Ok Jim, let me understand if I'm missing something with a short example.
Take this little section of the grammar
parameter_with_index
: parameterName = 'PARAM_' parameterIndex '=' numericLiteral
;
It shall be changed to
parameter_with_index
: parameterNameLiteral parameterIndex separator numericLiteral
;
parameterNameLiteral
: PARAM_TOKEN
;
separator
: SEPARATOR_TOKEN
;
SEPARATOR_TOKEN
: '='
;
PARAM_TOKEN
: 'PARAM_'
;
is it right?
Thank you all for your precious suggestions :)
________________________________
Da: Jim Idle <jimi at temporal-wave.com>
A: antlr-interest at antlr.org
Inviato: Lun 21 marzo 2011, 16:25:29
Oggetto: Re: [antlr-interest] Grammar issue
Take the 'literals' out of the parser and create real lexer tokens that
are declared before the other, more general tokens.
Using literals in the parser confuses you and will also make error
handling and tree walking more difficult.
Jim
> -----Original Message-----
> From: antlr-interest-bounces at antlr.org [mailto:antlr-interest-
> bounces at antlr.org] On Behalf Of maulattu
> Sent: Monday, March 21, 2011 1:53 AM
> To: antlr-interest at antlr.org
> Subject: Re: [antlr-interest] Grammar issue
>
> Hi all,
> for a matter of convenience I didn't write here all the rules for white
> spaces and so on, but I've them in order to skip white spaces, tab, \r,
> \n and so on.
> It seems the lexer recognize as "textLiteral" the parameter names
> (i.e.:
> "VARIABLE" and "MESSAGE_").
>
> Any suggestions?
>
> Thank you all :)
>
>
> /*----------------------------------------*/
> parametersList
> : parametersDeclaration+ EOF
> ;
>
> parametersDeclaration
> : parameter_with_index /* like PARAM_1 above*/
> | paramemter_without_index /* like VARIABILE above */
> ;
>
> parameter_with_index
> : parameterName = 'PARAM_' parameterIndex '=' numericLiteral
> | parameterName = 'MESSAGE_' parameterIndex '=' textLiteral
> ;
>
> paramemter_without_index
> : parameterName = 'VARIABLE' '=' textLiteral
> ;
> /*----------------------------------------*/
> parameterIndex
> : DECIMAL_LITERAL
> ;
>
> numericLiteral
> : HEX_LITERAL
> | DECIMAL_LITERAL
> ;
>
> textLiteral
> : TEXT_LITERAL
> ;
>
> HEX_LITERAL
> : '0' ('x' | 'X') HEX_DIGIT+
> ;
>
> DECIMAL_LITERAL
> : '0' | '1'..'9' '0'..'9'*
> ;
> /*----------------------------------------*/
> fragment
> TEXT_LITERAL
> : LETTER (LETTER | '0'..'9')*
> ;
>
> fragment
> HEX_DIGIT
> : ('0'..'9' | 'a'..'f' | 'A'..'F')
> ;
>
> fragment
> LETTER
> : 'A'..'Z'
> | 'a'..'z'
> | '_'
> | '+'
> ;
>
>
>
>
> ________________________________
> Da: Stephen Tuttlebee <themightystephen at googlemail.com>
> A: antlr-interest at antlr.org
> Inviato: Gio 17 marzo 2011, 13:00:59
> Oggetto: Re: [antlr-interest] Grammar issue
>
> Hi
>
> Is not the failure to parse:
> MESSAGE_1 = this is a message
> due to the whitespace between the words in the string following the
> equals sign?
>
> Do you have a whitespace lexer rule (often called WS)? I suspect the
> parser is failing due to the fact that it parses up to 'MESSAGE_1 =
> this' just fine but then after that point it is expecting another
> 'parametersDeclaration' to follow, for which the remaining input 'is a
> message' would not have any rules that matches it (the parser would be
> expecting one of three things next, either 'PARAM_', 'MESSAGE_' or
> 'VARIABLE').
>
> I'm not an expert on ANTLR, but I think that might be your problem.
>
> Thanks
> Stephen
>
> List: http://www.antlr.org/mailman/listinfo/antlr-interest
> Unsubscribe:
> http://www.antlr.org/mailman/options/antlr-interest/your-email-address
>
>
>
>
>
> List: http://www.antlr.org/mailman/listinfo/antlr-interest
> Unsubscribe: http://www.antlr.org/mailman/options/antlr-interest/your-
> email-address
List: http://www.antlr.org/mailman/listinfo/antlr-interest
Unsubscribe:
http://www.antlr.org/mailman/options/antlr-interest/your-email-address