1、抓取远程网页源码，这里要实现自动判断网页编码，否则有可能抓到乱码。我是先看应答的 http头的chareset，一般这个很准，但像csdn的新闻比较变态http应答的头里的chareset和网页的meta里声明的 chareset不一致，所以我手工加了一下判断，如果不一致再在内存流里用网页声明的编码读取一遍源码
2、把网页分割成几大块。试用了一下tidy的.net包装及HtmlParse的.net版本，都不太好用。于是我自己写了个算法，可以把网页里的div块，td块等都提取出来，支持嵌套的情况。一般只提取div的文字块儿就行了。
3、把汉字少于200的文本块去了，一般少于200字的文本块不会是正文，即便是正文，一般来说也不会有太多的价值，我直接去掉。
4、 因为div支持嵌套，所以剩下的文本块，有可能是重复的，一个是另一个的父节点，所以要把最里层的文本块找出来，最里层的文本块肯定是汉字最多的，而其它 文本最少的，所以要计算出剩余文本块中汉字占所有字符比例最高的文本块，基本上它就是正文的文本块了。当然有的网页正文里也可能还有div的文本块，这时 候可能会判断错误，但只要正文嵌套的Div文本块的汉字少于200字，我的算法还是能准确提取正文文本块的。这一步我用写了一个自定义的方法传递给 List的Sort方法。
5、把<p><br>等标签替换成特殊占位符[p][br]等，因为最终的正文需要保留段落和回车换行等格式。这一步用正则实现。
6、把最后剩下的文本块的html标签去掉，我用正则过滤的。
7、把[p]替换成回车换行加俩空格，把[br]替换成回车换行，这步也用正则。到此，正文提取完毕
主要代码：
public class GetMainContentHelper
{
///<summary>
/// 判断两段儿文本里哪个中文占的比例高
///</summary>
///<param name="x"></param>
///<param name="y"></param>
///<returns></returns>
public static int CompareDinosByChineseLength(string x, string y)
{
if (x == null)
{
if (y == null)
{
return 0;
}
else
{
return -1;
}
}
else
{
if (y == null)
{
return 1;
}
else
{
Regex r = new Regex("[\u4e00-\u9fa5]");
float xCount = (float)(r.Matches(x).Count) / (float)x.Length;
float yCount = (float)(r.Matches(y).Count) / (float)y.Length;
int retval = xCount.CompareTo(yCount);
if (retval != 0)
{
return retval;
}
else
{
return x.CompareTo(y);
}
}
}
}
///<summary>
/// 获取一个网页源码中的标签列表，支持嵌套，一般或去div，td等容器
///</summary>
///<param name="input"></param>
///<param name="tag"></param>
///<returns></returns>
public static List<string> GetTags(string input, string tag)
{
StringReader strReader = new StringReader(input);
int lowerThanCharCounter = 0;
int lowerThanCharPos = 0;
Stack<int> tagPos = new Stack<int>();
List<string> taglist = new List<string>();
int i = 0;
while (true)
{
try
{
int intCharacter = strReader.Read();
if (intCharacter == -1) break;
char convertedCharacter = Convert.ToChar(intCharacter);
if (lowerThanCharCounter > 0)
{
if (convertedCharacter == ‘>‘)
{
lowerThanCharCounter--;
string biaoqian = input.Substring(lowerThanCharPos, i - lowerThanCharPos + 1);
if (biaoqian.StartsWith(string.Format("<{0}", tag)))
{
tagPos.Push(lowerThanCharPos);
}
if (biaoqian.StartsWith(string.Format("</{0}", tag)))
{
if (tagPos.Count < 1)
continue;
int tempTagPos = tagPos.Pop();
string strdiv = input.Substring(tempTagPos, i - tempTagPos + 1);
taglist.Add(strdiv);
}
}
}
if (convertedCharacter == ‘<‘)
{
lowerThanCharCounter++;
lowerThanCharPos = i;
}
}
finally
{
i++;
}
}
return taglist;
}
///<summary>
/// 获取指定网页的源码，支持编码自动识别
///</summary>
///<param name="url"></param>
///<returns></returns>
public static string getDataFromUrl(string url)
{
string str = string.Empty;
HttpWebRequest request = (HttpWebRequest)HttpWebRequest.Create(url);
//设置http头
request.AllowAutoRedirect = true;
request.AllowWriteStreamBuffering = true;
request.Referer = "";
request.Timeout = 10 * 1000;
request.UserAgent = "";
HttpWebResponse response = null;
try
{
response = (HttpWebResponse)request.GetResponse();
if (response.StatusCode == HttpStatusCode.OK)
{
//根据http应答的http头来判断编码
string characterSet = response.CharacterSet;
Encoding encode;
if (characterSet != "")
{
if (characterSet == "ISO-8859-1")
{
characterSet = "gb2312";
}
encode = Encoding.GetEncoding(characterSet);
}
else
{
encode = Encoding.Default;
}
//声明一个内存流来保存http应答流
Stream receiveStream = response.GetResponseStream();
MemoryStream mStream = new MemoryStream();
byte[] bf = new byte[255];
int count = receiveStream.Read(bf, 0, 255);
while (count > 0)
{
mStream.Write(bf, 0, count);
count = receiveStream.Read(bf, 0, 255);
}
receiveStream.Close();
mStream.Seek(0, SeekOrigin.Begin);
//从内存流里读取字符串
StreamReader reader = new StreamReader(mStream, encode);
char[] buffer = new char[1024];
count = reader.Read(buffer, 0, 1024);
while (count > 0)
{
str += new String(buffer, 0, count);
count = reader.Read(buffer, 0, 1024);
}
//从解析出的字符串里判断charset，如果和http应答的编码不一直
//那么以页面声明的为准，再次从内存流里重新读取文本
Regex reg =
new Regex(@"<meta[\s\S]+?charset=(.*)""[\s\S]+?>",
RegexOptions.Multiline | RegexOptions.IgnoreCase);
MatchCollection mc = reg.Matches(str);
if (mc.Count > 0)
{
string tempCharSet = mc[0].Result("$1");
if (string.Compare(tempCharSet, characterSet, true) != 0)
{
encode = Encoding.GetEncoding(tempCharSet);
str = string.Empty;
mStream.Seek(0, SeekOrigin.Begin);
reader = new StreamReader(mStream, encode);
buffer = new char[255];
count = reader.Read(buffer, 0, 255);
while (count > 0)
{
str += new String(buffer, 0, count);
count = reader.Read(buffer, 0, 255);
}
}
}
reader.Close();
mStream.Close();
}
}
catch (Exception ex)
{
Trace.TraceError(ex.ToString());
}
finally
{
if (response != null)
response.Close();
}
return str;
}
///<summary>
/// 从一段网页源码中获取正文
///</summary>
///<param name="input"></param>
///<returns></returns>
public static string GetMainContent(string input)
{
string reg1 = @"<(p|br)[^<]*>";
string reg2 =
@"(\[([^=]*)(=[^\]]*)?\][\s\S]*?\[/\1\])|(?<lj>(?=[^\u4E00-\u9FA5\uFE30-\uFFA0,."");])<a\s+[^>]*>[^<]{2,}</a>(?=[^\u4E00-\u9FA5\uFE30-\uFFA0,."");]))|(?<Style><style[\s\S]+?/style>)|(?<select><select[\s\S]+?/select>)|(?<Script><script[\s\S]*?/script>)|(?<Explein><\!\-\-[\s\S]*?\-\->)|(?<li><li(\s+[^>]+)?>[\s\S]*?/li>)|(?<Html></?\s*[^> ]+(\s*[^=>]+?=[‘""]?[^""‘]+?[‘""]?)*?[^\[<]*>)|(?<Other>&[a-zA-Z]+;)|(?<Other2>\#[a-z0-9]{6})|(?<Space>\s+)|(\&\#\d+\;)";
//1、获取网页的所有div标签
List<string> list = GetTags(input, "div");
//2、去除汉字少于200字的div
List<string> needToRemove = new List<string>();
foreach (string s in list)
{
Regex r = new Regex("[\u4e00-\u9fa5]");
if (r.Matches(s).Count < 300)
{
needToRemove.Add(s);
}
}
foreach (string s in needToRemove)
{
list.Remove(s);
}
//3、把剩下的div按汉字比例多少倒序排列,
list.Sort(CompareDinosByChineseLength);
if (list.Count < 1)
{
return "";
}
input = list[list.Count - 1];
//4、把p和br替换成特殊的占位符[p][br]
input = new Regex(reg1, RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(input, "[$1]");
//5、去掉HTML标签，保留汉字
input = new Regex(reg2, RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(input, "");
//6、把特殊占维护替换成回车和换行
input = new Regex("\\[p]", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(input, "\r\n ");
input = new Regex("\\[br]", RegexOptions.Multiline | RegexOptions.IgnoreCase).Replace(input, "\r\n");
return input;
}
}