gnubol: Name is Select/Assign clause

["William M. Klein" <wmklein@ix.netcom.com>]

I was just cleaning up some old notes and read this.  FYI, on IBM mainframes
(today), *EITHER* a "data-name" or a "literal" in the Select/Assign clause
points to a DDName.  (actually, on CMS, there is a way to "convert" it to a
physical file name - but that is fairly unusual.) OTOH, in Micro Focus
(MERANT) COBOL - if you have a literal, it is always a "physical file" name
(the thing known to the operating system - this may or many not "assume" the
local directory).  But a data-name may *either* point to an environment
variable - or a data-name (in Working-Storage - for example) that had the
physical file name as its value.  Which of these last two options is meant in
Micro Focus COBOL - depends upon a compiler directive.

P.S.  If possible, would people please not use "MF COBOL" to describe EITHER
"Micro Focus COBOL" or "Mainframe COBOL" - as it does tend to be ambiguous.

> -----Original Message-----
> From: owner-gnu-cobol@wallace.lusars.net
> [mailto:owner-gnu-cobol@wallace.lusars.net]On Behalf Of Tim Josling
> Sent: Thursday, May 25, 2000 2:38 PM
> To: gnu-cobol@lusars.net
> Subject: Re: gnubol: Record delimiter clause and parse order
>
>
> Matthew,
>
> In the standard the meaning of implementer defined name (assign
> to printer) and literal-1 (assign to "xxx") are up to the
> implementer. The interpretation below is MF's definition. Other
> compilers do it differently. You can get the cobol 2002 standard
> off the links off my web page and it has a list of changes from
> cobol 85. MF's docs should say what is non-standard and what is
> vendor defined. If you have the MF docs that is acually a big win
> :-).
>
> From memory, IBM Cobol uses the imlementer defined name as a dd
> name in the JCL (similar to an environment variable, points to
> the file). I would want to be able to be compatible with that.
> Just try and keep it flexible and not hardwire any one compiler's
> interpretation - especially any nonstandard keywords need to be
> dialect dependent in the keyword table.
>
> Tim Josling
>
> Matthew Vanecek wrote:
> >
> > Tim Josling wrote:
> > >
> > > Matthew,
> > > What's a "device"? Where does it say in quotes is a device and
> > > otherwise it's a file?
> > > Tim Josling
> > >
> >
> > Backwards:
> > > (i.e., if it's in quotes, it's a file, if it's not, it's a device, but
> > > if it's not a device, it could be a file...wow, how twisted...)
> >
> > I was reading that in the MF System Reference, File Handling chapter or
> > somewhere in their manuals.  It's all a haze now...I'd have to find it
> > again for the exact wording, but IIRC, quoted identifiers refer to
> > files, and unquoted ones refer to devices (e.g., PRINTER or stdout/in
> > (32-bit only, I think)), or tape devices, etc.
> >
> > Like I said, that's (devices) too far down the road to worry about right
> > now.  Just setting up a basic file structure and getting that to work
> > correctly (or, "as expected"), and supporting as many idiosyncratic
> > implementations as is sane, is enough for the next little while.
> > --
> > Matthew Vanecek
>
> --
> This message was sent through the gnu-cobol mailing list.  To
> remove yourself
> from this mailing list, send a message to majordomo@lusars.net with the
> words "unsubscribe gnu-cobol" in the message body.  For more information on
> the GNU COBOL project, send mail to gnu-cobol-owner@lusars.net.


--
This message was sent through the gnu-cobol mailing list.  To remove yourself
from this mailing list, send a message to majordomo@lusars.net with the
words "unsubscribe gnu-cobol" in the message body.  For more information on
the GNU COBOL project, send mail to gnu-cobol-owner@lusars.net.