[antlr-interest] Tokens and ~()

[Cid Dennis]

I have a rule that should find all characters between / and /:

regex : 
	'/' (r=~('/'))+ '/' 
	; 

The problem is later in my grammar file  I have something that looks like:


fragment operator
	: o=('.pick'|'.length'|'.as'|'.head'|'.tail'|'.sort'
      	|'.filter'|'.map'|'.uc'|'.lc' |'.split' | '.join' | '.query'
      	| '.has' | '.union' | '.difference' | '.intersection' | '.unique' | '.once'
      	| '.duplicates') '(' (e=expr  (',' e1=expr )*)? ')'	
	;

When I parse a file that has this in it.

 prune = mdata.replace(/(?s)(.*v.directedBy..)(\w+\s+\w+)(<\/a.*)/,"$2");

I get mismatched character 'r' expecting 'f'.    This seems to be because I have '.difference'.   What I do not understand is why this that happening.  I would expect the ~('/') to say give me everything except "/" but it gets stuff on other tokens defined in the grammar.

Can someone shed some light on what I am doing wrong.  This is my first grammar with ANTLR and am confused.


------------------------------------------
Cid Dennis
ph : 303 915 9381
em: cid at kynetx.com