public override void Read(Tokeniser t, CharacterReader r)
{
if (r.IsEmpty())
{
t.EofError(this);
t.DoctypePending.ForceQuirks = true;
t.EmitDoctypePending();
t.Transition(Data);
return;
}
if (r.MatchesAny('\t', '\n', '\r', '\f', ' '))
{
r.Advance(); // ignore whitespace
}
else if (r.Matches('>'))
{
t.EmitDoctypePending();
t.AdvanceTransition(Data);
}
else if (r.MatchConsumeIgnoreCase("PUBLIC"))
{
t.Transition(AfterDoctypePublicKeyword);
}
else if (r.MatchConsumeIgnoreCase("SYSTEM"))
{
t.Transition(AfterDoctypeSystemKeyword);
}
else
{
t.Error(this);
t.DoctypePending.ForceQuirks = true;
t.AdvanceTransition(BogusDoctype);
}
}
public char?ConsumeCharacterReference(char?additionalAllowedCharacter, bool inAttribute)
{
if (_reader.IsEmpty())
{
return(null);
}
if (additionalAllowedCharacter != null && additionalAllowedCharacter == _reader.Current())
{
return(null);
}
if (_reader.MatchesAny('\t', '\n', '\r', '\f', ' ', '<', '&'))
{
return(null);
}
_reader.Mark();
if (_reader.MatchConsume("#"))
{ // numbered
bool isHexMode = _reader.MatchConsumeIgnoreCase("X");
string numRef = isHexMode ? _reader.ConsumeHexSequence() : _reader.ConsumeDigitSequence();
if (numRef.Length == 0)
{ // didn't match anything
CharacterReferenceError("Numeric reference with no numerals");
_reader.RewindToMark();
return(null);
}
if (!_reader.MatchConsume(";"))
{
CharacterReferenceError("Missing semicolon"); // missing semi
}
int charval = -1;
try
{
int numbase = isHexMode ? 16 : 10;
charval = Convert.ToInt32(numRef, numbase);
}
catch (FormatException)
{
} // skip
if (charval == -1 || (charval >= 0xD800 && charval <= 0xDFFF) || charval > 0x10FFFF)
{
CharacterReferenceError("Character outside of valid range");
return(ReplacementChar);
}
else
{
// todo: implement number replacement table
// todo: check for extra illegal unicode points as parse errors
return((char)charval);
}
}
else
{ // named
// get as many letters as possible, and look for matching entities. unconsume backwards till a match is found
string nameRef = _reader.ConsumeLetterThenDigitSequence();
bool looksLegit = _reader.Matches(';');
// found if a base named entity without a ;, or an extended entity with the ;.
bool found = (Entities.IsBaseNamedEntity(nameRef) || (Entities.IsNamedEntity(nameRef) && looksLegit));
if (!found)
{
_reader.RewindToMark();
if (looksLegit)
{
CharacterReferenceError(string.Format("Invalid named referenece '{0}'", nameRef));
}
return(null);
}
if (inAttribute && (_reader.MatchesLetter() || _reader.MatchesDigit() || _reader.MatchesAny('=', '-', '_')))
{
// don't want that to match
_reader.RewindToMark();
return(null);
}
if (!_reader.MatchConsume(";"))
{
CharacterReferenceError("Missing semicolon"); // missing semi
}
return(Entities.GetCharacterByName(nameRef));
}
}
public override void Read(Tokeniser t, CharacterReader r)
{
if (r.MatchConsume("--"))
{
t.CreateCommentPending();
t.Transition(CommentStart);
}
else if (r.MatchConsumeIgnoreCase("DOCTYPE"))
{
t.Transition(Doctype);
}
else if (r.MatchConsume("[CDATA["))
{
// todo: should actually check current namepspace, and only non-html allows cdata. until namespace
// is implemented properly, keep handling as cdata
//} else if (!t.currentNodeInHtmlNS() && r.matchConsume("[CDATA[")) {
t.Transition(CDataSection);
}
else
{
t.Error(this);
t.AdvanceTransition(BogusComment); // advance so this character gets in bogus comment data's rewind
}
}