IBM Mainframe Forum

by **bond_ajay** » Wed Apr 02, 2008 10:27 am

Below is the requirement for which we want to know if we can achieve it in 1 step using a DFSORT - splice/icetool? The headings used, namely, Item, Location, RTL and CST are not part of the file but only indicative.

File -1: (key is Item(6bytes) + Location(4bytes). CST field is a dump from a table and has a assumed 4 decimal spaces)

Select all

Item Location RTL CST
55037 5001 3198.4 2286.752
283846 5001 166301.52 129381.0672
330143 5001 30054.99 23547
330178 5001 30474.92 23876

File -2: (key is Item (6bytes)+ Location(4bytes). CST field is a dump from a table and has a assumed 4 decimal spaces)

Select all

Item Location RTL CST
55037 5001 3198.4 1286.752
283846 5001 166301.52 119381.06
330143 5001 30054.99 23547
330000 5001 30474.92 23876

Requirement:
1. We need to compare the key fields (item + location) of the two files, file-1 and file-2. for the matching records we need to write an output record which will have the "ITEM", "LOCATION" and a difference of CST field of file-1 and CST field of File-2. (CST-File-1 - CST-File-2).

2. When the record is found on file-1 and not on file-2, we need to write an output record which will have "ITEM", "LOCATION" and a negated value of the CST field of file-1.

3. When the record is found on file-2 and not on file-1, we need to write an output record into a different output file which will have "ITEM", "LOCATION" and "CST" fields as in the file.

Expected OUTPUT:
Output-File-1:

Select all

Item Location CST
55037 5001 1000.0000
283846 5001 10000.0072
330143 5001 0.0000
330178 5001 -23876

Output-File-2:

Select all

Item Location CST
330000 5001 23876

by **skolusu** » Thu Apr 03, 2008 2:32 am

The following DFSORT/ICETOOL JCL will give you the desired results

Select all

//STEP0100 EXEC PGM=ICETOOL
//TOOLMSG DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//IN1 DD *
----+----1----+----2----+----3----+----4----+
55037 5001 3198.4 2286.752
283846 5001 166301.52 129381.0672
330143 5001 30054.99 23547
330178 5001 30474.92 23876
//IN2 DD *
55037 5001 3198.4 1286.752
283846 5001 166301.52 119381.06
330143 5001 30054.99 23547
330000 5001 30474.92 23876
//T1 DD DSN=&&T1,DISP=(MOD,PASS),SPACE=(CYL,(1,1),RLSE)
//OUT1 DD SYSOUT=*
//OUT2 DD SYSOUT=*
//TOOLIN DD *
COPY FROM(IN1) USING(CTL1)
COPY FROM(IN2) USING(CTL2)
SORT FROM(T1) USING(CTL3)
//CTL1CNTL DD *
INREC PARSE=(%00=(ABSPOS=31,ENDAT=C'.',FIXLEN=6),
%01=(FIXLEN=4)),
BUILD=(1,15,C'+',%00,UFF,EDIT=(TTTTTT),C'.',%01)
OUTREC OVERLAY=(24:24,4,TRAN=ALTSEQ)
ALTSEQ CODE=(40F0)
OUTFIL FNAMES=T1,BUILD=(01,15,16,12,SFF,PD,LENGTH=8,24:C'1')
//CTL2CNTL DD *
INREC PARSE=(%00=(ABSPOS=31,ENDAT=C'.',FIXLEN=6),
%01=(FIXLEN=4)),
BUILD=(1,15,C'-',%00,UFF,EDIT=(TTTTTT),C'.',%01)
OUTREC OVERLAY=(24:24,4,TRAN=ALTSEQ)
ALTSEQ CODE=(40F0)
OUTFIL FNAMES=T1,BUILD=(01,15,16,12,SFF,PD,LENGTH=8,24:C'2')
//CTL3CNTL DD *
OPTION EQUALS
SORT FIELDS=(01,12,CH,A)
SUM FIELDS=(16,8,PD,24,1,ZD)

OUTFIL FNAMES=OUT1,INCLUDE=(24,1,SS,EQ,C'1,3'),
IFTHEN=(WHEN=(24,1,CH,EQ,C'1'),
BUILD=(1,15,16,8,PD,EDIT=(SIIIIIT.TTTT),SIGNS=('-',' ',,))),
IFTHEN=(WHEN=(24,1,CH,EQ,C'3'),
BUILD=(1,15,16,8,PD,EDIT=(SIIIIIT.TTTT),SIGNS=(' ','-',,)))

OUTFIL FNAMES=OUT2,SAVE,
BUILD=(1,15,16,8,PD,EDIT=(IIIIIIT.TTTT))

The output from the above job is

Out1 output

Select all

283846 5001 10000.0072
330143 5001 0.0000
330178 5001 -23876.0000
55037 5001 1000.0000

out2 output

Select all

330000 5001 23876.0000

by **bond_ajay** » Thu Apr 03, 2008 11:30 am

Thanks for the help skolusu.

I have just one doubt. If i remove the column RTL (and its values are removed from the file). So we will effectively have only Location, Item and the CST fields. If so, can we reduce the complexity of the icetool? what will the Sort card look like. Please reply with your comments. Thanks.

by **skolusu** » Thu Apr 03, 2008 11:34 pm

bond_ajay wrote:Thanks for the help skolusu.

I have just one doubt. If i remove the column RTL (and its values are removed from the file). So we will effectively have only Location, Item and the CST fields. If so, can we reduce the complexity of the icetool? what will the Sort card look like. Please reply with your comments. Thanks.

Bond_ajay,

Removing the RTL column isn't going to help. The complexity of the job depends on the way u have the input data and the way you want the output. And btw dont just get confused with the control cards. It is only using 1 sort pass and 2 copy operations which are quite efficient.

IBM Mainframe Forum

DFSORT - Need 1 step solution?

DFSORT - Need 1 step solution?

Re: DFSORT - Need 1 step solution?

Re: DFSORT - Need 1 step solution?

Re: DFSORT - Need 1 step solution?