[antlr-interest] nondeterminism question

[shmuel siegel]

Klaas Hofstra wrote:

>Hi,
>
>I encountered a nondeterminism warning but when I look at my grammar I don't 
>see the reason.
>
>When I check the generated code I see exactly what I expect. However, I must 
>be missing something because I get this warning:
>
>warning:nondeterminism between alts 1 and 2 of block upon
>     k==1:LPAREN
>
>The warning point to the "integer" line in this rule:
>
>typeDef
>    : "signed"
>    | "unsigned"
>    | "float"
>    | ("integer" ( LPAREN! range RPAREN! )? )      <= warning on this line
>    ;
>
>Selected parts of the rest of the grammar are shown below:
>
>-----------------------------------
>
>range 
>    : expression DOTDOT! expression
>    ;
>
>expression
>/* 
> standard (C like) expression grammar with an added bitIndex and 
> bitSlice operator as shown below
>*/
>
>
>postFixExpression
>    : (primaryExpression | (LPAREN! expression RPAREN!)) (
>      (bitIndexExpression) => bitIndexExpression
>    | bitSliceExpression 
>    | arrayIndexExpression
>     )?
>    ;
>
>arrayIndexExpression
>    : LBRACKET! logicalExpression RBRACKET!
>    ;
>
>bitIndexExpression
>    : LPAREN! logicalExpression RPAREN!
>    ;
>
>bitSliceExpression
>    : LPAREN! logicalExpression DOTDOT! logicalExpression RPAREN!
>    ;
>
>
>-----------------------------------
>
>I hope someone can explain to me why I get this warning so I can fix my 
>grammar.
>
>Thanks in advance,
>
>Klaas
>
>  
>
I am guessing at the structure of expression but it seems to me that 
range, and therefore the line in question, can't distinguish between a 
grouped expression and a bitSpliceExpression. Does changing k, the 
lookahead, value help.


-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.10.18/86 - Release Date: 8/31/2005