public static List <string> GetLinksByKey(string sContent, /*string sUrl,*/ List <string> listKey)
{
List <string> listResult = new List <string>();
List <string> list = new List <string>();
string sKey = "";
string strKey;
//提取链接
Regex re = new Regex(@"<a\s+[^>]*href\s*=\s*[^>]+>[\s\S]*?</a>", RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
MatchCollection mcs = re.Matches(sContent);
foreach (Match mc in mcs)
{
strKey = RemoveByReg(GetLink(mc.Value), @"^http.*/$");//屏蔽以“http”开头“/”结尾的链接地址
if (strKey.Length > 0)
{
list.Add(mc.Value);
}
}
//准备好关键字
foreach (string s in listKey)
{
sKey += "([\\s\\S]*" + s + "[\\s\\S]*)|";
}
if (sKey != "")
{
sKey = sKey.Substring(0, sKey.Length - 1);
}
if (sKey == "")
{
sKey = "[\\s\\S]+";
}
Regex reKey = new Regex(sKey, RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
Match tmpmc;
//链接的文字一定要5个字以上才算有效？
re = new Regex(@"<a\s+[^>]+>([\s\S]{5,})?</a>", RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
foreach (string s in list)
{
tmpmc = re.Match(s);
if (tmpmc.Success)
{
strKey = CString.ClearTag(tmpmc.Groups[1].Value.Trim());
strKey = RemoveByReg(strKey, @"更多|登录|添加|推荐|收藏夹|加盟|关于|订阅|阅读器|我的|有限|免费|公司|more|RSS|about|\.");
if (CString.GetLength(strKey) > 8)//最起码是5个是为了屏蔽垃圾信息。
{
if (reKey.Match(strKey).Success)
{
listResult.Add(s);
}
}
}
}
#region 对RSS的支持
if (listResult.Count == 0)
{
return(GetLinksByKeyFromRss(sContent, listKey));
}
#endregion
return(listResult);
}
private static void _GetLinks(string sContent, string sUrl, ref Dictionary <string, string> lisA)
{
const string sFilter =
@"首页|下载|中文|English|反馈|讨论区|投诉|建议|联系|关于|about|诚邀|工作|简介|新闻|掠影|风采
|登录|注销|注册|使用|体验|立即|收藏夹|收藏|添加|加入
|更多|more|专题|精选|热卖|热销|推荐|精彩
|加盟|联盟|友情|链接|相关
|订阅|阅读器|RSS
|免责|条款|声明|我的|我们|组织|概况|有限|免费|公司|法律|导航|广告|地图|隐私
|〖|〗|【|】|（|）|［|］|『|』|\.";
Regex re = new Regex(@"<a\s+[^>]*href\s*=\s*[^>]+>[\s\S]*?</a>", RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
Regex re2 = new Regex(@"""|'", RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
MatchCollection mcs = re.Matches(sContent);
//foreach (Match mc in mcs)
for (int i = mcs.Count - 1; i >= 0; i--)
{
Match mc = mcs[i];
string strHref = GetLink(mc.Value).Trim();
strHref = strHref.Replace("\\\"", "");//针对JS输出链接
strHref = strHref.Replace("\\\'", "");
string strTemp = RemoveByReg(strHref, @"^http.*/$");//屏蔽以“http”开头“/”结尾的链接地址
if (strTemp.Length < 2)
{
continue;
}
//过滤广告或无意义的链接
string strText = CString.ClearTag(GetTextByLink(mc.Value)).Trim();
strTemp = RemoveByReg(strText, sFilter);
if (CString.GetLength(strTemp) < 9)
{
continue;
}
if (re2.IsMatch(strText))
{
continue;
}
//换上绝对地址
strHref = CText.GetUrlByRelative(sUrl, strHref);
if (strHref.Length <= 18)//例如，http://www.163.com = 18
{
continue;
}
//计算#字符出现的位置，移除它后面的内容
//如果是域名地址，就跳过
int charIndex = strHref.IndexOf('#');
if (charIndex > -1)
{
strHref = strHref.Substring(0, charIndex);
}
strHref = strHref.Trim(new char[] { '/', '\\' });
string tmpDomainURL = CRegex.GetDomain(strHref);
if (strHref.Equals(tmpDomainURL, StringComparison.OrdinalIgnoreCase))
{
continue;
}
if (!lisA.ContainsKey(strHref) && !lisA.ContainsValue(strText))
{
lisA.Add(strHref, strText);
}
}
}