Appreciate all the help last time - I've got an illegal A correlative that I
need to somehow make legal. (D3 NT/Linux) Here's what I'm trying to do,
which is obviously failing miserably:

A;0(T1,5)(D4/)(G2/1):"-"0(D4/)(G0/1)):"-"0(D4/)(G1/1))

The file ID is a system(19), and I'm trying to extract the date and convert
it to YYYY-MM-DD format.

Thanks in advance!

..logic polish reverse understand don't who wimps for are correls A

F0;(T1,5);P;(D4Y);"-";:;_;P;(DM);(MR%2);"-";:;_;(DD);(MR%2);:;:

panzerboy (AT) gmail (DOT) com wrote:
POTW! =`:^>

--
frosty

On Feb 2, 2:56 pm, "frosty" <fros... (AT) bogus (DOT) tld> wrote:
My acronym circuit must be faulty, what is POTW? (or is that WTOP?)

Knowing most mere mortals prefer A correls I've tried to give the A
Correl a go
Neither
A 0(T1,5)(D4Y):"-":0(T1,5)(D4/)(G1/1):"-":0(T1,5)(D4/)(G0/1)
nor
A 0(T1-5)(D4Y):"-":0(T1-5)(DM)(MR%2):"-";0(T1-5)(DD)(MR%2)
Work, they get reject as invalid A correls.
Is there a limit on the number of processing codes you can use after a
attribute number?
A 0(T1-5)(D4Y):"-":0(T1-5)(DM):"-";0(T1-5)(DD)
Works fine but only gives single digit day & month for today, Scott
wants 2 digits.

Using the LPV allowed me to finally get a A correl that would work...
001 S
002 0
003
004
005
006
007
008 T1,5]D4/]A LPV(G2/1):"-":LPV(G1/1):"-":LPV(G0/1)

Hmm..., & you wonder why I prefer F-correls.

Jeremy Thomson

Scott - more parens - less parameters - minor corrections:

A;(0(T1,5)(D4/)(G2/1)):"-"0(T1,5)(D4-)(G0-2))

Tom

"Scott Deardorff" <smdamd (AT) bellsouth (DOT) net> wrote

Scott - different subject - system() calls.
Avoid these like the plague.
system(19) on Universe returns the logon id, so if you convert - beware.
If you absolutely must use system calls, then help yourself and wrap them in
vendor specific subroutines.
Conversions are bad enough as is, without adding to your workload.
Tom

"Tom Phillips" <squash (AT) computer (DOT) org> wrote

The real question here is why insist on the date formated as yyyy-mm-
dd? D3 does not support that via the D processing code, so you have
conditioned the users of the system to expect dates in a format for
which you will be forever building basic subroutines or complicated
correlatives (as proven by this thread) that are required everywhere
you display a date. I say use the tools the system provides for free:
give them mm-dd-yyyy (actually I prefer mm-dd-yy as it is more
compact) and then the answer is (as you probably already know) as
simple as A0(t1,5)(d4-) - which will actually be valid only for the
next 234 years, after which time the drawbacks of relying on the
embedded date in system(19) become apparent.

/Scott Ballinger
Pareto Corporation
Edmonds WA USA
206 713 6006

On Feb 1, 6:37 pm, "Tom Phillips" <squ... (AT) computer (DOT) org> wrote:

panzerboy (AT) gmail (DOT) com wrote:
Post Of The Week!

--
frosty

That doesnt work for me I get
[168] illegal a-correlative.: 'A;(0(T1,5)(D4/)(G2/1)):"-"0(T1,5)(D4-)
(G0-2))'

However
A;0(T1,5]D4/]G2/1):"-":0(T1,5]D4-U]G0-2)
works and doesnt require your system to be set for US dates.

Jeremy Thomson

On Feb 2, 3:37 pm, "Tom Phillips" <squ... (AT) computer (DOT) org> wrote:

On Feb 2, 3:31 am, panzer... (AT) gmail (DOT) com wrote:
On the subject of american dates: Why?

Is there *any* good reason for MMDDYY?

Can we get this changed please? Helpdesk?!