/// <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>
/// 作者名
/// </summary>
/// <param name="sInput">输入内容</param>
/// <param name="sRegex">表达式字符串</param>
public static string GetAuthor(string sInput, string sRegex)
{
string sAuthor = GetText(sInput, sRegex, "Author");
sAuthor = CString.ClearTag(sAuthor);
if (sAuthor.Length > 99)
{
sAuthor = sAuthor.Substring(0, 99);
}
return(sAuthor);
}
/// <summary>
/// 文章来源
/// </summary>
/// <param name="sInput">输入内容</param>
/// <param name="sRegex">表达式字符串</param>
public static string GetSource(string sInput, string sRegex)
{
string sSource = GetText(sInput, sRegex, "Source");
sSource = CString.ClearTag(sSource);
if (sSource.Length > 99)
{
sSource = sSource.Substring(0, 99);
}
return(sSource);
}
/// <summary>
/// 文章标题
/// </summary>
/// <param name="sInput">输入内容</param>
/// <param name="sRegex">表达式字符串</param>
public static string GetTitle(string sInput, string sRegex)
{
string sTitle = GetText(sInput, sRegex, "Title");
sTitle = CString.ClearTag(sTitle);
if (sTitle.Length > 99)
{
sTitle = sTitle.Substring(0, 99);
}
return(sTitle);
}
/// <summary>
/// 通过正则表达式获取内容
/// </summary>
/// <param name="sContent">待查找的内容</param>
/// <param name="sRegex">正则表达式</param>
/// <returns></returns>
public static string GetTextByReg(string sContent, string sRegex)
{
Regex re = new Regex(sRegex, RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace | RegexOptions.Multiline);
Match mc = re.Match(sContent);
string str = "";
if (mc.Success)
{
str = mc.Groups[0].Value;
}
while (str.EndsWith("_"))
{
str = CString.RemoveEndWith(str, "_");
}
return(str);
}
/// <summary>
/// 根据相对路径得到绝对路径
/// </summary>
/// <param name="sInput">输入内容</param>
/// <param name="sRelativeUrl">相对链接地址</param>
public static string GetUrl(string sInput, string sRelativeUrl)
{
string sReturnUrl = "";
string sUrl = _GetStandardUrlDepth(sInput);//返回了http://www.163.com/news/这种形式
if (sRelativeUrl.ToLower().StartsWith("http") || sRelativeUrl.ToLower().StartsWith("https"))
{
sReturnUrl = sRelativeUrl.Trim();
}
else if (sRelativeUrl.StartsWith("/"))
{
sReturnUrl = GetDomain(sInput) + sRelativeUrl;
}
else if (sRelativeUrl.StartsWith("../"))
{
sUrl = sUrl.Substring(0, sUrl.Length - 1);
while (sRelativeUrl.IndexOf("../") >= 0)
{
string temp = CString.GetPreStrByLast(sUrl, "/");
if (temp.Length > 6)
{//temp != "http:/"，否则的话，说明已经回溯到尽头了，"../"与网址的层次对应不上。存在这种情况，网页上面的链接是错误的，但浏览器还能正常显示
sUrl = temp;
}
sRelativeUrl = sRelativeUrl.Substring(3);
}
sReturnUrl = sUrl + "/" + sRelativeUrl.Trim();
}
else if (sRelativeUrl.StartsWith("./"))
{
sReturnUrl = sUrl + sRelativeUrl.Trim().Substring(2);
}
else if (sRelativeUrl.Trim() != "")
{//2007images/modecss.css
sReturnUrl = sUrl + sRelativeUrl.Trim();
}
else
{
sRelativeUrl = sUrl;
}
return(sReturnUrl);
}
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);
}