Re: [FWP] COBOL as Perl

[Chaim Frenkel <chaimf@pobox.com>]

>>>>> "BH" == Brand Hilton <bhilton@adc.com> writes:

BH> On Sun, Jul 04, 1999 at 12:18:22AM -0400, Chaim Frenkel wrote:
>> It's missing the processing steps. The loop has to decide which
>> stream to handle _next_. And what to do with it.

BH> Ok, I'm confused.  AFAICT, my code does this.  Do you have some test
BH> data against which it fails?  Maybe my code is just confusing because
BH> of all the negative logic.

I couldn't find the left or right leg processing path. For example

	Left		Right
	AAAAA	90	AAAA +33
	CCCCC	60	AAAA +50
	DDDDD	80	BBBB +30
			CCCC Delete

You want to end up with a resulting file

		AAAA	173
		BBBB	30
		DDDD	80

In other words, each leg, in general, needs specific processing. True,
under specific optimizations, you can collapse or reorder legs. But not
in the general case.

I'm currently trying to work out how to handle a data merge problem.
I have two large data stores, which are out of sync. (There was a lot
of manual intervention while they were being built and mirrored) I
have the original data sources (but not all). And I have a record
of what _should_ have existed. I now have to figure out which
goes where and what would be the most correct dataset to use.

I'm basically trying to generalize to an N-way merge.

<chaim>
-- 
Chaim Frenkel					     Nonlinear Knowledge, Inc.
chaimf@pobox.com				               +1-718-236-0183

==== Want to unsubscribe from Fun With Perl?  Well, if you insist...
==== Send email to <fwp-request@technofile.org> with message _body_
====   unsubscribe