J3/03-263r1 Date: November 13, 2003 To: J3 From: Rob James Subject: CLASS IS type guard statements and intrinsic types ========================================================== The response to this paper can be found in paper J3/03-267 ========================================================== When the changes involving select type, intrinsic types, and unlimited polymorphics were made, I don't believe the following case was ever considered. SELECT TYPE(a) TYPE IS (integer) ! If "a" is an integer, the following block ! will be executed. ... CLASS IS (integer) ! The following block can never be executed. ... END SELECT It seems weird to me that this should be allowed, since one of the blocks can't be executed. I don't think we should allow intrinsic types in CLASS IS type guard statements. There is no reason for them to appear there. Proposed edits to the Fortran 2003 FCD: --------------------------------------- [162:22+] Add the following constraint: C816.5 (R823) The in a CLASS IS shall not specify an intrinsic type.