployees_dwyg_itotalcount");
                lstparms.add("mdgridview_t_unitemployees_dwyg_ipagesize");
                lstparms.add("mdgridview_t_unitemployees_dwyg_ipagecount");
                lstParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentNum");

                mc = Regex.Matches(msg, @"<input[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?value=[""'](?<val>[^'""]*?)[""'][^<>]*?/?>|<input[^<>]*?name=[""'](?<name>[^'""]*?)[""'][^<>]*?[""'][^<>]*?/?>|<select[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?[""'][^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                foreach (Match mi in mc)
                {

                    if (lstParms.Contains(mi.Groups["name"].Value.Trim()))
                        continue;

                    if (dicSznyEmployeeParms.ContainsKey(mi.Groups["name"].Value.Trim()))
                        dicSznyEmployeeParms[mi.Groups["name"].Value.Trim()] = mi.Groups["val"].Value.Trim();
                    else
                        dicSznyEmployeeParms.Add(mi.Groups["name"].Value.Trim(), mi.Groups["val"].Value.Trim());
                }
                int cnt = int.Parse(dicSznyEmployeeParms["MdGridView_t_unitemployees_dwyg_iTotalCount"]);
                for (int i = 1; i <= cnt; i++)
                    queueSznyEmployeeInfo.Enqueue(i);

                //獲取TR
                //< tr[^<>] *? name = "SelectTR"[^<>] *?>.*?</ tr >
                //new Regex(@"<tr[^<>]*?name=""SelectTR""[^<>]*?>.*?</tr>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                mc = Regex.Matches(msg, @"<tr[^<>]*?name=""SelectTR""[^<>]*?>.*?</tr>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                foreach (Match mi in mc)
                {
                    //獲取td
                    //(?<=<td[^<>]*?>).*?(?=</td>)
                    //new Regex("(?<=<td[^<>]*?>).*?(?=</td>)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                    MatchCollection mic = Regex.Matches(mi.Value, "(?<=<td[^<>]*?>).*?(?=</td>)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                    int ix = int.Parse(mic[1].Value.Trim());
                    if (!dicSznyEmployees.ContainsKey(ix))
                    {
                        dicSznyEmployees.Add(ix, new Dictionary<string, string>());
                    }
                    queueSznyEmployeeInfo.Enqueue(ix);

                    dicSznyEmployees[ix].Add("UserName", mic[2].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("PersonID", mic[3].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Birthday", mic[4].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Sex", mic[5].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("HomePhone", mic[6].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("TelPhone", mic[7].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Mail", mic[8].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Address", mic[9].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("MinZu", mic[10].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("AddressJiGuan", mic[11].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("ZhengZhiMianmao", mic[12].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Paiqianshijian", mic[13].Value.Trim().Replace("&nbsp;", ""));
                    dicSznyEmployees[ix].Add("Remark", mic[14].Value.Trim().Replace("&nbsp;", ""));
                }
            }
            return true;
        }

這樣所有的人員資訊一次性採集到靜態變數字典中了，剩下的一個工號可以慢慢獲取了。

既然是這樣，老實的分析Post數據，按照格式Post數據把。

分析完Post的數據後，突發奇想，我是不是可以通過相同的__ViewState和__EVENTVALIDATION POST數據呢？說干就干。

寫程式碼跳轉到員工列表頁面，然後POST數據設置一頁顯示所有數據。

所有的POST的參數，保存到一個靜態變數中。

發現POST批量提交的時候，前3次正常，以後就直接未登錄。

果斷放棄，換思路。

那如果這樣不行可不可以把所有的數據放到一個頁面上，然後每次獲取一次頁面，然後根據順序號POST數據呢。

上面已經把所有的列表數據都採集完了，順序號也固定了，然後在POST數據的時候，發現有的人員和工號不對應。

這時候去分析為什麼數據會出現不對應的情況呢？發現正則表達式寫的還有問題。獲取頁面的Input的時候，屬性有可能使用雙引號，也有可能使用單引號。

正則表達式由原來的

<input[^<>]*?name="(?<name>[^"]*?)"[^<>]*?value="(?<val>[^"]*?)"[^<>]*?/?>|<input[^<>]*?name="(?<name>[^"]*?)"[^<>]*?"[^<>]*?/?>|<select[^<>]*?name="(?<name>[^"]*?)"[^<>]*?"[^<>]*?/?>

修改為

<input[^<>]*?name=["'](?<name>[^"']*?)["'][^<>]*?value=["'](?<val>[^'"]*?)["'][^<>]*?/?>|<input[^<>]*?name=["'](?<name>[^'"]*?)["'][^<>]*?["'][^<>]*?/?>|<select[^<>]*?name=["'](?<name>[^"']*?)["'][^<>]*?["'][^<>]*?/?>

由於網站非同步提交，也就是以前WEBForm採用的ScriptManager,提交的時候返回的HTML不是整個Document,沒有注意，以為沒有返回__ViewState。所以採用GET的時候獲取的__ViewState繼續執行獲取工號的操作。發現獲取的工號都是錯誤，人員與工號對不上

麻爪了，不知道該咋辦了。猶豫了一下後，上Fiddler吧，一點點的看提交的參數是否有區別。發現正常網站在Get到頁面後，通過調整每頁x條數據後，提交的ViewState與原來的不一致。尋尋覓覓覓覓尋尋最後發現非同步返回的HTML中，最後有ViewState….

由於返回的數據順序，每次也不一樣，也是造成人員、工號不一致的原因。

提交後正常了，但是1000多條的員工資訊，每次提交都是2000多個參數。看著冗長的POST數據，無語了。這樣提交先不說網站本身就慢。我提交這麼多網站會不會更慢，我的系統是不是也會更慢。

怎麼辦？

是不是有可能把分頁設置成每頁只有一條數據，然後每次翻頁，採集數據。簡單試試把

先修改獲取列表頁面數據，把數據設置成一條每頁，此時不再採集列表中的資訊。而是記錄總共多少頁，放入隊列中，共定時任務去分頁採集數據。列表資訊通過後面的分頁數據採集。

由於網站是內部系統，為了不影響系統的正常運行，每次只採集一條資訊，等待這條資訊採集完成後，在採集下一頁資訊。

採集列表

　　　　　public static void ReqSznyEmployeeList(int ix,Action<string> SuccessCallback, Action<string> FailCallback)
        {

            string url = $"//127.0.0.1/HumanResources/EmployeeManage/EmployeeInfoList.aspx";
            HttpWebRequest req = WebRequest.Create(url) as HttpWebRequest;
            req.Method = "GET";
            req.UserAgent = "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36";
            req.Accept = "*/*";
            req.KeepAlive = true;
            req.ServicePoint.ConnectionLimit = int.MaxValue;
            req.ServicePoint.Expect100Continue = false;
            req.CookieContainer = sznyCookie;
            req.Credentials = System.Net.CredentialCache.DefaultCredentials;

            req.BeginGetResponse(new AsyncCallback(RspSznyEmployeeList), new object[] { req, url,ix, SuccessCallback, FailCallback });
        }



　　　　　private static void RspSznyEmployeeList(IAsyncResult result)
        {
            object[] parms = result.AsyncState as object[];
            HttpWebRequest req = parms[0] as HttpWebRequest;
            string url = parms[1].ToString();
            int ix = int.Parse(parms[2].ToString());
            Action<string> SuccessCallback = parms[3] as Action<string>;
            Action<string> FailCallback = parms[4] as Action<string>;
            try
            {
                using (HttpWebResponse rsp = req.EndGetResponse(result) as HttpWebResponse)
                {
                    using (StreamReader reader = new StreamReader(rsp.GetResponseStream()))
                    {
                        string msg = "";
                        msg = reader.ReadToEnd();
                        //統計參數
                        //__doPostBack\('(?<name>[^']*?)'
                        //new Regex(@"__doPostBack\('(?<name>[^']*?)'", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                        string name = "";
                        MatchCollection mc = Regex.Matches(msg, @"__doPostBack\('(?<name>[^']*?)'", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        foreach (Match mi in mc)
                        {
                            name = mi.Groups["name"].Value.Trim();
                            break;
                        }

                        //(?<=<a[^<>]*?href="javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)"[^<>]*?>條/頁)
                        //new Regex(@"(?<=<a[^<>]*?href=""javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)""[^<>]*?>條/頁)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                        string smname = "MdGridView_t_unitemployees_dwyg$_SearchGo";
                        //Match m = Regex.Match(msg, @"(?<=<a[^<>]*?href=""javascript:__dopostback\()[^<>]*?(?=,[^<>]*?\)""[^<>]*?>條/頁)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        //<input[^<>]*?name=["'](?<name>[^"']*?)["'][^<>]*?value=["'](?<val>[^'"]*?)["'][^<>]*?/?>|<input[^<>]*?name=["'](?<name>[^'"]*?)["'][^<>]*?["'][^<>]*?/?>|<select[^<>]*?name=["'](?<name>[^"']*?)["'][^<>]*?["'][^<>]*?/?>
                        //new Regex(@"<input[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?value=[""'](?<val>[^'""]*?)[""'][^<>]*?/?>|<input[^<>]*?name=[""'](?<name>[^'""]*?)[""'][^<>]*?[""'][^<>]*?/?>|<select[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?[""'][^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                        Dictionary<string, string> dicParms = new Dictionary<string, string>();
                        dicParms.Add("ScriptManager1", $"UpdatePanel1|{smname}");
                        dicParms.Add("__EVENTTARGET", smname);
                        dicParms.Add("__EVENTARGUMENT", "");
                        dicParms.Add("__VIEWSTATE", "");
                        dicParms.Add("__EVENTVALIDATION", "");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_GridViewID", $"{name}");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentPage", ix.ToString());
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalPage", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalCount", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageSize", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageCount", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentNum", "0");
                        //dicParms.Add("MdGridView_t_unitemployees_dwyg$_SearchTextBox", ix.ToString());

                        dicParms.Add("XM", "ZXMCHECK");
                        List<string> lstParms = new List<string>() { "ScriptManager1", "XM", "MdGridView_t_unitemployees_dwyg_iCurrentNum",  "Button_Query", "__EVENTTARGET", "__EVENTARGUMENT", "Button_SelQuery", "Button_view", "Button_edit", "Button_out", "ImageButton_Tx", "ImageButton_xx1", "Button_qd",  "__ASYNCPOST", "MdGridView_t_unitemployees_dwyg__PageSetText" };
                        mc = Regex.Matches(msg, @"<input[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?value=[""'](?<val>[^'""]*?)[""'][^<>]*?/?>|<input[^<>]*?name=[""'](?<name>[^'""]*?)[""'][^<>]*?[""'][^<>]*?/?>|<select[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?[""'][^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        foreach (Match mi in mc)
                        {

                            if (lstParms.Contains(mi.Groups["name"].Value.Trim()))
                                continue;

                            if (dicParms.ContainsKey(mi.Groups["name"].Value.Trim()))
                                dicParms[mi.Groups["name"].Value.Trim()] = mi.Groups["val"].Value.Trim();
                            else
                                dicParms.Add(mi.Groups["name"].Value.Trim(), mi.Groups["val"].Value.Trim());
                        }
                        if (dicParms.ContainsKey("MdGridView_t_unitemployees_dwyg$_PageSetText"))
                            dicParms["MdGridView_t_unitemployees_dwyg$_PageSetText"] = "1";
                        else
                            dicParms.Add("MdGridView_t_unitemployees_dwyg$_PageSetText", "1");//1200條 每頁
                        if (dicParms.ContainsKey("MdGridView_t_unitemployees_dwyg_iPageCount"))
                            dicParms["MdGridView_t_unitemployees_dwyg_iPageCount"] = "1";
                        else
                            dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageCount", "1");
                        if (dicParms.ContainsKey("MdGridView_t_unitemployees_dwyg_iPageSize"))
                            dicParms["MdGridView_t_unitemployees_dwyg_iPageSize"] = "1";
                        else
                            dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageSize", "1");

                        if (dicParms.ContainsKey("MdGridView_t_unitemployees_dwyg$_SearchTextBox"))
                            dicParms["MdGridView_t_unitemployees_dwyg$_SearchTextBox"] = $"{ix}";
                        else
                            dicParms.Add("MdGridView_t_unitemployees_dwyg$_SearchTextBox", $"{ix}");/*第幾頁*/

                        dicParms["MdGridView_t_unitemployees_dwyg_iTotalPage"] = dicParms["MdGridView_t_unitemployees_dwyg_iTotalCount"];
                        msg = Post(url, dicParms, SuccessCallback, FailCallback);

                        //獲取TR
                        //<tr[^<>]*?name="SelectTR"[^<>]*?>.*?</tr>
                        //new Regex(@"<tr[^<>]*?name=""SelectTR""[^<>]*?>.*?</tr>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                        mc = Regex.Matches(msg, @"<tr[^<>]*?name=""SelectTR""[^<>]*?>.*?</tr>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        foreach (Match mi in mc)
                        {
                            //獲取td
                            //(?<=<td[^<>]*?>).*?(?=</td>)
                            //new Regex("(?<=<td[^<>]*?>).*?(?=</td>)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                            MatchCollection mic = Regex.Matches(mi.Value, "(?<=<td[^<>]*?>).*?(?=</td>)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                            if (!dicSznyEmployees.ContainsKey(ix))
                            {
                                dicSznyEmployees.Add(ix, new Dictionary<string, string>());
                            }
                            //queueSznyEmployeeInfo.Enqueue(ix);

                            if (!dicSznyEmployees[ix].ContainsKey("UserName"))
                                dicSznyEmployees[ix].Add("UserName", mic[2].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["UserName"] = mic[2].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("PersonID"))
                                dicSznyEmployees[ix].Add("PersonID", mic[3].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["PersonID"] = mic[3].Value.Trim().Replace("&nbsp;", "");

                            if (!dicSznyEmployees[ix].ContainsKey("Birthday"))
                                dicSznyEmployees[ix].Add("Birthday", mic[4].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Birthday"] = mic[4].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("Sex"))
                                dicSznyEmployees[ix].Add("Sex", mic[5].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Sex"] = mic[5].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("HomePhone"))
                                dicSznyEmployees[ix].Add("HomePhone", mic[6].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["HomePhone"] = mic[6].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("TelPhone"))
                                dicSznyEmployees[ix].Add("TelPhone", mic[7].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["TelPhone"] = mic[7].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("Mail"))
                                dicSznyEmployees[ix].Add("Mail", mic[8].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Mail"] = mic[8].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("Address"))
                                dicSznyEmployees[ix].Add("Address", mic[9].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Address"] = mic[9].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("MinZu"))
                                dicSznyEmployees[ix].Add("MinZu", mic[10].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["MinZu"] = mic[10].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("AddressJiGuan"))
                                dicSznyEmployees[ix].Add("AddressJiGuan", mic[11].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["AddressJiGuan"] = mic[11].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("ZhengZhiMianmao"))
                                dicSznyEmployees[ix].Add("ZhengZhiMianmao", mic[12].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["ZhengZhiMianmao"] = mic[12].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("Paiqianshijian"))
                                dicSznyEmployees[ix].Add("Paiqianshijian", mic[13].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Paiqianshijian"] = mic[13].Value.Trim().Replace("&nbsp;", "");
                            if (!dicSznyEmployees[ix].ContainsKey("Remark"))
                                dicSznyEmployees[ix].Add("Remark", mic[14].Value.Trim().Replace("&nbsp;", ""));
                            else
                                dicSznyEmployees[ix]["Remark"] = mic[14].Value.Trim().Replace("&nbsp;", "");

                        }
                        dicParms.Clear();
                        mc = Regex.Matches(msg, @"(?<name>__VIEWSTATE)\|(?<v>[^\|]+)|(?<name>__EVENTVALIDATION)\|(?<v>[^\|]+)", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        foreach (Match mi in mc)
                        {
                            dicParms.Add(mi.Groups["name"].Value.Trim(), mi.Groups["v"].Value.Trim());
                        }

                        dicParms.Add("HiddenField_param", "");
                        dicParms.Add("__EVENTTARGET", "");
                        dicParms.Add("__EVENTARGUMENT", "");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_GridViewID", $"{name}");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentPage", ix.ToString());
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalPage", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iTotalCount", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageSize", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iPageCount", "1");
                        dicParms.Add("MdGridView_t_unitemployees_dwyg_iCurrentNum", "0");
                        dicParms.Add("XM", "ZXMCHECK");
                        lstParms.Clear();
                        lstParms = new List<string>() { "XM", "__EVENTTARGET", "__EVENTARGUMENT", "Button_Query", "Button_SelQuery", };
                        lstParms.Add("Button_edit");
                        lstParms.Add("Button_out");
                        lstParms.Add("ImageButton_Tx");
                        lstParms.Add("ImageButton_xx1");
                        lstParms.Add("Button_qd");

                        mc = Regex.Matches(msg, @"<input[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?value=[""'](?<val>[^'""]*?)[""'][^<>]*?/?>|<input[^<>]*?name=[""'](?<name>[^'""]*?)[""'][^<>]*?[""'][^<>]*?/?>|<select[^<>]*?name=[""'](?<name>[^""']*?)[""'][^<>]*?[""'][^<>]*?/?>", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        foreach (Match mi in mc)
                        {
                            if (lstParms.Contains(mi.Groups["name"].Value.Trim()))
                                continue;
                            if (dicParms.ContainsKey(mi.Groups["name"].Value.Trim()))
                                dicParms[mi.Groups["name"].Value.Trim()] = mi.Groups["val"].Value.Trim();
                            else
                                dicParms.Add(mi.Groups["name"].Value.Trim(), mi.Groups["val"].Value.Trim());
                        }
                        ReqSznyEmployeeInfo(ix, dicParms, SuccessCallback, FailCallback);
                    }
                }
            }
            catch (Exception ex) {
                Business.queueSznyEmployeeInfo.Enqueue(ix);
                Business.queueMsg.Enqueue($"{DateTime.Now.ToString("yyy-MM-dd HH:mm:ss")}{ex.Message}");
            }
        }

獲取工號

　　　　public static void ReqSznyEmployeeInfo(int ix,Dictionary<string,string> dicParms, Action<string> SuccessCallback, Action<string> FailCallback) {
            StringBuilder data = new StringBuilder();
            foreach (var kv in dicParms)
            {
                if (kv.Key.StartsWith("header"))
                    continue;
                data.Append($"&{Common.UrlEncode(kv.Key, Encoding.UTF8)}={ Common.UrlEncode(kv.Value, Encoding.UTF8)}");
            }
            if (data.Length > 0)
                data.Remove(0, 1);
            HttpWebRequest req = WebRequest.Create("//127.0.0.1/HumanResources/EmployeeManage/EmployeeInfoList.aspx") as HttpWebRequest;
            req.Method = "POST";
            req.KeepAlive = true;
            req.CookieContainer = sznyCookie;
            req.ContentType = "application/x-www-form-urlencoded";
            req.Accept = "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9";
            if (dicParms.ContainsKey("ScriptManager1"))
            {
                req.Headers.Add("X-MicrosoftAjax", "Delta=true");
                req.Headers.Add("X-Requested-With", "XMLHttpRequest");
                req.ContentType = "application/x-www-form-urlencoded; charset=UTF-8";
                req.Accept = "*/*";
            }
            req.Headers.Add("Cache-Control", "max-age=0");
            req.UserAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.163 Safari/537.36";
            req.ServicePoint.ConnectionLimit = int.MaxValue;
            req.ServicePoint.Expect100Continue = false;
            req.AllowAutoRedirect = true;
            req.Credentials = System.Net.CredentialCache.DefaultCredentials;
            byte[] buffer = Encoding.UTF8.GetBytes(data.ToString());
            using (Stream reqStream = req.GetRequestStream())
            {
                reqStream.Write(buffer, 0, buffer.Length);
            }
            req.BeginGetResponse(new AsyncCallback(RspSznyEmployeeInfo), new object[] { req,ix, dicParms, SuccessCallback, FailCallback });
        }

        private static void RspSznyEmployeeInfo(IAsyncResult result)
        {
            object[] parms = result.AsyncState as object[];
            HttpWebRequest req = parms[0] as HttpWebRequest;
            int ix =int.Parse( parms[1].ToString());
            Dictionary<string, string> dicParms = parms[2] as Dictionary<string, string>;
            Action<string> SuccessCallback = parms[3] as Action<string>;
            Action<string> FailCallback = parms[4] as Action<string>;
            try
            {
                using (HttpWebResponse rsp = req.EndGetResponse(result) as HttpWebResponse)
                {
                    using (StreamReader reader = new StreamReader(rsp.GetResponseStream()))
                    {
                        string msg = "";
                        msg = reader.ReadToEnd();

                        string code = "無";
                        
                        //<input[^<>]*?name\s*?=\s*?["']TextBox_YG_Code_str["'][^<>]*?value\s*?=\s*?["'](?<code>[^"']*?)["']|<input[^<>]*?value\s*?=\s*?["'](?<code>[^"']*?)["'][^<>]*?name\s*?=\s*?["']TextBox_YG_Code_str["']
                        //new Regex(@"<input[^<>]*?name\s*?=\s*?[""']TextBox_YG_Code_str[""'][^<>]*?value\s*?=\s*?[""'](?<code>[^""']*?)[""']|<input[^<>]*?value\s*?=\s*?[""'](?<code>[^""']*?)[""'][^<>]*?name\s*?=\s*?[""']TextBox_YG_Code_str[""']", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture)
                        Match m = Regex.Match(msg, @"<input[^<>]*?name\s*?=\s*?[""']TextBox_YG_Code_str[""'][^<>]*?value\s*?=\s*?[""'](?<code>[^""']*?)[""']|<input[^<>]*?value\s*?=\s*?[""'](?<code>[^""']*?)[""'][^<>]*?name\s*?=\s*?[""']TextBox_YG_Code_str[""']", RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.ExplicitCapture);
                        if (m.Success)
                            code = m.Groups["code"].Value.Trim();
                        if (dicSznyEmployees[ix].ContainsKey("Code"))
                            dicSznyEmployees[ix]["Code"] = code;
                        else
                            dicSznyEmployees[ix].Add("Code", code);
                        queueSuccessEmployeeInfo.Enqueue(ix);
                    }
                }
            }
            catch (Exception ex) {
                Business.queueSznyEmployeeInfo.Enqueue(ix);
                Business.queueMsg.Enqueue($"{DateTime.Now.ToString("yyy-MM-dd HH:mm:ss")}{ex.Message}");
            }
        }

入庫

採集到的資訊，通過定時任務保存到資料庫。

　　　　　　　　Task.Factory.StartNew(() => {
                  while (true) {
                    if (Business.queueSuccessEmployeeInfo.Count <= 0) {
                        Thread.Sleep(1000);
                        continue;
                    }
                    List<Dictionary<string, string>> lst = new List<Dictionary<string, string>>();
                    while (Business.queueSuccessEmployeeInfo.Count > 0) {
                        Business.queueSuccessEmployeeInfo.TryDequeue(out int ix);
                        lst.Add(Business.dicSznyEmployees[ix]);
                        if (lst.Count >= 50)
                            break;
                    }
                    DbAccess.AddTran(lst, "SznyEmployee",new List<string>() { "UserName", "PersonID" });

                    Thread.Sleep(1);
                }
            });

總結

採集的時候，為了能利用已經採集到的資訊，而不是重複採集，在採集的時候對資料庫數據進行判斷是否存在。純粹是為了提高效率，WebForm的網站真是太慢，太慢了

以前寫非同步純粹是為了提高執行緒效率，在.NET中感覺不到快樂。

終於搞定了，數據已經成功入庫了。

.NET的沒落也是有原因的，網站的速度的確是慢，.net押寶.net core的新體驗了。

我討厭採集WEBForm網站，寫了這麼久的爬蟲，祈禱永遠不要在碰到WEBFORM了。

秀一下結果慶祝一下把

Tags: c# 爬蟲採集