C#抓取和分析网页的类

http://www.cnblogs.com/justwannaloveyou/archive/2010/10/24/1859645.html

抓取和分析网页的类。

主要功能有：

1、提取网页的纯文本，去所有html标签和javascript代码

2、提取网页的链接，包括href和frame及iframe

3、提取网页的title等(其它的标签可依此类推，正则是一样的)

4、可以实现简单的表单提交及cookie保存

/*
*  Author:Sunjoy at CCNU
*  如果您改进了这个类请发一份代码给我(ccnusjy 在gmail.com)
*/

using System;
using System.Data;
using System.Configuration;
using System.Net;
using System.IO;
using System.Text;
using System.Collections.Generic;
using System.Text.RegularExpressions;
using System.Threading;
using System.Web;
/// <summary>
/// 网页类
/// </summary>
public class WebPage
{

    #region 私有成员
    private Uri m_uri;   //网址
    private List<Link> m_links;    //此网页上的链接
    private string m_title;        //此网页的标题
    private string m_html;         //此网页的HTML代码
    private string m_outstr;       //此网页可输出的纯文本
    private bool m_good;           //此网页是否可用
    private int m_pagesize;       //此网页的大小
    private static Dictionary<string, CookieContainer> webcookies = new Dictionary<string, CookieContainer>();//存放所有网页的Cookie
    private string m_post;  //此网页的登陆页需要的POST数据
    private string m_loginurl;  //此网页的登陆页
    #endregion


    #region 私有方法
    /// <summary>
    /// 这私有方法从网页的HTML代码中分析出链接信息
    /// </summary>
    /// <returns>List<Link></returns>
    private List<Link> getLinks()
    {
        if (m_links.Count == 0)
        {
            Regex[] regex = new Regex[2];
            regex[0] = new Regex("(?m)<a[^><]+href=(\"|')?(?<url>([^>\"'\\s)])+)(\"|')?[^>]*>(?<text>(\\w|\\W)*?)</", RegexOptions.Multiline | RegexOptions.IgnoreCase);
            regex[1] = new Regex("<[i]*frame[^><]+src=(\"|')?(?<url>([^>\"'\\s)])+)(\"|')?[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase);
            for (int i = 0; i < 2; i++)
            {
                Match match = regex[i].Match(m_html);
                while (match.Success)
                {
                    try
                    {
                        string url = new Uri(m_uri, match.Groups["url"].Value).AbsoluteUri;
                        string text = "";
                        if (i == 0) text = new Regex("(<[^>]+>)|(\\s)|(&nbsp;)|&|\"", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(match.Groups["text"].Value, "");
                        Link link = new Link(url, text);
                        m_links.Add(link);
                    }
                    catch(Exception ex){Console.WriteLine(ex.Message); };
                    match = match.NextMatch();
                }
            }
        }
        return m_links;
    }
   
    /// <summary>
    /// 此私有方法从一段HTML文本中提取出一定字数的纯文本
    /// </summary>
    /// <param name="instr">HTML代码</param>
    /// <param name="firstN">提取从头数多少个字</param>
    /// <param name="withLink">是否要链接里面的字</param>
    /// <returns>纯文本</returns>
    private string getFirstNchar(string instr, int firstN, bool withLink)
    {
        if (m_outstr == "")
        {
            m_outstr = instr.Clone() as string;
            m_outstr = new Regex(@"(?m)<script[^>]*>(\w|\W)*?</script[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase ).Replace(m_outstr, "");
            m_outstr = new Regex(@"(?m)<style[^>]*>(\w|\W)*?</style[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase ).Replace(m_outstr, "");
            m_outstr = new Regex(@"(?m)<select[^>]*>(\w|\W)*?</select[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase ).Replace(m_outstr, "");
            if (!withLink) m_outstr = new Regex(@"(?m)<a[^>]*>(\w|\W)*?</a[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(m_outstr, "");
            Regex objReg = new System.Text.RegularExpressions.Regex("(<[^>]+?>)|&nbsp;", RegexOptions.Multiline | RegexOptions.IgnoreCase);
            m_outstr = objReg.Replace(m_outstr, "");
            Regex objReg2 = new System.Text.RegularExpressions.Regex("(\\s)+", RegexOptions.Multiline | RegexOptions.IgnoreCase);
            m_outstr = objReg2.Replace(m_outstr, " ");
        }
        return m_outstr.Length > firstN ? m_outstr.Substring(0, firstN) : m_outstr;
    }

    /// <summary>
    /// 此私有方法返回一个IP地址对应的无符号整数
    /// </summary>
    /// <param name="x">IP地址</param>
    /// <returns></returns>
    private uint getuintFromIP(IPAddress x)
    {
        Byte[] bt = x.GetAddressBytes();
        uint i = (uint)(bt[0] * 256 * 256 * 256);
        i += (uint)(bt[1] * 256 * 256);
        i += (uint)(bt[2] * 256);
        i += (uint)(bt[3]);
        return i;
    }

    #endregion


    #region 公有文法
    /// <summary>
    /// 此公有方法提取网页中一定字数的纯文本，包括链接文字
    /// </summary>
    /// <param name="firstN">字数</param>
    /// <returns></returns>
    public string getContext(int firstN)
    {
        return getFirstNchar(m_html, firstN, true);
    }

    /// <summary>
    /// 此公有方法提取网页中一定字数的纯文本，不包括链接文字
    /// </summary>
    /// <param name="firstN"></param>
    /// <returns></returns>
    public string getContextWithOutLink(int firstN)
    {
        return getFirstNchar(m_html, firstN, false);
    }

    /// <summary>
    /// 此公有方法从本网页的链接中提取一定数量的链接，该链接的URL满足某正则式
    /// </summary>
    /// <param name="pattern">正则式</param>
    /// <param name="count">返回的链接的个数</param>
    /// <returns>List<Link></returns>
    public List<Link> getSpecialLinksByUrl(string pattern,int count)
    {
        if(m_links.Count==0)getLinks();
        List<Link> SpecialLinks = new List<Link>();
        List<Link>.Enumerator i;
        i = m_links.GetEnumerator();
        int cnt = 0;
        while (i.MoveNext() && cnt<count)
        {
            if (new Regex(pattern, RegexOptions.Multiline | RegexOptions.IgnoreCase ).Match(i.Current.url).Success)
            {
                SpecialLinks.Add(i.Current);
                cnt++;
            }
        } 
        return SpecialLinks;
    }



    /// <summary>
    /// 此公有方法从本网页的链接中提取一定数量的链接，该链接的文字满足某正则式
    /// </summary>
    /// <param name="pattern">正则式</param>
    /// <param name="count">返回的链接的个数</param>
    /// <returns>List<Link></returns>
    public List<Link> getSpecialLinksByText(string pattern,int count)
    {
        if (m_links.Count == 0) getLinks();
        List<Link> SpecialLinks = new List<Link>();
        List<Link>.Enumerator i;
        i = m_links.GetEnumerator();
        int cnt = 0;
        while (i.MoveNext() && cnt < count)
        {
            if (new Regex(pattern, RegexOptions.Multiline | RegexOptions.IgnoreCase ).Match(i.Current.text).Success)
            {
                SpecialLinks.Add(i.Current);
                cnt++;
            }
        }
        return SpecialLinks;
    }
    /// <summary>
    /// 此公有方法获得所有链接中在一定IP范围的链接
    /// </summary>
    /// <param name="_ip_start">起始IP</param>
    /// <param name="_ip_end">终止IP</param>
    /// <returns></returns>
    public List<Link> getSpecialLinksByIP(string _ip_start, string _ip_end)
    {
        IPAddress ip_start = IPAddress.Parse(_ip_start);
        IPAddress ip_end = IPAddress.Parse(_ip_end);
        if (m_links.Count == 0) getLinks();
        List<Link> SpecialLinks = new List<Link>();
        List<Link>.Enumerator i;
        i = m_links.GetEnumerator();
        while (i.MoveNext())
        {
            IPAddress ip;
            try
            {
                ip = Dns.GetHostEntry(new Uri(i.Current.url).Host).AddressList[0];
            }
            catch { continue; }
            if(getuintFromIP(ip)>=getuintFromIP(ip_start) && getuintFromIP(ip)<=getuintFromIP(ip_end))
            {
                SpecialLinks.Add(i.Current);
            }
        }
        return SpecialLinks;
    }

    /// <summary>
    /// 这公有方法提取本网页的纯文本中满足某正则式的文字
    /// </summary>
    /// <param name="pattern">正则式</param>
    /// <returns>返回文字</returns>
    public string getSpecialWords(string pattern)
    {
        if (m_outstr == "") getContext(Int16.MaxValue);
        Regex regex = new Regex(pattern, RegexOptions.Multiline | RegexOptions.IgnoreCase );
        Match mc=regex.Match(m_outstr);
        if (mc.Success)
            return mc.Groups[1].Value;
        return string.Empty;
    }
    #endregion


    #region 构造函数
    
    private void Init(string _url)
    {
   
        try
        {
            m_uri = new Uri(_url);
            m_links = new List<Link>();
            m_html = "";
            m_outstr = "";
            m_title = "";
            m_good = true;
            if (_url.EndsWith(".rar") || _url.EndsWith(".dat") || _url.EndsWith(".msi"))
            {
                m_good = false;
                return;
            }
            HttpWebRequest rqst = (HttpWebRequest)WebRequest.Create(m_uri);
            rqst.AllowAutoRedirect = true;
            rqst.MaximumAutomaticRedirections = 3;
            rqst.UserAgent = "Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)";
            rqst.KeepAlive = true;
            rqst.Timeout = 30000;
            lock (WebPage.webcookies)
            {
                if (WebPage.webcookies.ContainsKey(m_uri.Host))
                    rqst.CookieContainer = WebPage.webcookies[m_uri.Host];
                else
                {
                    CookieContainer cc = new CookieContainer();
                    WebPage.webcookies[m_uri.Host] = cc;
                    rqst.CookieContainer = cc;
                }
            }

            HttpWebResponse rsps = (HttpWebResponse)rqst.GetResponse();

            Stream sm = rsps.GetResponseStream();
            if (!rsps.ContentType.ToLower().StartsWith("text/") || rsps.ContentLength > 1 << 22)
            {
                rsps.Close();
                m_good = false;
                return;
            }
            Encoding cding = System.Text.Encoding.Default;

　　　　cding = System.Text.Encoding.GetEncoding("utf-8");  //  解决c#__StreamReader读入中文乱码
            string contenttype=rsps.ContentType.ToLower();
            int ix = contenttype.IndexOf("charset=");
            if (ix != -1)
            {

                try
                {
                    cding = System.Text.Encoding.GetEncoding(rsps.ContentType.Substring(ix + "charset".Length + 1));
                }
                catch
                {
                    cding = Encoding.Default;
                }
                m_html = new StreamReader(sm, cding).ReadToEnd();
            }
            else
            {
                m_html = new StreamReader(sm, cding).ReadToEnd();
                Regex regex = new Regex("charset=(?<cding>[^=]+)?\"",RegexOptions.IgnoreCase);
                string strcding = regex.Match(m_html).Groups["cding"].Value;
                try
                {
                    cding = Encoding.GetEncoding(strcding);
                }
                catch{
                    cding = Encoding.Default;
                }
                byte[] bytes=Encoding.Default.GetBytes(m_html.ToCharArray());
                m_html = cding.GetString(bytes);
                if (m_html.Split('?').Length > 100)
                {
                    m_html=Encoding.Default.GetString(bytes);
                }
            }

            
            m_pagesize = m_html.Length;
            m_uri = rsps.ResponseUri;
            rsps.Close();
        }
        catch (Exception ex)
        {
            Console.WriteLine(ex.Message+m_uri.ToString());
            m_good = false;
            
        }
    }

    public WebPage(string _url)
    {
        string uurl = "";
        try
        {
            uurl = Uri.UnescapeDataString(_url);
            _url = uurl;
        }
        catch { };
        Regex re = new Regex("(?<h>[^\x00-\xff]+)");
        Match mc = re.Match(_url);
        if (mc.Success)
        {
            string han = mc.Groups["h"].Value;
            _url = _url.Replace(han, System.Web.HttpUtility.UrlEncode(han, Encoding.GetEncoding("GB2312")));
        }

        Init(_url);
    }

    public WebPage(string _url, string _loginurl, string _post)
    {
        string uurl = "";
        try
        {
            uurl = Uri.UnescapeDataString(_url);
            _url = uurl;
        }
        catch { };
        Regex re = new Regex("(?<h>[^\x00-\xff]+)");
        Match mc = re.Match(_url);
        if (mc.Success)
        {
            string han = mc.Groups["h"].Value;
            _url = _url.Replace(han, System.Web.HttpUtility.UrlEncode(han, Encoding.GetEncoding("GB2312")));
        }
        if (_loginurl.Trim() == "" || _post.Trim() == "" || WebPage.webcookies.ContainsKey(new Uri(_url).Host))
        {
            Init(_url);
        }
        else
        {
            #region 登陆
            string indata = _post;
            m_post = _post;
            m_loginurl = _loginurl;
            byte[] bytes = Encoding.Default.GetBytes(_post);
            CookieContainer myCookieContainer = new CookieContainer();
            try
            {

                //新建一个CookieContainer来存放Cookie集合 

                HttpWebRequest myHttpWebRequest = (HttpWebRequest)WebRequest.Create(_loginurl);
                //新建一个HttpWebRequest 
                myHttpWebRequest.ContentType = "application/x-www-form-urlencoded";
                myHttpWebRequest.AllowAutoRedirect = false;
                myHttpWebRequest.UserAgent = "Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)";
                myHttpWebRequest.Timeout = 60000;
                myHttpWebRequest.KeepAlive = true;
                myHttpWebRequest.ContentLength = bytes.Length;
                myHttpWebRequest.Method = "POST";
                myHttpWebRequest.CookieContainer = myCookieContainer;
                //设置HttpWebRequest的CookieContainer为刚才建立的那个myCookieContainer 
                Stream myRequestStream = myHttpWebRequest.GetRequestStream();
                myRequestStream.Write(bytes, 0, bytes.Length);
                myRequestStream.Close();
                HttpWebResponse myHttpWebResponse = (HttpWebResponse)myHttpWebRequest.GetResponse();

                foreach (Cookie ck in myHttpWebResponse.Cookies)
                {
                    myCookieContainer.Add(ck);
                }
                myHttpWebResponse.Close();
            }
            catch
            {
                Init(_url);
                return;
            }

            #endregion

            #region 登陆后再访问页面
            try
            {
                m_uri = new Uri(_url);
                m_links = new List<Link>();
                m_html = "";
                m_outstr = "";
                m_title = "";
                m_good = true;
                if (_url.EndsWith(".rar") || _url.EndsWith(".dat") || _url.EndsWith(".msi"))
                {
                    m_good = false;
                    return;
                }
                HttpWebRequest rqst = (HttpWebRequest)WebRequest.Create(m_uri);
                rqst.AllowAutoRedirect = true;
                rqst.MaximumAutomaticRedirections = 3;
                rqst.UserAgent = "Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)";
                rqst.KeepAlive = true;
                rqst.Timeout = 30000;
                rqst.CookieContainer = myCookieContainer;
                lock (WebPage.webcookies)
                {
                    WebPage.webcookies[m_uri.Host] = myCookieContainer;
                }
                HttpWebResponse rsps = (HttpWebResponse)rqst.GetResponse();

                Stream sm = rsps.GetResponseStream();
                if (!rsps.ContentType.ToLower().StartsWith("text/") || rsps.ContentLength > 1 << 22)
                {
                    rsps.Close();
                    m_good = false;
                    return;
                }
                Encoding cding = System.Text.Encoding.Default;
                int ix = rsps.ContentType.ToLower().IndexOf("charset=");
                if (ix != -1)
                {
                    try
                    {
                        cding = System.Text.Encoding.GetEncoding(rsps.ContentType.Substring(ix + "charset".Length + 1));
                    }
                    catch
                    {
                        cding = Encoding.Default;
                    }
                }


                m_html = new StreamReader(sm, cding).ReadToEnd();


                m_pagesize = m_html.Length;
                m_uri = rsps.ResponseUri;
                rsps.Close();
            }
            catch (Exception ex)
            {
                Console.WriteLine(ex.Message+m_uri.ToString());
                m_good = false;
            
            }
            #endregion
        }

    }

    #endregion


    #region 属性

    /// <summary>
    /// 通过此属性可获得本网页的网址，只读
    /// </summary>
    public string URL
    {
        get
        {
            return m_uri.AbsoluteUri;
        }
    }

    /// <summary>
    /// 通过此属性可获得本网页的标题，只读
    /// </summary>
    public string Title
    {
        get
        {
            if (m_title == "")
            {
                Regex reg = new Regex(@"(?m)<title[^>]*>(?<title>(?:\w|\W)*?)</title[^>]*>", RegexOptions.Multiline | RegexOptions.IgnoreCase );
                Match mc = reg.Match(m_html);
                if (mc.Success)
                    m_title= mc.Groups["title"].Value.Trim();
            }
            return m_title;
        }
    }
  

    /// <summary>
    /// 此属性获得本网页的所有链接信息，只读
    /// </summary>
    public List<Link> Links
    {
        get
        {
            if (m_links.Count == 0) getLinks();
            return m_links;
        }
    }


    /// <summary>
    /// 此属性返回本网页的全部纯文本信息，只读
    /// </summary>
    public string Context
    {
       get
       {
           if (m_outstr == "") getContext(Int16.MaxValue);
           return m_outstr;
       }
    }

    /// <summary>
    /// 此属性获得本网页的大小
    /// </summary>
    public int PageSize
    {
        get
        {
            return m_pagesize;
        }
    }
    /// <summary>
    /// 此属性获得本网页的所有站内链接
    /// </summary>
    public List<Link> InsiteLinks
    {
        get
        {
            return getSpecialLinksByUrl("^http://"+m_uri.Host,Int16.MaxValue);
        }
    }

    /// <summary>
    /// 此属性表示本网页是否可用
    /// </summary>
    public bool IsGood
    {
        get
        {
            return m_good;
        }
    }
    /// <summary>
    /// 此属性表示网页的所在的网站
    /// </summary>
    public string Host
    {
        get
        {
            return m_uri.Host;
        }
    }
    

    /// <summary>
    /// 此网页的登陆页所需的POST数据
    /// </summary>
    public string PostStr
    {
        get
        {
            return m_post;
        }
    }
    /// <summary>
    /// 此网页的登陆页
    /// </summary>
    public string LoginURL
    {
        get
        {
            return m_loginurl;
        }
    }
    #endregion
}

/// <summary>
/// 链接类
/// </summary>
public class Link
{
    public string url;   //链接网址
    public string text;  //链接文字
    public Link(string _url, string _text)
    {
        url = _url;
        text = _text;
    }
}

欢迎转载，转载请注明：转载自陈善波 [ http://justwannaloveyou.cnblogs.com/]