From: Stephen Leake <stephen.a.leake.1@gsfc.nasa.gov>
Subject: Re: Ada95 BNF
Date: 03 May 2001 12:24:01 -0400
Date: 2001-05-03T16:39:16+00:00 [thread overview]
Message-ID: <uk83yv1su.fsf@gsfc.nasa.gov> (raw)
In-Reply-To: 3af1541f$1@pull.gecm.com
"Martin Dowie" <martin.dowie@nospam.baesystems.com> writes:
> Looking through the BNF on AdaPower.com and the LRM that comes with
> adaguide, I noticed
> that package_unit_name is defined as a name, which in turn can be all sorts
> of things,
> including a 'name', which in turn can be lots of things - including a type
> conversion?!
Just because something is allowed by the syntax, doesn't mean it is
allowed by the language. See the appropriate LRM section for the real
rules. In this case, 'parent_unit_name' is first discussed in 10.1.1,
which doesn't actually help that much.
> Shouldn't package_unit_name be defined as an Identifier?..
A 'parent_unit_name' can be of the form ancestor.parent, which is not
an "identifier". So "name" is appropriate.
--
-- Stephe
next prev parent reply other threads:[~2001-05-03 16:24 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2001-05-03 12:23 Ada95 BNF Martin Dowie
2001-05-03 16:24 ` Stephen Leake [this message]
2001-05-03 19:18 ` Martin Dowie
2001-05-04 9:27 ` Marius Amado Alves
2001-05-08 14:53 ` Tucker Taft
2001-05-09 6:18 ` Martin Dowie
2001-05-09 13:47 ` Ted Dennison
-- strict thread matches above, loose matches on Subject: below --
2001-05-04 5:48 Christoph Grein
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox