然后我们利用正则完全删除ViewState
// 过滤ViewState
private string ViewStateFilter( string strHTML ) {
string matchString1 = "type=\"hidden\" name=\"__VIEWSTATE\" id=\"__VIEWSTATE\"";
string matchString2 = "type=\"hidden\" name=\"__EVENTVALIDATION\" id=\"__EVENTVALIDATION\"";
string matchString3 = "type=\"hidden\" name=\"__EVENTTARGET\" id=\"__EVENTTARGET\"";
string matchString4 = "type=\"hidden\" name=\"__EVENTARGUMENT\" id=\"__EVENTARGUMENT\"";
string positiveLookahead1 = "(?=.*(" + Regex.Escape( matchString1 ) + "))";
string positiveLookahead2 = "(?=.*(" + Regex.Escape( matchString2 ) + "))";
string positiveLookahead3 = "(?=.*(" + Regex.Escape( matchString3 ) + "))";
string positiveLookahead4 = "(?=.*(" + Regex.Escape( matchString4 ) + "))";
RegexOptions opt = RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.CultureInvariant | RegexOptions.Compiled;
Regex[] arrRe = new Regex[] {
new Regex("\\s*<div>" + positiveLookahead1 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead2 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead3 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead3 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead4 + "(.*?)</div>\\s*", opt)
};
foreach ( Regex re in arrRe ) {
strHTML = re.Replace( strHTML, "" );
}
return strHTML;
}
以下是删除页面空白的方法:
// 删除空白
private Regex tabsRe = new Regex( "\\t", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex carriageReturnRe = new Regex( ">\\r\\n<", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex carriageReturnSafeRe = new Regex( "\\r\\n", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex multipleSpaces = new Regex( " ", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex spaceBetweenTags = new Regex( ">\\s<", RegexOptions.Compiled | RegexOptions.Multiline );
private string WhitespaceFilter( string html ) {
html = tabsRe.Replace( html, string.Empty );
html = carriageReturnRe.Replace( html, "><" );
html = carriageReturnSafeRe.Replace( html, " " );
while ( multipleSpaces.IsMatch( html ) )
html = multipleSpaces.Replace( html, " " );
html = spaceBetweenTags.Replace( html, "><" );
html = html.Replace( "//<![CDATA[", "" );
html = html.Replace( "//]]>", "" );
return html;
}
以下是删除ASP.NET控件的垃圾UniqueID名称方法:
// 过滤NamingContainer
private string NamingContainerFilter( string html ) {
RegexOptions opt =
RegexOptions.IgnoreCase |
RegexOptions.Singleline |
RegexOptions.CultureInvariant |
RegexOptions.Compiled;
Regex re = new Regex( "( name=\")(?=.*(" + Regex.Escape( "$" ) + "))([^\"]+?)(\")", opt );
html = re.Replace( html, new MatchEvaluator( delegate( Match m ) {
int lastDollarSignIndex = m.Value.LastIndexOf( '$' );
if ( lastDollarSignIndex >= 0 ) {
return m.Groups[ 1 ].Value + m.Value.Substring( lastDollarSignIndex + 1 );
}
else {
return m.Value;
}
} ) );
return html;
}
最后,我们把以上过滤方法整合到CommonFilter类的Write方法:
public override void Write( byte[] buffer, int offset, int count ) {
// 转换buffer为字符串
byte[] data = new byte[ count ];
Buffer.BlockCopy( buffer, offset, data, 0, count );
string html = System.Text.Encoding.UTF8.GetString( buffer );
//
// 以下整合过滤方法
//
html = NamingContainerFilter( html );
html = ViewStateFilter( html );
html = WhitespaceFilter( html );
byte[] outdata = System.Text.Encoding.UTF8.GetBytes( html );
// 写入磁盘
_cacheStream.Write( outdata, 0, outdata.GetLength( 0 ) );
_responseStream.Write( outdata, 0, outdata.GetLength( 0 ) );
}
private string ViewStateFilter( string strHTML ) {
string matchString1 = "type=\"hidden\" name=\"__VIEWSTATE\" id=\"__VIEWSTATE\"";
string matchString2 = "type=\"hidden\" name=\"__EVENTVALIDATION\" id=\"__EVENTVALIDATION\"";
string matchString3 = "type=\"hidden\" name=\"__EVENTTARGET\" id=\"__EVENTTARGET\"";
string matchString4 = "type=\"hidden\" name=\"__EVENTARGUMENT\" id=\"__EVENTARGUMENT\"";
string positiveLookahead1 = "(?=.*(" + Regex.Escape( matchString1 ) + "))";
string positiveLookahead2 = "(?=.*(" + Regex.Escape( matchString2 ) + "))";
string positiveLookahead3 = "(?=.*(" + Regex.Escape( matchString3 ) + "))";
string positiveLookahead4 = "(?=.*(" + Regex.Escape( matchString4 ) + "))";
RegexOptions opt = RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.CultureInvariant | RegexOptions.Compiled;
Regex[] arrRe = new Regex[] {
new Regex("\\s*<div>" + positiveLookahead1 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead2 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead3 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead3 + "(.*?)</div>\\s*", opt),
new Regex("\\s*<div>" + positiveLookahead4 + "(.*?)</div>\\s*", opt)
};
foreach ( Regex re in arrRe ) {
strHTML = re.Replace( strHTML, "" );
}
return strHTML;
}
以下是删除页面空白的方法:
// 删除空白
private Regex tabsRe = new Regex( "\\t", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex carriageReturnRe = new Regex( ">\\r\\n<", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex carriageReturnSafeRe = new Regex( "\\r\\n", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex multipleSpaces = new Regex( " ", RegexOptions.Compiled | RegexOptions.Multiline );
private Regex spaceBetweenTags = new Regex( ">\\s<", RegexOptions.Compiled | RegexOptions.Multiline );
private string WhitespaceFilter( string html ) {
html = tabsRe.Replace( html, string.Empty );
html = carriageReturnRe.Replace( html, "><" );
html = carriageReturnSafeRe.Replace( html, " " );
while ( multipleSpaces.IsMatch( html ) )
html = multipleSpaces.Replace( html, " " );
html = spaceBetweenTags.Replace( html, "><" );
html = html.Replace( "//<![CDATA[", "" );
html = html.Replace( "//]]>", "" );
return html;
}
以下是删除ASP.NET控件的垃圾UniqueID名称方法:
// 过滤NamingContainer
private string NamingContainerFilter( string html ) {
RegexOptions opt =
RegexOptions.IgnoreCase |
RegexOptions.Singleline |
RegexOptions.CultureInvariant |
RegexOptions.Compiled;
Regex re = new Regex( "( name=\")(?=.*(" + Regex.Escape( "$" ) + "))([^\"]+?)(\")", opt );
html = re.Replace( html, new MatchEvaluator( delegate( Match m ) {
int lastDollarSignIndex = m.Value.LastIndexOf( '$' );
if ( lastDollarSignIndex >= 0 ) {
return m.Groups[ 1 ].Value + m.Value.Substring( lastDollarSignIndex + 1 );
}
else {
return m.Value;
}
} ) );
return html;
}
最后,我们把以上过滤方法整合到CommonFilter类的Write方法:
public override void Write( byte[] buffer, int offset, int count ) {
// 转换buffer为字符串
byte[] data = new byte[ count ];
Buffer.BlockCopy( buffer, offset, data, 0, count );
string html = System.Text.Encoding.UTF8.GetString( buffer );
//
// 以下整合过滤方法
//
html = NamingContainerFilter( html );
html = ViewStateFilter( html );
html = WhitespaceFilter( html );
byte[] outdata = System.Text.Encoding.UTF8.GetBytes( html );
// 写入磁盘
_cacheStream.Write( outdata, 0, outdata.GetLength( 0 ) );
_responseStream.Write( outdata, 0, outdata.GetLength( 0 ) );
}