一个实用的C#网页抓取类代码分享
一个实用的C# 网页抓取类 模拟蜘蛛,类中定义了超多的C#采集文章、网页抓取文章的基础技巧,下面分享代码:
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; namespace MyWebPage { public class Link { public string Text { get; set; } public string NavigateUrl { get; set; } } /// <summary> /// 网页类 /// </summary> public class WebPage { #region 私有成员 private Uri m_uri; //url private List<Link> m_links=new List<Link>(); //此网页上的链接 private string m_title=""; //标题 private string m_html=""; //HTML代码 private string m_outstr=""; //网页可输出的纯文本 private bool m_good; //网页是否可用 private int m_pagesize; //网页的大小 private static Dictionary<string, CookieContainer> webcookies = new Dictionary<string, CookieContainer>();//存放所有网页的Cookie #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; } } public string M_html { get { if (m_html == null) { m_html = ""; } return m_html; } } /// <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; } } #endregion /// <summary> /// 从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(@"<a\shref\s*=""(?<URL>[^""]*).*?>(?<title>[^<]*)</a>", RegexOptions.IgnoreCase | RegexOptions.Singleline); regex[1] = new Regex("<[i]*frame[^><]+src=(\"|')?(?<url>([^>\"'\\s)])+)(\"|')?[^>]*>", RegexOptions.IgnoreCase); for (int i = 0; i < 2; i++) { Match match = regex[i].Match(m_html); while (match.Success) { try { string url = HttpUtility.UrlDecode(new Uri(m_uri, match.Groups["URL"].Value).AbsoluteUri); string text = ""; if (i == 0) text = new Regex("(<[^>]+>)|(\\s)|(?)|&|\"", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(match.Groups["text"].Value, ""); Link link = new Link(); link.Text = text; link.NavigateUrl = url; 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("(<[^>]+?>)|?", 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; } #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.NavigateUrl).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> /// 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 = 10000; 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; 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; } //该处视情况而定 有的需要解码 //m_html = HttpUtility.HtmlDecode(new StreamReader(sm, cding).ReadToEnd()); m_html = new StreamReader(sm, cding).ReadToEnd(); } else { //该处视情况而定 有的需要解码 //m_html = HttpUtility.HtmlDecode(new StreamReader(sm, cding).ReadToEnd()); 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) { } } public WebPage(string _url) { string uurl = ""; try { uurl = Uri.UnescapeDataString(_url); _url = uurl; } catch { }; Init(_url); } #endregion } /// <summary> /// 主程序编写作者:大黑 /// (WebPage类编写者另有其人,类Link的定义在他博文的留言区,这是他的博文地址http://blog.csdn.net/yysyangyangyangshan/article/details/6661886) /// 描述:由C#编写的多线程异步抓取网页的网络爬虫控制台程序 /// 功能:目前只能提取网络链接,所用的两个记录文件并不需要很大。网页文本、图片、视频和html代码暂时不能抓取,请见谅。 /// 但需要注意,网页的数目是非常庞大的,如下代码理论上大概可以把整个互联网网页链接都抓下来。 /// 但事实上,由于处理器功能和网络条件(主要是网速)限制,一般的家用电脑最多能胜任12个线程左右的抓取任务,抓取速度有限。可以抓取,但需要时间和耐心。 /// 当然,这个程序把所有链接抓下来是可能的,因为链接占系统空间并不多,而且有记录文件的帮助,已抓取网页的数量可以堆积下去, /// 甚至可以把所有的互联网网络链接都存取下来,当然,最好是分批次。建议设置maxNum为500-1000左右,慢慢累积下去。 /// 另外因为是控制台程序,有时候显示字符过多会系统会暂停显示,这时候只要点击控制台按下回车键就可以了。程序假死的时候,可以按回车键(Enter)试试。 /// 使用本程序,请确保已创建相应的记录文件,出于简化代码的考虑,本程序做的并不健壮,请见谅。 /// 默认的文件创建在E盘根目录“已抓取网址.txt”和“待抓取网址.txt”这两个文本文件中,使用者需要自行创建这两个文件,注意后缀名不要搞错。 /// 这两个文件里面的链接基本都是有效链接,可以单独处理使用。 /// 本爬虫程序的速度如下: /// 10线程最快大概400个链接每分钟 /// 6-8线程最快大概200-300个链接每分钟 /// 2-4线程最快大概150-200个链接每分钟 /// 单线程最快大概70-100个链接每分钟 /// 之所以用多线程异步抓取完全是出于效率考虑,本程序多线程同步并不能带来速度的提升,只要抓取的网页不要太多重复和冗余就可以,异步并不意味着错误。 /// </summary> class Program { //默认的种子网址,可以改动。尽量保持不同的网址,可以提高效率。 //另外注意改的时候“,”符号要注意,别改错了,要英文逗号。 //默认是3线程3个种子网址,最多12线程12个网址,如果要12个以上,可以自己在后面添加。线程设定为12个以上时,必须添加网址,不然会出错 //每个网址对应一个线程,即6线程对应6网址,12线程对应12网址,改默认值的时候要注意parts和数组linkArray的对应关系。 //实测当CPU是i5处理器时,最多运行8-10线程,会伴有严重的卡顿。线程数越高,对CPU要求越高。一般CPU 2-4线程就可以。 static string[] linkArray = {"http://www.163.com/","http://www.sohu.com/","http://www.srcfans.com"}; //设定存放已抓取网址列表的文件路径 static string urlCapturedDir = @"E:\已抓取网址.txt"; //设定存放未抓取网址列表的文件路径 static string myLinkDir = @"E:\待抓取网址.txt"; //同时抓取网址的线程数 private static int parts=4; //目前抓取的网页数量 private static int count=0; //设定的最多抓取网页的数量 private static int maxNum=2000; //待抓取的网址链表 private static List<string> myLink=new List<string>(); //已抓取的网址链表 private static List<string> urlCaptured=new List<string>(); //用于临时存放待抓取的某一个网址 private static string strLink = ""; public Program(string link) { myLink.Add(link); } public void doWork() { Console.WriteLine("working..."); while ((myLink.Count != 0) && (count < maxNum)) { strLink = myLink[0]; if (!urlCaptured.Contains(strLink)) { //加入已抓取队列 urlCaptured.Add(strLink); //当前抓取的网页数量 count++; //通过WebPage类抓取 WebPage tempWeb = new WebPage(strLink); Console.WriteLine("网址" + strLink + "共有" + tempWeb.Links.Count + "个链接"); foreach (Link li in tempWeb.Links) { //确保不重复抓取 if (!myLink.Contains(li.NavigateUrl)) myLink.Add(li.NavigateUrl); } Console.WriteLine("待抓取网页的数目是:" + myLink.Count); //确保不重复抓取,此句多线程运行会报错 while(myLink.Remove(strLink)); } else { //删除重复网址 while (myLink.Remove(strLink)) ; } Console.WriteLine("目前共抓取了" + count + "个网页"); } } static void printLinks(List<string> links) { foreach(string s in links) { Console.WriteLine(s); } } static void Main(string[] args) { //string seedUrl = "http://www.srcfans.com/"; //WebPage webInfo = new WebPage(seedUrl); //Console.WriteLine( webInfo.Context);//不包含html标签的所有内容 //Console.WriteLine( webInfo.M_html);//包含html标签的内容 //urlCaptured.Add(seedUrl); //读取已处理网址列表(urlCaptured) string[] lines = System.IO.File.ReadAllLines(urlCapturedDir); int urlCapturedIndex = 0; foreach (string line in lines) { //if (!urlCaptured.Contains(line)) { urlCaptured.Add(line); urlCapturedIndex++; if(urlCapturedIndex%2000==0) { Console.WriteLine("已导入" + urlCapturedIndex + "项已抓取网址"); } } } // 显示已抓取网址列表 Console.WriteLine("已读取已处理网址列表,共"+urlCaptured.Count+"条信息:"); //foreach (string line in urlCaptured) //{ // Console.WriteLine(line); //} //Console.WriteLine("已处理网址列表显示完毕"); //未抓取网址初始化 string[] lines2 = System.IO.File.ReadAllLines(myLinkDir); int myLinkIndex = 0; foreach (string line in lines2) { //Contains方法太耗时 //if (!myLink.Contains(line)) { myLink.Add(line); myLinkIndex++; if(myLinkIndex%2000==0) { Console.WriteLine("已导入" + myLinkIndex + "项待抓取网址"); } } } // 显示未抓取网址列表 Console.WriteLine("已读取未抓取网址列表,共" + myLink.Count + "条信息:"); //foreach (string line in myLink) //{ // Console.WriteLine(line); //} //Console.WriteLine("未抓取网址列表显示完毕"); System.Threading.Thread[] threads = new System.Threading.Thread[parts]; for (int i = 0; i < parts; i++) { string str_l = linkArray[i]; Program p = new Program(str_l); System.Threading.Thread t = new System.Threading.Thread(new System.Threading.ThreadStart(p.doWork)); threads[i] = t; threads[i].Name = "Thread" + i.ToString(); } for (int i = 0; i < parts; i++) { threads[i].Start(); } //为了确保 Main 函数不会尝试在辅助线程有机会执行结束之前将它终止 //Main 函数将一直循环,直到辅助线程对象的 IsAlive 属性设置为 false //必须用while,表示线程都已经执行完了 for (int i = 0; i < parts; i++) { while (threads[i].IsAlive) { ; } } Console.WriteLine("一共抓取了" + urlCaptured.Count + "个网址"); Console.WriteLine("待处理的网址共有:" + myLink.Count + "个"); Console.WriteLine("检查重复项...."); List<string> temp_l = new List<string>(); foreach (string str in urlCaptured) { if (!temp_l.Contains(str)) { temp_l.Add(str); } else { Console.WriteLine("网址" + str + "重复"); } } Console.WriteLine("共有" + (urlCaptured.Count - temp_l.Count) + "项是重复网址"); urlCaptured = temp_l; Console.WriteLine("已抓取网址写入文件..."); //对已抓取的网址进行排序 urlCaptured.Sort(); using (System.IO.StreamWriter file = new System.IO.StreamWriter(urlCapturedDir)) { foreach(string s in urlCaptured) { file.WriteLine(s); } } Console.WriteLine("已抓取网址写入完成"); Console.WriteLine("待处理网址写入文件..."); myLink.Sort(); using (System.IO.StreamWriter file = new System.IO.StreamWriter(myLinkDir)) { foreach (string s in myLink) { file.WriteLine(s); } } Console.WriteLine("待抓取网址写入完成"); //下面的句子表示保持显示当前控制台命令行,不然控制台会一闪而过 Console.WriteLine("按任意键退出..."); Console.ReadKey(); } } }