Re: gnubol: Record delimiter clause and parse order

[Tim Josling <tej@melbpc.org.au>]

Matthew Vanecek wrote:
> 
> For shits-n-grins, and personal aggrandizement, I did a little test in
> MF today at work.  Sequential is sequential, with no delimiters,
> assuming fixed length records.  Line sequential is delimited by "\r\n"
> (hex 0D 0A).  At least, that's how they come out on OS/2.  So, for fixed
> length sequential, that's the way I'll do it.  Line sequential I'll
> throw in because it seems that so many non-mainframe compilers seem to
> advertise support it--"Shop here!! We support MicroFocus' non-standard
> proprietary extensions!!"
> 

Good.

> > Someone mentioned that fixed vs variable changes the filling of
> > the records. True.
> >
> 
> I also did a couple of variable length writes, so I could get an idea of
> what MF does with writing variable length files.  I'm hard pressed to
> decipher them, however, even with a hex editor.  I've attached it, and
> the program I used to write it.  I can understand the basic
> structure--starts out at 6 bytes, then increases to 10 bytes, and
> presumably would add another 4 bytes when rec-len reaches 14, to give 18
> bytes, and so on. Any pointers on interpreting would be very much
> appreciated.  We don't really use variable length records in files.
> Actually, we do much more DB work than we do file IO at all...

Something went wrong with the output file you sent. Could you
send to me direct and call the attachment vr.bin or something
that cannot be text (ie not vr.txt). - I think the mailing list
may be messing it up. I remember someone decribing the MF format
once, I hope I kept it... will have a look. Anyway I am very
comfortable with your approach....

Tim Josling

--
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.