IBM Mainframe Forum

by **ranga_subham** » Tue Jun 05, 2012 7:31 pm

Hi,

Request you to convert the below given two step solution into one step ICETOOL solution. We are using "Z/OS DFSORT V1R10".

Select all

//*=======================================================
//STEP0001 EXEC PGM=SORT
//SORTIN DD DSN=INPUT.GDG.FILE(0),DISP=SHR
//SORTOUT DD DUMMY
//SORTXSUM DD DSN=&&TEMP1,DISP=(NEW,PASS,DELETE),
// UNIT=SYSDA,SPACE=(CYL,(100,100),RLSE)
//SYSOUT DD SYSOUT=*
//SYSIN DD *
INCLUDE COND=(40,3,CH,NE,C' ')
INREC FIELDS=(26,10,X,40,3,X,36,3)
SORT FIELDS=(1,14,CH,A),EQUALS
SUM FIELDS=NONE,XSUM
//*=======================================================
//STEP0002 EXEC PGM=SORT
//SORTJNF1 DD DSN=INPUT.GDG.FILE(0),DISP=SHR
//SORTJNF2 DD DSN=&&TEMP1,DISP=(OLD,PASS,DELETE)
//SORTOUT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SYSIN DD *
SORT FIELDS=COPY
JOINKEYS FILE=F1,FIELDS=(26,10,A,40,3,A)
JOINKEYS FILE=F2,FIELDS=(1,10,A,12,3,A)
REFORMAT FIELDS=(F1:1,43)
/*

Please help.

Thanks.

by **skolusu** » Tue Jun 05, 2012 9:23 pm

The above job doesn't make sense at all.

1. You are reading latest generation of gdg file and including the records that are greater than spaces in pos 40 for 3 bytes.

2. You are creating a new record with 10 bytes from pos 26 and 3 bytes from pos 40 as a key. I have no idea as to why you are using the 3 bytes from pos 36 as it not used any where.

3. You sum sort on the newly created 14 bytes and write out the duplicates to SORTXSUM.

4. Now you take the same latest generation of gdg file(which is the same file you used to get the sortxsum dataset) and match it with the SORTXSUM dataset using JOINKEYS? Aren't you getting a many to many match as the sortxsum dataset has only the duplicates? What is the purpose of this JCL?

It would be a good idea to show us a sample of 10 unique records (which may have duplicates) and also show us the desired output. I am guessing the lrecl's of the file is 43 and RECFM=FB. Is that correct?

And for the record it can be done in a single pass of data.

by **ranga_subham** » Fri Jun 08, 2012 2:17 pm

Hi,

The job has definitely produced results for me......that way it made some sense........however, its a roundabout method because I am NOT a SORT expert and seeking your advice

The GDG file has lots of duplicates and I got one of them into SORTXSUM and using that as input again I am trying to get all the duplicates......I know I have not done a clean job but got the results in a crooked manner.

Input:

Select all

----+----1----+----2----+----3----+----4----+
20120606221337705765ASRTA ABC123AGT OO1
20120606221337705765ASRTA ABC123AGV OO2
20120606221337705765ASRTA ABC123APC
20120606221337705765ASRTA ABC123ATU KUR
20120606221337705765ASRTA ABC123AKJ PIC
20120606221337705765ASRTA ABC123ALU OO2

The key fields are at 30th(6 characters), 36th(3 characters) and 40th(3 characters) positions. The main ones are fields at 30th and 40th columns. My expected output is as shown below:

Select all

----+----1----+----2----+----3----+----4----+
20120606221337705765ASRTA ABC123AGV OO2
20120606221337705765ASRTA ABC123ALU OO2

Thanks.

by **skolusu** » Fri Jun 08, 2012 10:12 pm

Ranga_subham,

You definitely have complicated a simple request. You just need ALLDUPS after eliminating any records with spaces.

Here is a simple 1 pass ICETOOL solution. Your JCL showed the key as starting from 26 for a length of 10 bytes , but your latest post shows that the key starts from pos 30th for a length of 6 bytes. Your sample data shows that they byte 26-29 are just spaces , so I took the key as starting from pos 30 for a length of 6 bytes.

Select all

//STEP0100 EXEC PGM=ICETOOL
//TOOLMSG DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//IN DD *
----+----1----+----2----+----3----+----4----+----5----+----6----+----
20120606221337705765ASRTA ABC123AGT OO1
20120606221337705765ASRTA ABC123AGV OO2
20120606221337705765ASRTA ABC123APC
20120606221337705765ASRTA ABC123ATU KUR
20120606221337705765ASRTA ABC123AKJ PIC
20120606221337705765ASRTA ABC123ALU OO2
//OUT DD SYSOUT=*
//TOOLIN DD *
SELECT FROM(IN) TO(OUT) ON(30,6,CH) ON(40,3,CH) ALLDUPS USING(CTL1)
//CTL1CNTL DD *
INCLUDE COND=(40,3,CH,NE,C' ')
//*

The output from this job is

Select all

20120606221337705765ASRTA ABC123AGV OO2
20120606221337705765ASRTA ABC123ALU OO2

by **ranga_subham** » Sun Jun 10, 2012 10:33 am

Hi,

Thanks a lot.

IBM Mainframe Forum

help me to convert this to ICETOOL code

help me to convert this to ICETOOL code

Re: help me to convert this to ICETOOL code

Re: help me to convert this to ICETOOL code

Re: help me to convert this to ICETOOL code

Re: help me to convert this to ICETOOL code