HTMLHelper

using System;
using System.Collections.Generic;
using System.Text;
using System.Net;
using System.IO;
using System.Xml.Serialization;
using System.Text.RegularExpressions;

namespace HttpHelper
{
    public class HttpHelper
    {
        public Stream GetStream(string url, ref CookieContainer cookies)
        {
            Stream strm = null;
            try
            {
                WebRequest req = WebRequest.Create(url);
                HttpWebRequest httpreg = (HttpWebRequest)req;
                httpreg.CookieContainer = cookies;
                httpreg.Method = "GET";
                //httpreg.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                httpreg.UserAgent = "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0; Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1) ; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022)";                WebResponse resp = httpreg.GetResponse();
                strm = resp.GetResponseStream();
            }
            catch (Exception ex)
            {
                return null;
            }
            return strm;
        }

        public string GetHTML(string url)
        {
            string result = null;
            try
            {
                WebRequest req = WebRequest.Create(url);
                HttpWebRequest httpreg = (HttpWebRequest)req;
                //httpreg.CookieContainer = cookies;
                httpreg.Method = "GET";
                httpreg.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                WebResponse resp = httpreg.GetResponse();
                StreamReader reader = new StreamReader(resp.GetResponseStream(), Encoding.GetEncoding(936));
                result = reader.ReadToEnd();
            }
            catch
            {
                return "wrong page";
            }
            return result;
        }

        public string doGet(string url, ref CookieContainer cookies)
        {
            string result = null;
            try
            {
                WebRequest req = WebRequest.Create(url);
                HttpWebRequest httpreg = (HttpWebRequest)req;
                httpreg.CookieContainer = cookies;
                httpreg.Method = "GET";
                httpreg.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                WebResponse resp = httpreg.GetResponse();
                StreamReader reader = new StreamReader(resp.GetResponseStream(), Encoding.GetEncoding(936));
                result = reader.ReadToEnd();
            }
            catch (Exception ex)
            {
                return "";
            }
            return result;
        }


        public Stream PostStream(string action, string data, ref CookieContainer cookies)
        {
            Stream strm = null;
            try
            {
                byte[] bytes = Encoding.UTF8.GetBytes(data);
                WebRequest req = WebRequest.Create(action);
                HttpWebRequest httpreq = (HttpWebRequest)req;
                httpreq.Method = "POST";
                httpreq.ContentType = "application/x-www-form-urlencoded";
                httpreq.Accept = "application/x-shockwave-flash, image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, application/vnd.ms-excel, application/vnd.ms-powerpoint, application/msword, application/x-silverlight, */*";
                httpreq.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                httpreq.ContentLength = bytes.Length;
                httpreq.CookieContainer = cookies;
                strm = httpreq.GetRequestStream();
            }
            catch (Exception ex)
            {
                return null;
            }
            return strm;
        }


        public string doPost(string action, string data, ref CookieContainer cookies)
        {
            string result = null;
            try
            {
                byte[] bytes = Encoding.UTF8.GetBytes(data);
                WebRequest req = WebRequest.Create(action);
                HttpWebRequest httpreq = (HttpWebRequest)req;
                httpreq.Method = "POST";
                httpreq.ContentType = "application/x-www-form-urlencoded";
                httpreq.Accept = "application/x-shockwave-flash, image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, application/vnd.ms-excel, application/vnd.ms-powerpoint, application/msword, application/x-silverlight, */*";
                httpreq.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                httpreq.ContentLength = bytes.Length;
                httpreq.CookieContainer = cookies;
                Stream strm = httpreq.GetRequestStream();
                strm.Write(bytes, 0, bytes.Length);
                strm.Close();
                WebResponse resq = httpreq.GetResponse();
                StreamReader reader = new StreamReader(resq.GetResponseStream(), Encoding.GetEncoding(936));
                result = reader.ReadToEnd();

            }
            catch (Exception ex)
            {
                return "";
            }
            return result;
        }

        public static string doGet(string url)
        {
            string result = null;
            try
            {
                WebRequest req = WebRequest.Create(url);
                HttpWebRequest httpreg = (HttpWebRequest)req;
                httpreg.Method = "GET";
                httpreg.UserAgent = "Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; CIBA)";
                WebResponse resp = httpreg.GetResponse();
                StreamReader reader = new StreamReader(resp.GetResponseStream(), Encoding.GetEncoding("gb2312"));
                result = reader.ReadToEnd();
            }
            catch (Exception ex)
            {
                return "";
            }
            return result;
        }

        /// <summary>    
        /// 返回匹配多個的集合值    
        /// </summary>    
        /// <param name="start">開始html tag</param>    
        /// <param name="end">結束html tag</param>    
        /// <param name="html">html</param>    
        ///    

        public static IList<string> GetHtmls(string start, string end, string html)
        {
            IList<string> list = new List<string>();
            try
            {
                string pattern = string.Format("{0}(?<g>(.|[\r\n])+?){1}", start, end);//匹配URL的模式,並分組    
                MatchCollection mc = Regex.Matches(html, pattern);//滿足pattern的匹配集合    
                if (mc.Count != 0)
                {
                    foreach (Match match in mc)
                    {
                        GroupCollection gc = match.Groups;
                        list.Add(gc["g"].Value);
                    }
                }
            }
            catch { }
            return list;
        }

        public static string GetHtml(string start, string end, string html)
        {
            string ret = string.Empty;
            try
            {
                string pattern = string.Format("{0}(?<g>(.|[\r\n])+?)?{1}", start, end);//匹配URL的模式,並分組    
                ret = Regex.Match(html, pattern).Groups["g"].Value;

            }
            catch { }

            return ret;
        }
        public static string funcGetSrc(string aText)
        {
            string pattern = @"(?is)<img (?:title=(['""]?)?(?<title>[^'""]*?)\1\s*|src=(['""]?)?(?<url>[^'""]*?)\2\s*)*/>";
            Regex r = new Regex(pattern, RegexOptions.IgnoreCase);
            MatchCollection mc = r.Matches(aText);
            string imgname = "";
            foreach (Match m in mc)
            {
                imgname += m.Groups[1].Value;
            }
            return imgname;
        }

        public static List<string> Get_A_Href(string html)
        {
            List<String> links = new List<String>();

            MatchCollection matches = Regex.Matches(html, "<a(?:\\s+.+?)*?\\s+href=\"([^\"]*?)\".+>(.*?)</a>", RegexOptions.IgnoreCase);

            foreach (Match match in matches)
            {
                string s = match.Groups[1].Value;
                links.Add(s);
            }
            return links;
        }

        /// <summary>   
        /// 獲取字符中指定標籤的值   
        /// </summary>   
        /// <param name="str">字符串</param>   
        /// <param name="title">標籤</param>   
        /// <returns>值</returns>   
        public static List<string> GetTitleContent(string str, string title)
        {
            string tmpStr = string.Format("<{0}[^>]*?>(?<Text>[^<]*)</{1}>", title, title); //獲取<title>之間內容   

            MatchCollection TitleMatchs = Regex.Matches(str, tmpStr, RegexOptions.IgnoreCase);
            List<string> list = new List<string>();
            if (TitleMatchs.Count != 0)
            {
                foreach (Match TitleMatch in TitleMatchs)
                {
                    string result = TitleMatch.Groups["Text"].Value;
                    list.Add(result);
                }
            }
            return list;
        }

        /// <summary>   
        /// 獲取字符中指定標籤的值   
        /// </summary>   
        /// <param name="str">字符串</param>   
        /// <param name="title">標籤</param>   
        /// <param name="attrib">屬性名</param>   
        /// <returns>屬性</returns>   
        public static List<string> GetTitleContent(string str, string title, string attrib)
        {

            string tmpStr = string.Format("<{0}[^>]*?{1}=(['\"\"]?)(?<url>[^'\"\"\\s>]+)\\1[^>]*>", title, attrib); //獲取<title>之間內容   

            MatchCollection TitleMatchs = Regex.Matches(str, tmpStr, RegexOptions.IgnoreCase);
            List<string> list = new List<string>();
            if (TitleMatchs.Count != 0)
            {
                foreach (Match TitleMatch in TitleMatchs)
                {
                    string result = TitleMatch.Groups["url"].Value;
                    list.Add(result);
                }
            }
            return list;
        }

        //調用
        private void DoIt()
        {
            HttpHelper helper = new HttpHelper();
            string result = helper.GetHTML("http://www.17173.com");
            result = HttpHelper.GetHtml("<div class=\"header-3-2-2-1\">", "<div class=header-3-3>", result);
            List<string> list_content = HttpHelper.GetTitleContent(result, "a");
            List<string> list_attrib = HttpHelper.GetTitleContent(result, "a", "href");
        }
    }
}