/// <summary>
/// 获取连接信息
/// </summary>
/// <param name="sContent"></param>
/// <param name="sUrl"></param>
/// <param name="lisA"></param>
private static void _GetLinks(string sContent, string sUrl, ref Dictionary <string, string> lisA)
{
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, Const.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);
}
}
}
/// <summary>
/// 获取连接信息
/// </summary>
/// <param name="sContent">网页内容</param>
/// <param name="sUrl">网页url</param>
/// <param name="lisDes">连接列表</param>
/// <returns></returns>
public static Dictionary <string, string> GetLinks(string sContent, string sUrl, ref Dictionary <string, string> lisDes)
{
Dictionary <string, string> lisA = new Dictionary <string, string>();
_GetLinks(sContent, sUrl, ref lisA);
string domain = CRegex.GetDomain(sUrl).ToLower();
//抓取脚本输出的链接
Regex re = new Regex(@"<script[^>]+src\s*=\s*(?:'(?<src>[^']+)'|""(?<src>[^""]+)""|(?<src>[^>\s]+))\s*[^>]*>", RegexOptions.Multiline | RegexOptions.IgnorePatternWhitespace | RegexOptions.IgnoreCase);
MatchCollection mcs = re.Matches(sContent);
//foreach (Match mc in mcs)
for (int i = mcs.Count - 1; i >= 0; i--)
{
Match mc = mcs[i];
string subUrl = CRegex.GetUrl(sUrl, mc.Groups["src"].Value);
if (domain.CompareTo(CRegex.GetDomain(subUrl).ToLower()) != 0)
{
//同一域的才提炼
continue;
}
string subContent = CSocket.GetHtmlByUrl(subUrl);
if (subContent.Length == 0)
{
continue;
}
_GetLinks(subContent, subUrl, ref lisA);
}
if (lisA.Count == 0)
{
return(GetLinksFromRss(sContent, sUrl, ref lisDes));
}
return(lisA);
}