From owner-sc22wg5+sc22wg5-dom8=www.open-std.org@open-std.org Mon Mar 12 21:55:47 2012 Return-Path: X-Original-To: sc22wg5-dom8 Delivered-To: sc22wg5-dom8@www.open-std.org Received: by www.open-std.org (Postfix, from userid 521) id 161069DB113; Mon, 12 Mar 2012 21:55:46 +0100 (CET) Delivered-To: sc22wg5@open-std.org Received: from mailrelay1.lrz-muenchen.de (mailrelay1.lrz-muenchen.de [129.187.254.106]) by www.open-std.org (Postfix) with ESMTP id ED3B09DB112 for ; Mon, 12 Mar 2012 21:55:45 +0100 (CET) Received: from postout1.mail.lrz.de ([10.156.6.18] [10.156.6.18]) by mailrelay1.lrz-muenchen.de with ESMTP; Mon, 12 Mar 2012 21:55:40 +0100 Received: from BADWLRZ-SWHBT1.ads.mwn.de (BADWLRZ-SWHBT1.ads.mwn.de [IPv6:2001:4ca0:0:108::125]) (using TLSv1 with cipher AES128-SHA (128/128 bits)) (No client certificate requested) by postout1.mail.lrz.de (Postfix) with ESMTPS id 9F018C5823; Mon, 12 Mar 2012 21:55:40 +0100 (CET) Received: from BADWLRZ-SWMBX1.ads.mwn.de ([fe80::11b4:b130:c4e2:2d0e]) by BADWLRZ-SWHBT1.ads.mwn.de ([fe80::e42f:e9f5:bde9:f99b%16]) with mapi id 14.01.0355.002; Mon, 12 Mar 2012 21:55:40 +0100 From: "Bader, Reinhold" To: WG5 , "j3@j3-fortran.org" Subject: Comment on N1904 cont'd Thread-Topic: Comment on N1904 cont'd Thread-Index: Ac0AkXf3Bf9BOgwWSVCPILpVVFzrsQ== Date: Mon, 12 Mar 2012 20:55:39 +0000 Message-Id: <166ED263DF83324D9A3BA67FB6772B2B47FF8580@BADWLRZ-SWMBX1.ads.mwn.de> Accept-Language: de-DE, en-US Content-Language: de-DE X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [129.187.48.252] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Sender: owner-sc22wg5@open-std.org Precedence: bulk Hello all,=20 It appears to me to be a good idea to enforce interoperable=20 type and type parameters for non-assumed-type dummy entities. This leaves less room for misuse, and by pushing all non- interoperable stuff to the TYPE(*) case the programmer is=20 warned by the notation that she/he's aiming at the space between=20 her/his toes. This would obsolete my issues (H) to (J), and=20 (K) with the possible exception of CLASS(*). [ Remaining TYPE(*) issues are for a different thread. ] With respect to John's point about C1255 I am not sure I agree;=20 the constraint points to paragraphs where interoperability of variables is defined, and I think it is there that the extensions to the concept of interoperability for variables should be placed;=20 this would also have the advantage that the restrictions on coarrays as well as the topic of the paragraph above are already included=20 there. It would simply be necessary to remove some restrictions spelled out in 15.3.5 and 15.3.6 and spell out what the new stuff interoperates with. 15.3.5 of F2008 could be reworded to "A named scalar Fortran variable is interoperable if and only if it is not a coarray, and either (a) its type and type parameters are interoperable, and it is a=20 dummy argument if it has the ALLOCATABLE or POINTER attribute, or (b) it is assumed-type. An interoperable Fortran variable that has the ALLOCATABLE or=20 POINTER attribute, or is of type character with a length that is assumed or declared by an expression that is not a constant=20 expression is interoperable with a C descriptor. If it is assumed-type, it is interoperable with a pointer to void.=20 Otherwise, it is interoperable with a scalar C entity that has=20 the same type and type parameters." <<>> 15.3.6 of F2008 could be reworded to para 1: "A Fortran variable that is a named array is interoperable if and=20 only if it is not a coarray and either (a) its type and type parameters are interoperable, it is of=20 explicit shape, assumed size, assumed shape, assumed rank,=20 or is a dummy argument with the ALLOCATABLE or POINTER=20 attributes, or (b) it is assumed-type, and is of assumed-size, assumed-shape, or assumed-rank." para 2 stays as-is. Delete NOTE 15.16 and add para 3: "An interoperable Fortran named array that=20 (a) has the ALLOCATABLE or POINTER attribute, or is of type character=20 with a length that is assumed or declared by an expression that=20 is not a constant expression, or (b) is of assumed-type and assumed-shape or assumed-rank is interoperable with a C descriptor. An interoperable Fortran=20 named array that is assumed-type and assumed-size is interoperable with a pointer to void." =20 <<>> =20 -- Dr. Reinhold Bader Leibniz Supercomputing Centre (http://www.lrz.de) / HPC Support =20 Tel. +49 89 35831 8825 - Fax +49 89 35831 8625