[antlr-interest] Grammar handling of whitespaces

[Jim Idle]

Use a semantic (probably gated is what is needed) predicate to look at 
the previous off channel token. If it is a whitespace, then disallow the 
following rule, issue an error or whatever it is you need:

{( (TokenStream)input ).get( input.index()-1 ).getType() != WS }?=> 
optionalrule. The get() call does not skip off channel tokens.

Jim



> -----Original Message-----
> From: Shawn Poulson [mailto:spoulson3 at yahoo.com]
> Sent: Thursday, February 07, 2008 5:19 AM
> To: antlr-interest at antlr.org
> Subject: [antlr-interest] Grammar handling of whitespaces
> 
> Hello all,
> 
> I've had success so far with generating my grammar for my C# target,
> with help from several individuals here.  One of the tips I've used is
> to use a lexer rule to skip whitespaces using a rule like:
> 
> WS : (' '|'\r'|'\t'|'\n') { $channel=HIDDEN; }
> 
> This is very helpful for most parser rules, so I don't have to
> explicitly put whitespaces throughout my parser rules.  But, there are
> times where I don't want whitespaces between certain elements.  Is
> there a way to handle this?
> 
> Thanks.
> 
> ---
> Shawn Poulson
> spoulson at explodingcoder.com