MVSFORUMS.com

singhnarender79 · Posted: Mon May 23, 2005 11:21 am Post subject:

Yes, you have got it right, I want the output records to be ordered according to the key.
_________________
"Hold fast to dreams, for if dreams die, life is a broken winged bird that cannot fly."

-- Langston Hughes

Frank Yaeger · Posted: Mon May 23, 2005 1:04 pm Post subject:

Naren,

Here's a DFSORT/ICETOOL job that will do what you asked for. You'll need DFSORT's Dec, 2004 PTF to use the new IFTHEN, OVERLAY, KEEPBASE and VLENOVLY functions.

singhnarender79 · Posted: Tue May 24, 2005 3:21 pm Post subject:

Hi Frank,

I tried to analyse the code you had send across as I was not able to execute it on our system (The PTF is getting installed on 29 May).

I took the following i/p files

File 1
X'01' Pol1
X'02'
X'02'
X'01' Pol2
X'02'
X'02'

File 2
X'01' Pol2
X'02'
X'02'
X'01' Pol4
X'02'
X'02'

Our T3 file looks like this...

Frank Yaeger · Posted: Tue May 24, 2005 5:29 pm Post subject:

Naren,

What you're saying makes no sense to me. You say "On using the SPLICE from T3 to OUT, I am not getting the expected result". But you also say you don't have the Dec, 2004 PTF installed yet. So how can you be running the SPLICE operator when it uses KEEPBASE and VLENOVLY which you can't use yet because you don't have the PTF?

You told me in an earlier post that your "key" starts in position 25, is 8 bytes long and consists of these four fields:

MP-POL-DEPT PIC S9 COMP-3.
MP-POL-LET PIC XX.
MP-POL-YEAR PIC S9(3) COMP-3.
MP-POL-SERIAL PIC S9(5) COMP-3.

That's what I set up the job to use for the keys.

I don't know what your input of:

X'01' Pol1
...

represents, but it doesn't look like the keys you described and for which the job is set up.

Also, T3 is a MOD data set used as the output for TWO operators. I have no idea which operator your T3 records resulted from. Again, if it's the output of the SPLICE operator, how did you use KEEPBASE and VLENOVLY when you don't have the PTF installed yet?

I don't really understand what it is you're doing or what you want me to analyze. I tested the job using input I created based on what you told me about the input. The job gave the expected output. If you want me to try the job with different input, you're going to have to supply that input in its exact form to me.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

Frank Yaeger · Posted: Tue May 24, 2005 5:38 pm Post subject:

Naren,

BTW, I assumed that your 8-byte key starts in position 25 counting the 4-byte RDW. That is, it looks like this:

singhnarender79 · Posted: Wed May 25, 2005 3:41 am Post subject:

Frank,

I was trying to understand the code which you had given me. Since you have used some new functions, I had looked for them in the Application Programming Guide and tried to understand what exactly they are doing in code you gave me. And since the PTF is not yet installed, I tried to figure out the output manually (On Paper). (i.e by writing down some input records, then figuring out what each step was doing and writing the output from each step)

The key which i mentioned is right and it starts from 25 (excluding the RDW, but that we had figured out) i.e

P-POL-DEPT PIC S9 COMP-3.
MP-POL-LET PIC XX.
MP-POL-YEAR PIC S9(3) COMP-3.
MP-POL-SERIAL PIC S9(5) COMP-3.

What I mentioned above was the first two bytes that comes before the key

X'0101'SMOPHRT
X'0202'SMOPHNAM
X'0302'SMOPHADD
X'0302'SMOPHADD
X'0402'SMOPHOCC

and I think u have also used that (Correct me if i am wrong)

Frank Yaeger · Posted: Wed May 25, 2005 10:12 am Post subject:

Yes, I used the X'01' in position 5 as the indicator that a new group was starting.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

singhnarender79 · Posted: Tue May 31, 2005 5:11 am Post subject:

Frank,

We have got the new PTF installed and I have executed the code you gave me. I am very happy to say that it's working like a charm.

and also i found out, where i had done a mistake while doing it manually.

Thanks for all the help.

Cheers,
Naren
_________________
"Hold fast to dreams, for if dreams die, life is a broken winged bird that cannot fly."

-- Langston Hughes

Frank Yaeger · Posted: Tue May 31, 2005 10:07 am Post subject:

Thanks for the feedback. I'm glad I could help.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

singhnarender79 · Posted: Thu Jun 02, 2005 11:17 am Post subject:

Frank,

There is one more interesting development that has come in my requirement. I have a workaround for that, but i was just wondering if it can be possible using DFSORT. So, if u can help us out it would be great.

The orinal req. is same but with one more addition. The addition is that when we are merging the 2 files, giving priority to records from file 2(which are also present in file 1), we need to extract some info (a date field) from the record in File1 and overlay it on the same record from file 2, and put it in the output file. If this is confusing, here is an example which explains it....

Pol means Policy field.

Input File 1

..SMOPHRT Pol1
..SMOPHNAM date1
..SMOPHADD
..SMOPHRT Pol2
..SMOPHNAM date2
..SMOPHADD

Input File 2
..SMOPHRT Pol3
..SMOPHNAM date3
..SMOPHADD
..SMOPHRT Pol2
..SMOPHNAM date4
..SMOPHADD

Output File

..SMOPHRT Pol1
..SMOPHNAM date1
..SMOPHADD
..SMOPHRT Pol2
..SMOPHNAM date2
..SMOPHADD
..SMOPHRT Pol3
..SMOPHNAM date3
..SMOPHADD

As you can see above Pol2 was common in both but taken from File2, and only the date has to be taken from Pol2 record from file1. So in the Output file, date4 was replaced by date2 for Pol2 record. As you can see above, date is not present in the ROOT record but in the next record.

The date field is defined as PIC S9(4) COMP and the Position is 56 (excluding RDW).

Thanks,
Naren
_________________
"Hold fast to dreams, for if dreams die, life is a broken winged bird that cannot fly."

-- Langston Hughes

Frank Yaeger · Posted: Thu Jun 02, 2005 11:54 am Post subject:

So you want the date from the X'02' record in file1 for matching groups? Can it be the entire X'02' record in file1, or does it have to be just the date from the X'02' record in file1 and the rest of the fields from the X'02' record in file2?
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

singhnarender79 · Posted: Thu Jun 02, 2005 12:31 pm Post subject:

It has to be just the date from the X'02' record in file1 and the rest of the fields from the X'02' record in file2 for the matching group only.
_________________
"Hold fast to dreams, for if dreams die, life is a broken winged bird that cannot fly."

-- Langston Hughes

Frank Yaeger · Posted: Thu Jun 02, 2005 1:15 pm Post subject:

Naren,

I think I've come up with a way to do it, but it will take me a little time to work it out. I'll post the solution later today.
_________________
Frank Yaeger - DFSORT Development Team (IBM)
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
DFSORT is on the Web at:
www.ibm.com/storage/dfsort

Frank Yaeger · Posted: Thu Jun 02, 2005 2:35 pm Post subject:

Naren,

Here's the revised DFSORT/ICETOOL job to handle the new requirement.

singhnarender79 · Posted: Mon Jun 06, 2005 6:28 am Post subject:

Hi Frank,

Thanks again for all the help. It is also working perfectly, but had to modify it slightly to get the records in the same order as in the input file. (I used the sequence number in sort rather then the Binary values).

Cheers,
Naren
_________________
"Hold fast to dreams, for if dreams die, life is a broken winged bird that cannot fly."

-- Langston Hughes