Syncsort - Remove Duplicates until encountering a '('

vjai6977 · New User Joined: 08 Aug 2008 Posts: 16 Location: Chennai

Hi All,

I am using the following JCL to remove the duplicates.

//STEP01 EXEC PGM=SORT
//SYSOUT DD SYSOUT=*
//SORTIN DD DISP=SHR,DSN=TST01.TUSER.INFILE
//SORTOUT DD DSN= TST01.TUSER.OUTFILE,
// UNIT=DISK,DISP=(,CATLG,DELETE),
// DCB=(RECFM=FB,LRECL=133,BLKSIZE=0),
// SPACE=(CYL,(1,5),RLSE)
//SORTWK01 DD UNIT=DISK,SPACE=(CYL,100)
//SORTWK02 DD UNIT=DISK,SPACE=(CYL,100)
//SYSIN DD *
SORT FIELDS=(10,25,A),FORMAT=CH
INCLUDE COND=(3,5,CH,EQ,C'TSTIN')
SUM FIELDS=NONE
/*

SortIn / SortOut file definition�s for the above JCL:

Organization . . . : PS
Record format . . . : FB
Record length . . . : 133
Block size . . . . : 27930

ON THE BELOW EXAMPLES THE EACH ASTRIC �*� IS A BLANK SPACE. JUST FOR READABILITY I HAD PUT IT DOWN.

Input file:

123456789012345678901234567890123456789012345678901234567890
__3___7__0_________0_________0____5______________________133
**TSTIN**TEST01.FEED.FILE(-1)
**TSTIN**TEST01.FEED.FILE(-8)
**TSTIN**TEST01.FEED.FILE(0)
**TSTIN**TEST01.FEED.FILE(-4)
**TSTIN**TEST01.OUT1.FILE001(-1)
**TSTIN**TEST01.OUT1.FILE001(-1)
**TSTIN**TEST01.OUT2.FILE(-1)

Required Output:

123456789012345678901234567890123456789012345678901234567890
__3___7__0_________0_________0____5______________________133
**TSTIN**TEST01.FEED.FILE(-4)
**TSTIN**TEST01.OUT1.FILE001(-1)
**TSTIN**TEST01.OUT2.FILE(-1)

As according to the sysin parameters supplied on the above JCL the syncsort considers the version limit like (-1), (-8) etc., after every dataset and the present output file looks like below.

123456789012345678901234567890123456789012345678901234567890
__3___7__0_________0_________0____5______________________133
**TSTIN**TEST01.FEED.FILE(-1)
**TSTIN**TEST01.FEED.FILE(-8)
**TSTIN**TEST01.FEED.FILE(0)
**TSTIN**TEST01.FEED.FILE(-4)
**TSTIN**TEST01.OUT1.FILE001(-1)
**TSTIN**TEST01.OUT2.FILE(-1)

But the first 4 records has a redundant datasets, irrespective of the different values in �()�. My requirement is sort the input file for all rows between columns 10+25, for each column position, the sort should consider the text before '(' only.

This way i can exclude redundant datasets. I welcome any suggestions.

Regards,
Jai

arcvns · Posted: Tue Sep 16, 2008 10:51 am Post subject:

vjai6977,

If you have Syncsort for z/OS Release 1.3 , you could use the PARSE statement to achieve this. Do a forum Search for more examples.

http://ibmmainframes.com/viewtopic.php?t=25044

Thanks,
Arun

Sambhaji · Posted: Tue Sep 16, 2008 12:17 pm Post subject:

arcvns · Posted: Tue Sep 16, 2008 1:01 pm Post subject:

Hello,

PARSE will definitely work if you have Syncsort for z/OS Release 1.3 which is the latest release of Syncsort.

Thanks,
Arun

karthikr44

HI,

The following sort card will give u solution. And i assumed that maximum length of file name including '**TSTIN**' is 50. U can change as ur requirement.

karthikr44

Hi,

Change

SORT FIELDS=(1,20,CH,A)

to

SORT FIELDS=(1,50,CH,A)

Regards
R KARTHIK

arcvns · Posted: Tue Sep 16, 2008 1:33 pm Post subject:

karthikr44,

Did you have a look at the output records posted in the original requirement. I believe you need to modify he card to something like this.
I cant test this now as I dont have the recent version of SYNCSORT

karthikr44

HI Arun,

Thanks for spotting the mistake. Ur code provides the following output....

arcvns · Posted: Tue Sep 16, 2008 3:13 pm Post subject:

Sambhaji · Posted: Tue Sep 16, 2008 3:35 pm Post subject:

Yes read correctly..
thats why i wrote "IF..." otherwise i wouldnt have

dick scherrer · Posted: Tue Sep 16, 2008 8:12 pm Post subject:

Hello,

There is no reason (even with qualification) to post a DFSORT code suggestion/reply to a Syncsort topic.

It has been removed. . .

vjai6977 · New User Joined: 08 Aug 2008 Posts: 16 Location: Chennai

Hi,

Thanks for the reply posts.

I tried to execute the solutions proposed by Karthik and sambhaji, either of resulted me on the following error message.

When Used for SORT:
SYSIN :
INCLUDE COND=(3,5,CH,EQ,C'TESTIN')
INREC PARSE=(%01=(ENDBEFR=C'(',FIXLEN=80)),
*
BUILD=(%01)
SORT FIELDS=(1,80,CH,A)
SUM FIELDS=NONE
WER901I **WARNING** SYNCSORT 1.2.3.1 WILL EXPIRE IN 77 DAYS
WER268A INREC STATEMENT : SYNTAX ERROR
WER211B SYNCSMF CALLED BY SYNCSORT; RC=0000
WER449I SYNCSORT GLOBAL DSM SUBSYSTEM ACTIVE

When Used for ICETOOL:

CTL1CNTL :
SORT FIELDS=COPY
INREC PARSE=(%00=(ABSPOS=10,FIXLEN=25,ENDBEFR=C'('),
*
%01=(FIXLEN=2,ENDBEFR=C')')),
BUILD=(1,9,%00,35:C'(',%01,JFY=(SHIFT=RIGHT),C')')
WER428I CALLER-PROVIDED IDENTIFIER IS "0001"
WER901I **WARNING** SYNCSORT 1.2.3.1 WILL EXPIRE IN 77 DAYS
WER268A INREC STATEMENT : SYNTAX ERROR
WER211B SYNCSMF CALLED BY SYNCSORT; RC=0000
WER449I SYNCSORT GLOBAL DSM SUBSYSTEM ACTIVE

Not too sure about this stuffs, I am using ICETOOL SPLICE for another operation and it works fine for me even now.

Any other suggestions for making the duplicate removal successful.

Regards,
Jai

vjai6977 · New User Joined: 08 Aug 2008 Posts: 16 Location: Chennai

my shop uses SYNCSORT for z/os version 1.2.3 , instead of using a parse stement is there any way to achieve the result.

Regars,
Jai

dick scherrer · Posted: Wed Sep 17, 2008 12:56 am Post subject:

Hello,

When your system renews the license that will expire soon, it might be a good time to upgrade to the current release of Syncsort. . .

vjai6977 · New User Joined: 08 Aug 2008 Posts: 16 Location: Chennai

Hi,

I appreciate and recommend for that, but now I would like to get a solution on the current version issues. Any ideas that i can solve it for now. And latter can use the above mentioned solution if we our shop is installed with release 1.3 of syncsort

Regards,
Jai