fortran - Does the finalization routine need to be elemental in order to be called on the elements of allocatable array that goes out of scope? -

if have allocatable array of finalizable derived type, finalizer called on every individual element when array goes out of scope?

here small code example illustrates question:

module leakytypemodule     implicit none    private     type, public :: leakytype       real, pointer :: dontleakme(:) => null()    contains       procedure :: new       final     :: finalizer    end type  contains     subroutine new(self, n)       class(leakytype), intent(out) :: self       integer         , intent(in)  :: n       allocate(self%dontleakme(n))       self%dontleakme = 42.0    end subroutine     subroutine finalizer(self)       type(leakytype), intent(inout) :: self       if (associated(self%dontleakme)) deallocate(self%dontleakme)    end subroutine  end module   program leak     use leakytypemodule    implicit none     type(leakytype), allocatable :: arr(:)     allocate(arr(1))    call arr(1)%new(1000)    deallocate(arr)  end program 

note program leaks dontleakme array allocated in new() method of leakytype. @ first bit surprising me but, discovered problem can fixed declaring finalizer elemental. both gfortran , ifort behave in same way, assume behaviour following fortran 2003 standard.

can confirm this? honest have hard time time understanding standard says on particular point.

right can't see use in not declaring finalizers elemental. have application i'm overlooking?

the rules determining whether final procedure invoked, , final procedure invoked, same resolution of generic procedures in terms of rank matching requirements.

noting question tagged fortran 2003...

elemental procedures in fortran 2003 , prior have pure. if finalizer needs incompatible pure attribute (which reasonably common) finalizer cannot elemental, , need write rank specific variants.

fortran 2008 introduces concept of impure elemental, quite handy writing finalizers.