MVSFORUMS.com

coolman · Last edited by coolman on Sat Feb 05, 2011 1:41 am; edited 1 time in total

Consider the following input

Frank Yaeger · Posted: Wed Mar 30, 2005 4:52 pm Post subject:

It's not clear what you're trying to do. All I see from your example is that you removed the header/detail records for RPT2 and RPT4. Are you trying to remove the second, fourth, sixth, etc report lines, or are you trying to do something else?
Please be more clear about what you want. Also, what is the RECFM and LRECL of your input file?
_________________
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

coolman · Posted: Wed Mar 30, 2005 5:09 pm Post subject:

Frank,

Thanks for your quick response. All I'm trying is filter out only specific sections of the report. In this example, I want RPT1 header followed by its detail records until the next report header. Then RPT3 followed by its detail records until the next and so on. I know upfront what are the different header that I want. Here in this example, its RPT1, RPT3 and RPT5.

The LRECL/RECFM can be 80/FB.
________
buy vaporgenie

kolusu · Posted: Wed Mar 30, 2005 5:28 pm Post subject:

Coolman,

There is similar thread about sorting headers and detail records. Check this link

http://www.mvsforums.com/helpboards/viewtopic.php?t=3432

The following solution is based of that trick.

coolman · Posted: Wed Mar 30, 2005 6:06 pm Post subject:

Kolusu,

Thanks for the quick solution. I already looked at that posting in mainframe challenge.

What if the report headers are not RPT1, RPT3, RPT5 and are something like this

RPT.012, RPT.023. RPT.045 which needs to be filtered.

The file contains report id's that vary from RPT.001 to RPT.100 within the same file

Please suggest. Can you also possibly show a EZTRIVE solution?
________
vaporizer affiliate programs

coolman · Posted: Wed Mar 30, 2005 6:08 pm Post subject:

On the same lines, is there a way, a header can be propapaged to all of the details records?

Say

HDR1
DET
DET
DET
HDR2
DET

and the o/p as
HDR1
DET HDR1
DET HDR1
DET HDR1
HDR2
DET HDR2
________
Chevrolet Big-Block engine specifications

Frank Yaeger · Posted: Wed Mar 30, 2005 7:44 pm Post subject:

Coolman,

Do you want to "filter out" the sections by hardcoding the report header id (e.g. 'RPT.012', 'RPT.023', etc) as an INCLUDE condition or some other way?
_________________
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

coolman · Posted: Thu Mar 31, 2005 9:24 am Post subject:

Frank,

Okay. It appear that I'm not clear. I have a report which has a bunch of report id's which vary from 1 to 100. Now, there are times, when I would have to filter only some specific report id's such as RPT.012, RPT.023 as in this example. I do not want to hardcode these values as the header containing these report id's may have different headings. So, I would need those header records(which contains the report id) followed by the list of details records until I hit another report id. Now, if the subsequent report id matches my criteria, I would be writing those to the output else I would have to skip and this processing needs to be done until the end of the file.

Hope I have made it clear this time
________
one vaporizer

Frank Yaeger · Posted: Thu Mar 31, 2005 12:17 pm Post subject:

What identifies a report id (that is, the first record of a report section)? Does it have 'RPT' as the first three characters or something else? If there's nothing that identifies a report id, then how do we know where a report section starts?

You say you don't want to hardcode these values (the report ids), but you don't say how you want to input the values to the job. For example, if you want the reports for ids 'RPT.012' and 'RPT.023', how are you going to communicate that to the job? I suggested INCLUDE because you could put whatever report ids you need for that particular run into the INCLUDE statement. Alternatively, you could set the report ids up as DFSORT symbols and use those symbols in the INCLUDE statement. Or do you have another way of communicating the report ids you want to the job that you haven't made clear yet? Like just the number of the report (e.g. 12, 23), and again how are you going to communicate that to the job?

coolman · Posted: Thu Mar 31, 2005 1:52 pm Post subject:

Frank,

Your assumption is right. I would know the report id's when the job is executed (which is the criteria) such as RPT.012, RPT.023. Please let me know if you need any more information.

Cheers,
Coolman
________
headshop

coolman · Posted: Thu Mar 31, 2005 2:20 pm Post subject:

As you stated, we can have them as a INCLUDE statement or use DFSORT symbols

Cheers,
Coolman
________
herbalaire

Frank Yaeger · Posted: Thu Mar 31, 2005 3:57 pm Post subject:

Coolman,

Below is a DFSORT/ICETOOL job that will do what I think you want. You'll need z/OS DFSORT V1R5 PTF UQ95214 or DFSORT R14 PTF UQ95213 (Dec, 2004) to use the new DFSORT/ICETOOL IFTHEN, OVERLAY and KEEPBASE functions.

coolman · Posted: Thu Mar 31, 2005 4:34 pm Post subject:

As always, Thanks Frank for your assistance. But I don't have the latest PTFs with me to use this one.

Cheers,
Coolman
________
los angeles dispensary

Frank Yaeger · Posted: Thu Mar 31, 2005 5:36 pm Post subject:

Coolman,

Ask your System Programmer to install the Dec, 2004 PTF (it's free).
_________________
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