// in data state, gather characters until a character reference or tag is found
public override void Read(Tokeniser t, CharacterReader r)
{
switch (r.Current())
{
case '&':
t.AdvanceTransition(CharacterReferenceInData);
break;
case '<':
t.AdvanceTransition(TagOpen);
break;
case _nullChar:
t.Error(this); // NOT replacement character (oddly?)
t.Emit(r.Consume());
break;
case _eof:
t.Emit(new Token.EOF());
break;
default:
string data = r.ConsumeToAny('&', '<', _nullChar);
t.Emit(data);
break;
}
}
/// handles data in title, textarea etc
public override void Read(Tokeniser t, CharacterReader r)
{
switch (r.Current())
{
case '&':
t.AdvanceTransition(CharacterReferenceInRcData);
break;
case '<':
t.AdvanceTransition(RcDataLessThanSign);
break;
case _nullChar:
t.Error(this);
r.Advance();
t.Emit(_replacementChar);
break;
case _eof:
t.Emit(new Token.EOF());
break;
default:
string data = r.ConsumeToAny('&', '<', _nullChar);
t.Emit(data);
break;
}
}
public void matchesAny()
{
char[] scan = { ' ', '\n', '\t' };
CharacterReader r = new CharacterReader("One\nTwo\tThree");
Assert.IsFalse(r.MatchesAny(scan));
Assert.AreEqual("One", r.ConsumeToAny(scan));
Assert.IsTrue(r.MatchesAny(scan));
Assert.AreEqual('\n', r.Consume());
Assert.IsFalse(r.MatchesAny(scan));
}
public void consumeToAny()
{
CharacterReader r = new CharacterReader("One &bar; qux");
Assert.AreEqual("One ", r.ConsumeToAny('&', ';'));
Assert.IsTrue(r.Matches('&'));
Assert.IsTrue(r.Matches("&bar;"));
Assert.AreEqual('&', r.Consume());
Assert.AreEqual("bar", r.ConsumeToAny('&', ';'));
Assert.AreEqual(';', r.Consume());
Assert.AreEqual(" qux", r.ConsumeToAny('&', ';'));
}
public override void Read(Tokeniser t, CharacterReader r)
{
char c = r.Current();
switch (c)
{
case '-':
t.Emit(c);
t.AdvanceTransition(ScriptDataDoubleEscapedDash);
break;
case '<':
t.Emit(c);
t.AdvanceTransition(ScriptDataDoubleEscapedLessthanSign);
break;
case _nullChar:
t.Error(this);
r.Advance();
t.Emit(_replacementChar);
break;
case _eof:
t.EofError(this);
t.Transition(Data);
break;
default:
string data = r.ConsumeToAny('-', '<', _nullChar);
t.Emit(data);
break;
}
}
public override void Read(Tokeniser t, CharacterReader r)
{
if (r.IsEmpty())
{
t.EofError(this);
t.Transition(Data);
return;
}
switch (r.Current())
{
case '-':
t.Emit('-');
t.AdvanceTransition(ScriptDataEscapedDash);
break;
case '<':
t.AdvanceTransition(ScriptDataEscapedLessThanSign);
break;
case _nullChar:
t.Error(this);
r.Advance();
t.Emit(_replacementChar);
break;
default:
string data = r.ConsumeToAny('-', '<', _nullChar);
t.Emit(data);
break;
}
}
// from < or </ in data, will have start or end tag pending
public override void Read(Tokeniser t, CharacterReader r)
{
// previous TagOpen state did NOT Consume, will have a letter char in current
string tagName = r.ConsumeToAny('\t', '\n', '\r', '\f', ' ', '/', '>', _nullChar).ToLowerInvariant();
t.TagPending.AppendTagName(tagName);
switch (r.Consume())
{
case '\t':
case '\n':
case '\r':
case '\f':
case ' ':
t.Transition(BeforeAttributeName);
break;
case '/':
t.Transition(SelfClosingStartTag);
break;
case '>':
t.EmitTagPending();
t.Transition(Data);
break;
case _nullChar: // replacement
t.TagPending.AppendTagName(_replacementStr);
break;
case _eof: // should Emit pending tag?
t.EofError(this);
t.Transition(Data);
break;
// no default, as covered with above ConsumeToAny
}
}
public override void Read(Tokeniser t, CharacterReader r)
{
char c = r.Current();
switch (c)
{
case '-':
t.AdvanceTransition(CommentEndDash);
break;
case _nullChar:
t.Error(this);
r.Advance();
t.CommentPending.Data.Append(_replacementChar);
break;
case _eof:
t.EofError(this);
t.EmitCommentPending();
t.Transition(Data);
break;
default:
t.CommentPending.Data.Append(r.ConsumeToAny('-', _nullChar));
break;
}
}
public override void Read(Tokeniser t, CharacterReader r)
{
string value = r.ConsumeToAny('\t', '\n', '\r', '\f', ' ', '&', '>', _nullChar, '"', '\'', '<', '=', '`');
if (value.Length > 0)
{
t.TagPending.AppendAttributeValue(value);
}
char c = r.Consume();
switch (c)
{
case '\t':
case '\n':
case '\r':
case '\f':
case ' ':
t.Transition(BeforeAttributeName);
break;
case '&':
char? reference = t.ConsumeCharacterReference('>', true);
if (reference != null)
{
t.TagPending.AppendAttributeValue(reference.Value);
}
else
{
t.TagPending.AppendAttributeValue('&');
}
break;
case '>':
t.EmitTagPending();
t.Transition(Data);
break;
case _nullChar:
t.Error(this);
t.TagPending.AppendAttributeValue(_replacementChar);
break;
case _eof:
t.EofError(this);
t.Transition(Data);
break;
case '"':
case '\'':
case '<':
case '=':
case '`':
t.Error(this);
t.TagPending.AppendAttributeValue(c);
break;
// no default, handled in Consume to any above
}
}
public override void Read(Tokeniser t, CharacterReader r)
{
string value = r.ConsumeToAny('\'', '&', _nullChar);
if (value.Length > 0)
{
t.TagPending.AppendAttributeValue(value);
}
char c = r.Consume();
switch (c)
{
case '\'':
t.Transition(AfterAttributeValueQuoted);
break;
case '&':
char? reference = t.ConsumeCharacterReference('\'', true);
if (reference != null)
{
t.TagPending.AppendAttributeValue(reference.Value);
}
else
{
t.TagPending.AppendAttributeValue('&');
}
break;
case _nullChar:
t.Error(this);
t.TagPending.AppendAttributeValue(_replacementChar);
break;
case _eof:
t.EofError(this);
t.Transition(Data);
break;
// no default, handled in Consume to any above
}
}
// from before attribute name
public override void Read(Tokeniser t, CharacterReader r)
{
string name = r.ConsumeToAny('\t', '\n', '\r', '\f', ' ', '/', '=', '>', _nullChar, '"', '\'', '<');
t.TagPending.AppendAttributeName(name.ToLowerInvariant());
char c = r.Consume();
switch (c)
{
case '\t':
case '\n':
case '\r':
case '\f':
case ' ':
t.Transition(AfterAttributeName);
break;
case '/':
t.Transition(SelfClosingStartTag);
break;
case '=':
t.Transition(BeforeAttributeValue);
break;
case '>':
t.EmitTagPending();
t.Transition(Data);
break;
case _nullChar:
t.Error(this);
t.TagPending.AppendAttributeName(_replacementChar);
break;
case _eof:
t.EofError(this);
t.Transition(Data);
break;
case '"':
case '\'':
case '<':
t.Error(this);
t.TagPending.AppendAttributeName(c);
break;
// no default, as covered in ConsumeToAny
}
}