Hello,

thank you for the explanations and comments.

A last remark regarding the NIL token in the SPARQL grammar. It might be worth 
a thought, to remove it from the lexer and substitute it by a parser rule. If 
this does not lead to ambiguousness for an LL(1) parser, special treatment for 
nil might be removed in InLineDataFull, ArgList and BuiltInCall ::= … | BNODE 
'(' ExpressionList? ')' …

And I admit that this is purely a question of style.

Kind regards
Juergen Pfundt


Reply via email to