Inserting records based on conditions

vickey_dw · New User Joined: 10 Jun 2015 Posts: 51 Location: India

Hi All

My input data set 1 example

Nic Clouston · Posted: Wed Feb 22, 2017 2:28 pm

You can easily use the code tags - either hand-craft them or use the full editor and the facilities that it provides.

I do not understand the logic - after the first record from data set 1 you add 3 record from data set 2, after the second record from data set 1 you have indicated oe record from data set 2 and ditto for the third record. Why 3 - 1- 1? What rules?

I will code your original post for you.

Bill Woodger · Posted: Wed Feb 22, 2017 2:30 pm

Why would you be unable to use the code tags?

There has to be some condition under which you'd know what records go where. I can't discern the condition from the data sample, other an "numbers only" after A, "not letters or numbers" after B and "letters only" after C.

Even if that is what you want, is your data "regular" within that (no other mixtures possible) and is your data "in order" within that?

Or some completely different conditions?

vickey_dw · New User Joined: 10 Jun 2015 Posts: 51 Location: India

Thanks for the reply!!!

My requirement is to add records first 10 from data set 2 once it finds AAA in data set 1.Then add next 10 record a from data set 2 once it finds BBB in data set 2.

So out will like

Nic Clouston · Posted: Wed Feb 22, 2017 4:12 pm

I have corrected your code tags - end tags are denoted by a '/' between the '[' and the 'code'. Using the full editor buttons this will be done automatically for you. Use the 'Preview' button to check your post before submitting it. I have also changed your 'file' to 'data set' as you are dealing with data sets on the mainframe - not files.

Arun Raj · Posted: Wed Feb 22, 2017 9:02 pm

vickey_dw · New User Joined: 10 Jun 2015 Posts: 51 Location: India

only two AAA and BBB records are there in the dataset 1.
Dataset two has 20 records out of which first 10 i need to copy after AAA and next 10 after BBB.
These values wont change in run time also.

Dataset1,Dataset 2 and outfile all have LRECL 80 and RECFM is FB

Arun Raj · Posted: Wed Feb 22, 2017 10:21 pm

Thanks. Can you try answering the questions you skipped?

vickey_dw · New User Joined: 10 Jun 2015 Posts: 51 Location: India

I think i missed below question

Bill Woodger · Posted: Wed Feb 22, 2017 11:49 pm

JOINKEYS.

Use JNF1CNTL, INREC, WHEN=INIT to assign a sequence number to each (of the two) record.

Use JNF2CNTL, INREC, WHEN=GROUP, RECORDS=10 to assign ID for each group.

The key is the sequence and ID.

SORTED,NOSEQCK

REFORMAT contains all original data for F1 and F2.

In the main task, INREC, WHEN=GROUP, RECORDS=10, append a SEQ.

OUTFIL, IFTHEN=(WHEN=(logical expression to test for one in the SEQ-field appended, BUILD with the slash-operator (/) to output two records (from the F1 and the F2 parts of the REFORMAT record). WHEN=NONE to just BUILD from from the F2 part of the REFORMAT record.