How can I write a parser rule that specifically disallows a match on some other rule

[The-Futurist]

Is there a way?

I have a situation in my grammar, I allow an end keyword to be optionally followed by the keyword it is ending, e.g.

loop while (true)
   // stuff
end loop

The closing loop here is optional. But this will fail:

loop while (true)
   // stuff
end 
loop while (true)
   // stuff
end 

Because it will see the loop keyword for the second loop, as being the optional keyword for the end of the first. So I want to have a rule for the end that says to see either END LOOP or END and to match on the latter when the tokens that follow match a legal statement.

This rule "works"

ifEnd
  : END activeStmt? 
  | END IF? 
  ;

But attaches that possible valid statement to the parse tree as if it was part of the statement. I want to only match the END in that first case so that the activeStmt is parsed just like any other.

ifEnd
  : END activeStmt -> END  // pretend we only saw 'end' and push all other tokens back into the stream and let the rest of the parse continue.
  | END IF?    // what follows 'end' is not another legal statement, so if its an 'if' consume that and let the rest of the parse continue.
  ;

That kind of thing, but I have no idea if Antlr supports this.

[The-Futurist]

I'm closing this. The reason the parse failed was legitimate in the actual test case, also I think the behavior I want is already present as I found, described in this post.

#4130