[antlr-interest] Re: ANTLR 3: Lexer problem

Oliver Zeigermann oliver.zeigermann at gmail.com
Mon Sep 12 07:32:40 PDT 2005


Forgot to mention, the problem is that the end tag "</t>" can not be
parsed as the generated lexer upon seing '<' predicts a second "<t>"
and chokes when it sees the '/'.

Oliver

2005/9/12, Oliver Zeigermann <oliver.zeigermann at gmail.com>:
> Again me - I am pretty sure to start bothering ;)
> 
> Anyway, any idea why this lexer grammar
> 
> ELEMENT
>     : "<t>"
>             (ELEMENT
>             | (options {greedy=true;} : ~'<')*
>             )*
>       "</t>"
>     ;
> 
> fails to parse this
> 
> <t>Huhu</t>
> 
> while this works ok:
> 
> ELEMENT
>     : "<t>"
>             (ELEMENT
>             | ~'<'
>             )*
>       "</t>"
>     ;
> 
> The second grammar is not suitable for me as I want text (everything,
> not starting with '>') to be reported in a bunch, and not as single
> characters. And - of course - this is not my real grammar, but a
> simplified version of it.
> 
> Oliver
>


More information about the antlr-interest mailing list