Re: [Fwd: gnubol: Record delimiter clause and parse order]

[Fred Mobach <fred@mobach.nl>]

Ralph Jones wrote:

> For variable length records, does the spec. call for full word padding?  I'm trying to recall, from my mainframe days, and I believe the records were not padded?

At the mainframes I have worked on I have not seen any padding on the record level. But I remember that some other systems could only process records with a multiple of the wordsize of
memory. On disc files the last part of a block can be left unused as happens until today.


> The sequential files has to be platform independent, so you can read the file produced by any compiler on any machine.  The conversion of big little endian data is still up the
> programmer.  This does not include line sequential which is not platform independent.

I am afraid that there exists more than one format for sequential files. In a previous post in this thread I have given the format of just one of them.


> Do you still have the record type 'undefined'?  This was usually used with tape files. The I-O routines delivered a block of data into buffer up to the defined size of the buffer.  I do
> not remember what happened when a block was longer than record size.

That happens to be used in combination with ORGANIZATION IS DIRECT and SEQUENTIAL. It is still in use.

Regards,

Fred


--
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.