Austin Hastings wrote: > No, sir. I was referring to "The imaginary token must be defined > elsewhere in a grammar or in the tokens section." An example: IMAGINARY:; Empty rules, which aren't called elsewhere. Best regards, Johannes Luber