If record (part that I sort on) "already" exists, remove all these records
(or if "exists" and there is a record with 999 in col8-10)
IN:
AAAAAA 001
BBBBBBB 009
AAAAAA 999
OUT:
BBBBBB 009
//Lasse
Dont keep any, if duplicate
Re: Dont keep any, if duplicate
by BillyBoyo » Wed Dec 14, 2011 9:37 pm
Do you mean
Gets you
?
AAAAAA 001
AAAAAA 002
BBBBBB 789
CCCCCC 346
DDDDDD 999
EEEEEE 100
AAAAAA 002
BBBBBB 789
CCCCCC 346
DDDDDD 999
EEEEEE 100
Gets you
BBBBBB 789
CCCCCC 346
EEEEEE 100
CCCCCC 346
EEEEEE 100
?
- BillyBoyo
- Global moderator
- Posts: 3804
- Joined: Tue Jan 25, 2011 12:02 am
- Has thanked: 22 times
- Been thanked: 265 times
Re: Dont keep any, if duplicate
by Frank Yaeger » Wed Dec 14, 2011 11:14 pm
If you want to keep only non-duplicates, you can do that with the SELECT NODUPS function of DFSORT's ICETOOL.
For complete information on the SELECT operator, see:
http://publibz.boulder.ibm.com/cgi-bin/ ... 0608113434
It isn't clear what you mean by this. You need to explain more clearly with a better example of input records ad expected output records that covers all cases.
Also, give the RECFM and LRECL of the input file.
SELECT FROM(IN) TO(OUT) ON(1,6,CH) NODUPS
For complete information on the SELECT operator, see:
http://publibz.boulder.ibm.com/cgi-bin/ ... 0608113434
(or if "exists" and there is a record with 999 in col8-10)
It isn't clear what you mean by this. You need to explain more clearly with a better example of input records ad expected output records that covers all cases.
Also, give the RECFM and LRECL of the input file.
Frank Yaeger - DFSORT Development Team (IBM) - yaeger@us.ibm.com
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
=> DFSORT/MVS is on the Web at http://www.ibm.com/storage/dfsort
Specialties: JOINKEYS, FINDREP, WHEN=GROUP, ICETOOL, Symbols, Migration
=> DFSORT/MVS is on the Web at http://www.ibm.com/storage/dfsort
-
Frank Yaeger - Global moderator
- Posts: 1079
- Joined: Sat Jun 09, 2007 8:44 pm
- Has thanked: 0 time
- Been thanked: 15 times
Re: Dont keep any, if duplicate
by LasseH » Thu Dec 15, 2011 1:49 pm
Hi
Standard FB80
Sorting on 1,6 (key)
If 8,3 = 999 do not keep any of the records with the same key
IN:
aaaaaa 001
aaaaaa 999
bbbbbb 012
ccccccc 012
dddddd 013
dddddd 014
dddddd 999
OUT:
bbbbbb 012
ccccccc 012
//Lasse
Standard FB80
Sorting on 1,6 (key)
If 8,3 = 999 do not keep any of the records with the same key
IN:
aaaaaa 001
aaaaaa 999
bbbbbb 012
ccccccc 012
dddddd 013
dddddd 014
dddddd 999
OUT:
bbbbbb 012
ccccccc 012
//Lasse
Re: Dont keep any, if duplicate
by BillyBoyo » Thu Dec 15, 2011 2:52 pm
Sorry, I think I took a wrong track from the subject you gave the topic.
So, if a particular key (1, 6) has one record with an indicator value of "999" then the whole group of records should be excluded, everything else should be included? Have I got it now?
So, if a particular key (1, 6) has one record with an indicator value of "999" then the whole group of records should be excluded, everything else should be included? Have I got it now?
- BillyBoyo
- Global moderator
- Posts: 3804
- Joined: Tue Jan 25, 2011 12:02 am
- Has thanked: 22 times
- Been thanked: 265 times
Re: Dont keep any, if duplicate
by LasseH » Thu Dec 15, 2011 4:08 pm
BillyBoyo wrote:Sorry, I think I took a wrong track from the subject you gave the topic.
So, if a particular key (1, 6) has one record with an indicator value of "999" then the whole group of records should be excluded, everything else should be included? Have I got it now?
Yep, that's correct
Re: Dont keep any, if duplicate
by BillyBoyo » Thu Dec 15, 2011 8:38 pm
I'm sure a proper solution will arrive later.
You could exclude all but "999" to one file. Then use JOINKEYS with that file as the second against the original file. Take those that are unmatched and you have your data.
I'm thinking something like the following:
Unfortunately I can't test it, and I'm only learning about DFSORT as well. I have hopes that it would work, but no idea or if optimal. My first JOINKEYS attempt
You could exclude all but "999" to one file. Then use JOINKEYS with that file as the second against the original file. Take those that are unmatched and you have your data.
I'm thinking something like the following:
//DROP999 EXEC PGM=SORT
//SYSOUT DD SYSOUT=*
//INA DD *
AAAAAA 001
AAAAAA 999
BBBBBB 012
CCCCC 012
DDDDDD 013
DDDDDD 014
DDDDDD 999
EEEEEE 010
EEEEEE 011
//INB DD SYSOUT=*
//SYMNAMES DD *
FILE-KEY,1,6,CH
RECORD-DONT-WANT-INDICATOR,8,3,CH
RECORD-DONT-WANT-VALUE,C'999'
//SORTOUT DD SYSOUT=*
//SYSIN DD *
OPTION COPY
JOINKEYS F1=INA,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOINKEYS F2=INB,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOIN UNPAIRED,F1,ONLY
//*
//JNF2CNTL DD *
OPTION COPY
INCLUDE COND=(RECORD-DONT-WANT-INDICATOR,EQ,RECORD-DONT-WANT-VALUE)
//*
//SYSOUT DD SYSOUT=*
//INA DD *
AAAAAA 001
AAAAAA 999
BBBBBB 012
CCCCC 012
DDDDDD 013
DDDDDD 014
DDDDDD 999
EEEEEE 010
EEEEEE 011
//INB DD SYSOUT=*
//SYMNAMES DD *
FILE-KEY,1,6,CH
RECORD-DONT-WANT-INDICATOR,8,3,CH
RECORD-DONT-WANT-VALUE,C'999'
//SORTOUT DD SYSOUT=*
//SYSIN DD *
OPTION COPY
JOINKEYS F1=INA,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOINKEYS F2=INB,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOIN UNPAIRED,F1,ONLY
//*
//JNF2CNTL DD *
OPTION COPY
INCLUDE COND=(RECORD-DONT-WANT-INDICATOR,EQ,RECORD-DONT-WANT-VALUE)
//*
Unfortunately I can't test it, and I'm only learning about DFSORT as well. I have hopes that it would work, but no idea or if optimal. My first JOINKEYS attempt
- BillyBoyo
- Global moderator
- Posts: 3804
- Joined: Tue Jan 25, 2011 12:02 am
- Has thanked: 22 times
- Been thanked: 265 times
Re: Dont keep any, if duplicate
by skolusu » Thu Dec 15, 2011 10:13 pm
BillyBoyo wrote:Unfortunately I can't test it, and I'm only learning about DFSORT as well. I have hopes that it would work, but no idea or if optimal. My first JOINKEYS attempt
Good Job for a first attempt. You got the syntax right, however your JCL had an error. Your INB DDname is pointing to sysout instead of the actual DSN name. JNF2CNTL control cards are starting in pos 1. You don't have to specify OPTION COPY on JNF2CNTL as the SORTED,NOSEQCK will treat it as COPY operation. Here is the updated version of the job
//DROP999 EXEC PGM=SORT
//SYSOUT DD SYSOUT=*
//INA DD DSN=Your input file,DISP=SHR
//INB DD DSN=Same input file again,DISP=SHR
//SYMNAMES DD *
FILE-KEY,1,6,CH
RECORD-DONT-WANT-INDICATOR,8,3,CH
RECORD-DONT-WANT-VALUE,C'999'
//SORTOUT DD SYSOUT=*
//SYSIN DD *
OPTION COPY
JOINKEYS F1=INA,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOINKEYS F2=INB,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOIN UNPAIRED,F1,ONLY
//*
//JNF2CNTL DD *
INCLUDE COND=(RECORD-DONT-WANT-INDICATOR,EQ,
RECORD-DONT-WANT-VALUE)
//*
//SYSOUT DD SYSOUT=*
//INA DD DSN=Your input file,DISP=SHR
//INB DD DSN=Same input file again,DISP=SHR
//SYMNAMES DD *
FILE-KEY,1,6,CH
RECORD-DONT-WANT-INDICATOR,8,3,CH
RECORD-DONT-WANT-VALUE,C'999'
//SORTOUT DD SYSOUT=*
//SYSIN DD *
OPTION COPY
JOINKEYS F1=INA,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOINKEYS F2=INB,FIELDS=(FILE-KEY,A),SORTED,NOSEQCK
JOIN UNPAIRED,F1,ONLY
//*
//JNF2CNTL DD *
INCLUDE COND=(RECORD-DONT-WANT-INDICATOR,EQ,
RECORD-DONT-WANT-VALUE)
//*
Return to DFSORT/ICETOOL/ICEGENER
-
- Related topics
- Replies
- Views
- Last post
-
- eliminating duplicate but values in duplicate record copied
by pradeepgowda » Thu Dec 29, 2011 1:24 am - 1 Replies
- 2297 Views
- Last post by skolusu
Fri Dec 30, 2011 3:32 am
- eliminating duplicate but values in duplicate record copied
-
- Duplicate removal and sum
1, 2by storage » Wed Dec 16, 2009 11:40 pm - 12 Replies
- 5477 Views
- Last post by Frank Yaeger
Tue Dec 22, 2009 9:01 pm
- Duplicate removal and sum
-
- Splice on duplicate records
by mmt_bit » Wed Aug 19, 2009 8:57 am - 1 Replies
- 2802 Views
- Last post by skolusu
Wed Aug 19, 2009 8:53 pm
- Splice on duplicate records
-
- Want to remove duplicate records
by pahi » Mon Aug 09, 2010 5:06 pm - 6 Replies
- 10810 Views
- Last post by dick scherrer
Fri Jan 06, 2012 12:39 am
- Want to remove duplicate records
-
- removing duplicate records.
by jvinoth » Wed Jun 06, 2012 5:39 pm - 6 Replies
- 3022 Views
- Last post by NicC
Thu Jun 07, 2012 12:08 pm
- removing duplicate records.