Identifier classification

In order to resolve the type name ambiguity, the scanner needs to distinguish ordinary identifiers that are defined as type names from those that are not. Two additional identifier classes can be used to embody important semantic properties:

• typedef_name: An applied occurrence of an ordinary identifier defined as a type name. This category of identifier is necessary to resolve an ambiguity.

• typedef_def: A defining occurrence of an ordinary identifier defined as a type name. This category of identifier is useful because the actions needed to define an identifier representing a type are different from those needed to define an identifier representing a typed object.

• member_def: A defining occurrence of a member identifier. This category of identifier is useful because a member identifier belongs to a different name space than an ordinary identifier. Applied occurrences of member identifiers occur in contexts that are syntactically distinct, but there is no such distinction for defining occurrences. This classification provides that distinction.

• identifier: Identifier occurrences whose processing requirements can be determined from their gramatical construct.