!standard 13.1.1(17/3)                                    16-06-06  AI12-0194-1/01
!class binding interpretation 16-06-06
!status Amendment 1-2012 16-07-21
!status WG9 Approved 16-10-08
!status ARG Approved 11-0-0  16-06-11
!status work item 16-06-06
!status received 16-02-03
!priority Low
!difficulty Easy
!qualifier Omission
!subject Language-defined aspects and entry bodies
!summary

Language-defined aspects are not allowed on entry bodies.

!question

!recommendation

(See Summary.)

!wording

Modify 13.1.1(17/3):

There are no language-defined aspects that may be specified on a
renaming_declaration, a generic_formal_parameter_declaration, a subunit,
a package_body, a task_body, a protected_body, {an entry_body, }or a
body_stub other than a subprogram_body_stub. 

!discussion

13.1.1(17/3) applies to all kinds of bodies other than those that can act as
specifications. Entry_Body should be included; this clearly was an oversight
of AI12-0169-1 (which added these aspect_specifications).

!corrigendum 13.1.1(17/3)

@drepl
There are no language-defined aspects that may be specified on a @fa<renaming_declaration>,
a @fa<generic_formal_parameter_declaration>, a @fa<subunit>, a @fa<package_body>,
a @fa<task_body>, a @fa<protected_body>, or a @fa<body_stub> other than a
@fa<subprogram_body_stub>.
@dby
There are no language-defined aspects that may be specified on a @fa<renaming_declaration>,
a @fa<generic_formal_parameter_declaration>, a @fa<subunit>, a @fa<package_body>,
a @fa<task_body>, a @fa<protected_body>, an @fa<entry_body>, or a @fa<body_stub>
other than a @fa<subprogram_body_stub>.

!ASIS

No ASIS effect.

!ACATS test

An ACATS B-Test should be created.

!appendix

From: Randy Brukardt
Sent: Wednesday, February  3, 2016  7:54 PM

n AI12-0169-1, we allowed aspect specifications on entry bodies.

However, we only made a syntax change. In particular, we didn't add a rule 
like 13.1.1(17/3) or 13.1.1(18/4) for entry bodies.

As such, it doesn't appear that any rule prevents putting Pre on an entry
body:

    entry Fooey (P : in out Natural) with Pre => P mod 2 = 0 is
       ...
    end Fooey;

This appears legal, while the similar subprogram body would be illegal.

It's bizarre that we ban language-defined aspects on all bodies *except* entry
bodies. I have to think this is an oversight (probably happened because we
were too tired after arguing about the syntax).

****************************************************************

From: Tucker Taft
Sent: Wednesday, February  3, 2016  9:41 PM

Agreed. Shouldn't be legal.

****************************************************************