/// <summary> Simplifies percent-encoded and entity-encoded characters to their
/// simplest form so that they can be properly validated. Attackers
/// frequently use encoding schemes to disguise their attacks and bypass
/// validation routines.
///
/// Handling multiple encoding schemes simultaneously is difficult, and
/// requires some special consideration. In particular, the problem of
/// double-encoding is difficult for parsers, and combining several encoding
/// schemes in double-encoding makes it even harder. Consider decoding
///
/// [PRE]
/// &lt;
/// [/PRE]
///
/// or
///
/// [PRE]
/// %26lt;
/// [/PRE]
///
/// or
///
/// [PRE]
/// &lt;
/// [/PRE].
///
/// This implementation disallows ALL double-encoded characters and throws an
/// IntrusionException when they are detected. Also, named entities that are
/// not known are simply removed.
///
/// Note that most data from the browser is likely to be encoded with URL
/// encoding (FIXME: RFC). The web server will decode the URL and form data
/// once, so most encoded data received in the application must have been
/// double-encoded by the attacker. However, some HTTP inputs are not decoded
/// by the browser, so this routine allows a single level of decoding.
///
/// </summary>
/// <param name="input">Unvalidated input from an HTTP request.
/// </param>
/// <returns> The canonicalized string.
/// </returns>
/// <seealso cref="Owasp.Esapi.Interfaces.IEncoder.Canonicalize(string)">
/// </seealso>
public string Canonicalize(string input)
{
StringBuilder sb = new StringBuilder();
EncodedStringReader reader = new EncodedStringReader(input);
while (reader.HasNext())
{
EncodedCharacter c = reader.NextCharacter;
if (c != null)
{
sb.Append(c.Unencoded);
}
}
return(sb.ToString());
}
/// <summary> HTML Entity encode utility method. To avoid double-encoding, this method
/// logs a warning if HTML entity encoded characters are passed in as input.
/// Double-encoded characters in the input cause an exception to be thrown.
/// </summary>
/// <param name="input">The input to encode.
/// </param>
/// <param name="immune">The immune characters.
/// </param>
/// <param name="baseChars">The base characters.
/// </param>
/// <returns> The encoded string.
/// </returns>
private string EntityEncode(string input, char[] baseChars, char[] immune)
{
System.Text.StringBuilder sb = new System.Text.StringBuilder();
EncodedStringReader reader = new EncodedStringReader(input);
while (reader.HasNext())
{
EncodedCharacter c = reader.NextCharacter;
if (c != null)
{
if (IsContained(baseChars, c.Unencoded) || IsContained(immune, c.Unencoded))
{
sb.Append(c.Unencoded);
}
else
{
sb.Append(c.GetEncoded(ENTITY_ENCODING));
}
}
}
return(sb.ToString());
}
/// <summary> HTML Entity encode utility method. To avoid double-encoding, this method
/// logs a warning if HTML entity encoded characters are passed in as input.
/// Double-encoded characters in the input cause an exception to be thrown.
/// </summary>
/// <param name="input">The input to encode.
/// </param>
/// <param name="immune">The immune characters.
/// </param>
/// <param name="baseChars">The base characters.
/// </param>
/// <returns> The encoded string.
/// </returns>
private string EntityEncode(string input, char[] baseChars, char[] immune)
{
System.Text.StringBuilder sb = new System.Text.StringBuilder();
EncodedStringReader reader = new EncodedStringReader(input);
while (reader.HasNext())
{
EncodedCharacter c = reader.NextCharacter;
if (c != null)
{
if (IsContained(baseChars, c.Unencoded) || IsContained(immune, c.Unencoded))
{
sb.Append(c.Unencoded);
}
else
{
sb.Append(c.GetEncoded(ENTITY_ENCODING));
}
}
}
return sb.ToString();
}
/// <summary> Simplifies percent-encoded and entity-encoded characters to their
/// simplest form so that they can be properly validated. Attackers
/// frequently use encoding schemes to disguise their attacks and bypass
/// validation routines.
///
/// Handling multiple encoding schemes simultaneously is difficult, and
/// requires some special consideration. In particular, the problem of
/// double-encoding is difficult for parsers, and combining several encoding
/// schemes in double-encoding makes it even harder. Consider decoding
///
/// [PRE]
/// &lt;
/// [/PRE]
///
/// or
///
/// [PRE]
/// %26lt;
/// [/PRE]
///
/// or
///
/// [PRE]
/// &lt;
/// [/PRE].
///
/// This implementation disallows ALL double-encoded characters and throws an
/// IntrusionException when they are detected. Also, named entities that are
/// not known are simply removed.
///
/// Note that most data from the browser is likely to be encoded with URL
/// encoding (FIXME: RFC). The web server will decode the URL and form data
/// once, so most encoded data received in the application must have been
/// double-encoded by the attacker. However, some HTTP inputs are not decoded
/// by the browser, so this routine allows a single level of decoding.
///
/// </summary>
/// <param name="input">Unvalidated input from an HTTP request.
/// </param>
/// <returns> The canonicalized string.
/// </returns>
/// <seealso cref="Owasp.Esapi.Interfaces.IEncoder.Canonicalize(string)">
/// </seealso>
public string Canonicalize(string input)
{
StringBuilder sb = new StringBuilder();
EncodedStringReader reader = new EncodedStringReader(input);
while (reader.HasNext())
{
EncodedCharacter c = reader.NextCharacter;
if (c != null)
{
sb.Append(c.Unencoded);
}
}
return sb.ToString();
}
