private static unsafe void HtmlEncode(string value, LowLevelTextWriter output)
{
if (value == null)
{
return;
}
if (output == null)
{
throw new ArgumentNullException("output");
}
int index = IndexOfHtmlEncodingChars(value, 0);
if (index == -1)
{
output.Write(value);
return;
}
Debug.Assert(0 <= index && index <= value.Length, "0 <= index && index <= value.Length");
int cch = value.Length - index;
fixed (char* str = value)
{
char* pch = str;
while (index-- > 0)
{
output.Write(*pch++);
}
for (; cch > 0; cch--, pch++)
{
char ch = *pch;
if (ch <= '>')
{
switch (ch)
{
case '<':
output.Write("<");
break;
case '>':
output.Write(">");
break;
case '"':
output.Write(""");
break;
case '\'':
output.Write("'");
break;
case '&':
output.Write("&");
break;
default:
output.Write(ch);
break;
}
}
else
{
int valueToEncode = -1; // set to >= 0 if needs to be encoded
#if ENTITY_ENCODE_HIGH_ASCII_CHARS
if (ch >= 160 && ch < 256)
{
// The seemingly arbitrary 160 comes from RFC
valueToEncode = ch;
}
else
#endif // ENTITY_ENCODE_HIGH_ASCII_CHARS
if (s_htmlEncodeConformance == UnicodeEncodingConformance.Strict && Char.IsSurrogate(ch))
{
int scalarValue = GetNextUnicodeScalarValueFromUtf16Surrogate(ref pch, ref cch);
if (scalarValue >= UNICODE_PLANE01_START)
{
valueToEncode = scalarValue;
}
else
{
// Don't encode BMP characters (like U+FFFD) since they wouldn't have
// been encoded if explicitly present in the string anyway.
ch = (char)scalarValue;
}
}
if (valueToEncode >= 0)
{
// value needs to be encoded
output.Write("&#");
output.Write(valueToEncode.ToString(CultureInfo.InvariantCulture));
output.Write(';');
}
else
{
// write out the character directly
output.Write(ch);
}
}
}
}
}
private static void HtmlDecode(string value, LowLevelTextWriter output)
{
if (value == null)
{
return;
}
if (output == null)
{
throw new ArgumentNullException("output");
}
if (!StringRequiresHtmlDecoding(value))
{
output.Write(value); // good as is
return;
}
int l = value.Length;
for (int i = 0; i < l; i++)
{
char ch = value[i];
if (ch == '&')
{
// We found a '&'. Now look for the next ';' or '&'. The idea is that
// if we find another '&' before finding a ';', then this is not an entity,
// and the next '&' might start a real entity (VSWhidbey 275184)
int index = value.IndexOfAny(s_htmlEntityEndingChars, i + 1);
if (index > 0 && value[index] == ';')
{
string entity = value.Substring(i + 1, index - i - 1);
if (entity.Length > 1 && entity[0] == '#')
{
// The # syntax can be in decimal or hex, e.g.
// å --> decimal
// å --> same char in hex
// See http://www.w3.org/TR/REC-html40/charset.html#entities
bool parsedSuccessfully;
uint parsedValue;
if (entity[1] == 'x' || entity[1] == 'X')
{
parsedSuccessfully = UInt32.TryParse(entity.Substring(2), NumberStyles.AllowHexSpecifier, CultureInfo.InvariantCulture, out parsedValue);
}
else
{
parsedSuccessfully = UInt32.TryParse(entity.Substring(1), NumberStyles.Integer, CultureInfo.InvariantCulture, out parsedValue);
}
if (parsedSuccessfully)
{
switch (s_htmlDecodeConformance)
{
case UnicodeDecodingConformance.Strict:
// decoded character must be U+0000 .. U+10FFFF, excluding surrogates
parsedSuccessfully = ((parsedValue < HIGH_SURROGATE_START) || (LOW_SURROGATE_END < parsedValue && parsedValue <= UNICODE_PLANE16_END));
break;
case UnicodeDecodingConformance.Compat:
// decoded character must be U+0001 .. U+FFFF
// null chars disallowed for compat with 4.0
parsedSuccessfully = (0 < parsedValue && parsedValue <= UNICODE_PLANE00_END);
break;
case UnicodeDecodingConformance.Loose:
// decoded character must be U+0000 .. U+10FFFF
parsedSuccessfully = (parsedValue <= UNICODE_PLANE16_END);
break;
default:
Debug.Assert(false, "Should never get here!");
parsedSuccessfully = false;
break;
}
}
if (parsedSuccessfully)
{
if (parsedValue <= UNICODE_PLANE00_END)
{
// single character
output.Write((char)parsedValue);
}
else
{
// multi-character
char leadingSurrogate, trailingSurrogate;
ConvertSmpToUtf16(parsedValue, out leadingSurrogate, out trailingSurrogate);
output.Write(leadingSurrogate);
output.Write(trailingSurrogate);
}
i = index; // already looked at everything until semicolon
continue;
}
}
else
{
i = index; // already looked at everything until semicolon
char entityChar = HtmlEntities.Lookup(entity);
if (entityChar != (char)0)
{
ch = entityChar;
}
else
{
output.Write('&');
output.Write(entity);
output.Write(';');
continue;
}
}
}
}
output.Write(ch);
}
}
