Re: What compiler to use
From: Richard E Maine (nospam_at_see.signature)
Date: 01/07/05
- Next message: beliavsky_at_aol.com: "Re: What compiler to use"
- Previous message: Richard E Maine: "Re: Opening an already-connected unit with status='unknown'"
- In reply to: Greg Lindahl: "Re: What compiler to use"
- Next in thread: Greg Lindahl: "Re: What compiler to use"
- Reply: Greg Lindahl: "Re: What compiler to use"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 07 Jan 2005 11:10:09 -0800
In article <41ded9b6$1@news.meer.net>, lindahl@pbm.com (Greg Lindahl)
wrote:
> In article <1105118247.034218.142480@f14g2000cwb.googlegroups.com>,
> <marvin.avery@hrdc-drhc.gc.ca> wrote:
>
> >Non-integers used for array subscripting
> >NUM parameter in READ statements
> >-- READ (IUNIT, NUM=NBYTES) LINE
> >Declarations like CHARACTER C*(*), D(N)*(*)
> >-- I have no idea what this is
>
> The last 2 are extremely unusual. The array subscripting extension
> used to be common but is getting less so over time.
Eh?
I agree that the NUM parameter is unusual - enough so that I wouldn't
have even been able to guess what it did without the clue of the nbytes
name here (and that could be leading me astray). I'd guess that this is
likely to need some porting work - I'd need more context to guess
whether the work would be trivial or not.
I'm sure I've seen the non-integer subscripts, but it has been a while.
I'd call them pretty rare, though I guess not as rare as the NUM
parameter. In principle, this is a simple porting job, though - just add
int() in the right places; the only problem is if there are a lot of
"right places" and it is hard to identify them.
So far I'm "with" you, Greg.
But the last one above looks like perfectly standard-conforming f77 to
me, and not at all rare. Of course, it is only standard-conforming in
the right context, but that's true of almost everything. Looks to me
like declarations of assumed-length character dummy arguments. Or the
scalar one could be a named constant, but I'd guess that a dummy
argument was more likely, particularly as the array case can't be a
named constant in f77 (probably ok in f90, but I don't recall ever
seeing such an assumed length array case for a named constant).
-- Richard Maine | Good judgment comes from experience; email: my first.last at org.domain | experience comes from bad judgment. org: nasa, domain: gov | -- Mark Twain
- Next message: beliavsky_at_aol.com: "Re: What compiler to use"
- Previous message: Richard E Maine: "Re: Opening an already-connected unit with status='unknown'"
- In reply to: Greg Lindahl: "Re: What compiler to use"
- Next in thread: Greg Lindahl: "Re: What compiler to use"
- Reply: Greg Lindahl: "Re: What compiler to use"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]