Need help to develop sort card or easytrive:

Vidya Kajale · New User Joined: 01 Jun 2010 Posts: 43 Location: India

Hi,

I have file1 having diffent fields in below format:
Plastic no. , IA cd, Country cd , ZIP , Currency cd, Fees, Currency fees

the data in these fileds is been repeated for same plastic no. 2 times

And file 2 having fileds as plastic no. ,country cd,currency cd and currency fees for each plastic no. repeated 2 times for two different countries.

I need to put data from file2 to file1 for each plastic no. present in file1 for the fields Country cd, currency cd and currency fees.

Can i achive these using easytrive or sort card.
Please educate me on these.

chandan.inst · Posted: Mon Feb 04, 2013 4:26 pm

Hi,

What you have tries so far?

Check for Sort Join keys. Your requirement seems to be easily doable with JOINKEYS

Regards,
Chandan

Pandora-Box · Posted: Mon Feb 04, 2013 4:33 pm

Hi Vidya,

You could explain your requirement in simpler and better by posting the sameple records of
file1 and file2 and just post the needed output file quoting the keys

Ofcourse,while doing so please do it within code tags

Vidya Kajale · New User Joined: 01 Jun 2010 Posts: 43 Location: India

Hi,

If i need to repeate it 200 times in the same file for same platic no. for 200 different countries shall i write sort join 200 times for each country code.

enrico-sorichetti · Posted: Mon Feb 04, 2013 4:42 pm

1) tell us what sort product You are using ...
( ICE .... IBM DFSORT, WER... SYNCSORT )

2) please refrain from using local jargon and filed names

we take the trouble of understanding Your problem...
please spare us the trouble of decrypting Your organization jargon

post ( as usual for <sort> requests ... )

recfm/lrecl of the input/output file

position/length/format of the fields of interest ( no cobol jargon please )

clear description of the process

sample of the input file[s]

example of the outputs expected

Vidya Kajale · New User Joined: 01 Jun 2010 Posts: 43 Location: India

Hi,
We are using sort product as SYNCSORT.
file1:

Bill Woodger · Posted: Mon Feb 04, 2013 5:43 pm

Which version of SyncSort?

Do you have support for JNFnCNTL files for JOINKEYS?

Pandora-Box · Posted: Mon Feb 04, 2013 6:04 pm

Bill,

From the input and output

I dont think TS needs to use JNFnCNTL

enrico-sorichetti · Posted: Mon Feb 04, 2013 6:08 pm

I am not a <sort> expert ...
but the logic tells me that the <TS> will have to add a sequence number to each entry in each group in order to match in strict sequence

Pandora-Box · Posted: Mon Feb 04, 2013 6:31 pm

Enrico,

The <TS> needs to join based on Plastic no. and consider the file is sorted so he gets the output as is.(If I understand <TS> correctly)

enrico-sorichetti · Posted: Mon Feb 04, 2013 6:34 pm

NOPE

each <plastic> contains three entries and even if the TS posted irrelevant info
I would assume ( to be on the safe side ) 1<>1 2<>2 3<>3

Pandora-Box · Posted: Mon Feb 04, 2013 6:45 pm

Ah I was so dumb on that .. I missed out a cartesian product

Thanks for the clarification Enrico

Apologies Bill

Bill Woodger · Posted: Mon Feb 04, 2013 6:45 pm

The key is a six digit number. There are three on each file. Which will get you nine joins per key. If one file has duplicates removed (information is duplicated, so should not be a problem) then it will make processing more straighforward. Even de-dupe both, if possible, and output three records per match at the end.