Question

How would you translate this portion of code written in ANTLR 3 into ANTLR 4?

expr: (Identifier '.')=> (refIdentifier)
  | (Identifier '!')=> (refIdentifier)
  | (Identifier '=>')=> (lambdaExpression);

I mean this kind of semantic predicate does not seem to exist now. What could I use Instead?

Was it helpful?

Solution

In ANTLR v4, there are no longer gated semantic predicates, { ... }?=>, and there are also no longer syntactic predicates, ( ... )=>, because the parsing algorithm used in v4 can resolve the ambiguities (the need for such predicates are no longer needed). So, this should just work for you:

expr
 : refIdentifier
 | refIdentifier
 | lambdaExpression
 ;

Note that there is just one type of predicate in v4: semantic predicates, { ... }?. If you need to inspect the contents of a token, for example, you can do it like this:

id_capitals_only
 : {_input.LT(1).getText().matches("[A-Z]+")}? ID
 ;

ID
 : [a-zA-Z]+
 ;

EDIT

And as Sam Harwell mentions in the comments:

The semantic predicates {...}? in V4 work like the gated semantic predicates did in V3. The ungated predicates from V3 do not have a counterpart in ANTLR 4.

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top