[antlr-interest] Matching the * character

[Steve O'Hara]

Thanks Gavin for the tip.

Steve

-----Original Message-----
From: info at mirality.co.nz [mailto:info at mirality.co.nz] On Behalf Of
Gavin Lambert
Sent: 28 March 2008 07:50
To: Steve O'Hara; Loring Craymer; antlr-interest at antlr.org
Subject: Re: [antlr-interest] Matching the * character

At 13:58 28/03/2008, Steve O'Hara wrote:
>Thanks for your help Loring - unfortunately, it hasn't helped.
>I've condensed the example down to something really simple;
>
>subExpressionText :  QuotedString STAR? WS?;
>QuotedString :'\'' (~'\'')* '\'' ( '\'' (~'\'')* '\'' )* ;
>WS : (' ' | '\t' | '\n' | '\r')+ ;
>STAR : '*' ;

That QuotedString rule seems like a very weird 
construction to me.  I guess it will work anyway, 
but I'd write it more like this (seems more 
explicit, and leaves room to add additional 
escapes later if you want to):

   QuotedString : '\'' (~'\'' | '\'\'')* '\'';

Not really relevant to your problem, but I'm not 
sure why you're getting the behaviour you're 
describing unless it's something like what Ter 
suggested and something else is messing with the 
input before it gets to the parser.  Or unless 
you've got additional rules you haven't shown us 
that are interfering.