|
View previous topic :: View next topic
|
| Author |
Message |
Sikkandhar
New User
Joined: 03 Dec 2007
Posts: 61
Location: Bangalore
|
|
|
|
Hi,
I have a file of length 80. The requirement is to check
1.if the 1st 9 bytes of the file. If there are duplicates then check the next set of 9 bytes (10-18). If that also matches, then it should be written into Output file F1.
2. If there are no duplicates in the 1-9 bytes of the file, then also it should be written into Output file F1.
3. If the bytes 1-9 are same and 10-18 are different, then it needs to be written into Output File F2.
Currently am planning to have 2 Sort Steps.
Step 1
Filter out all duplicate records (on 1-9 bytes) from Input file using ICETOOL - ALLDUPS (Output file F2).
Non Duplicate records pushed into Output file F1
Step 2
Sort the output of Step 1 on bytes 1-9 and 10-18. If there are duplicates, i write into Output file F1 (using MOD)
If they don't match, then write into Output file F3.
Can these 2 steps be clubbed into one single step
Thanks
Sikkandhar |
|
| Back to top |
|
 |
Pandora-Box
Global Moderator
.jpg)
Joined: 07 Sep 2006
Posts: 1592
Location: Andromeda Galaxy
|
|
|
|
Your explanation is pretty vague for me
Can you please post your requirement with sample data ? |
|
| Back to top |
|
|
Sikkandhar
New User
Joined: 03 Dec 2007
Posts: 61
Location: Bangalore
|
|
|
|
The input file looks like
| Code: |
A123456782012121
A123456782012121
A123456782012121
B123456782012121
B123456782012121
B123456782012122
C123456782012121
D123456782012121 |
The 1st 3 records, has same value in 1-18 bytes, so it has to be written into Output file F1 (Requirement 1 in the above post).
Records 4 through 6, has same value for 1st 9 bytes (B12345678), but the next 9 bytes have different value. Hence it has to be written into Output file F2 (Requirement 3 in the above post).
In Last 2 reocrds, there are no duplicate values for the bytes 1 through 9, hence it needs to be written into Output file F1 (Requirement 2 in the above post). |
|
| Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007
Posts: 10900
Location: italy
|
|
|
|
Your description is still confusing ...
looks like You want to process not single records but groups of records
and ... when You post data/jcl use the code tags ,
they preserve the spacing and provide better readability |
|
| Back to top |
|
|
Sikkandhar
New User
Joined: 03 Dec 2007
Posts: 61
Location: Bangalore
|
|
|
|
The input file looks like
| Code: |
Column 1 - 9 Column 10 -18
A12345678 2012121
A12345678 2012121
A12345678 2012121
B12345678 2012121
B12345678 2012121
B12345678 2012122
B12345678 2012122
C12345678 2012121
D12345678 2012121 |
The first 9 bytes (1-9) are Account numbers and the next set (10-19) are dates.
My Output file F1 should not contain duplicate accounts (.i.e. There should be only one record for Account number - date combination). Records 8 and 9 meet this requirement.
If there are multiple records with the same Account number and Same date (records 1 through 3 meet this requirement), then that should also be considered .i.e. Records 1 to 3 should be written into Output file F1
If there are multiple records with same account number but if the dates are different (records 4 through 7 meet this requirement. There are different dates for the same Account number), Then all these four records should be written into a separate Output file F2. |
|
| Back to top |
|
|
Pandora-Box
Global Moderator
Joined: 07 Sep 2006
Posts: 1592
Location: Andromeda Galaxy
|
|
|
|
What is your sort product and level ?
To check please execute the below code and let us know the SYSOUT message
| Code: |
//S1 EXEC PGM=ICEMAN
//SYSOUT DD SYSOUT=*
//SORTIN DD *
RECORD
/*
//SORTOUT DD DUMMY
//SYSIN DD *
OPTION COPY
/* |
|
|
| Back to top |
|
|
Sikkandhar
New User
Joined: 03 Dec 2007
Posts: 61
Location: Bangalore
|
|
|
|
| Code: |
SYNCSORT FOR Z/OS 1.4.0.1R U.S. PATENTS: 4210961, 5117495 (C) 2010 SYNCSORT INC. DATE=2012/352 TIME=03.49.23
z/OS 1.13.0
SYNCSORT LICENSED FOR CPU SERIAL NUMBER 1B593, MODEL 2097 734 LICENSE/PRODUCT EXPIRATION DATE: 02 MAR 2013
SYSIN :
OPTION COPY
WER276B SYSDIAG= 12561134, 21228076, 21228076, 8622750
WER164B 6,884K BYTES OF VIRTUAL STORAGE AVAILABLE, MAX REQUESTED,
WER164B 0 BYTES RESERVE REQUESTED, 332K BYTES USED
WER146B 32K BYTES OF EMERGENCY SPACE ALLOCATED
WER108I SORTIN : RECFM=FB ; LRECL= 80; BLKSIZE= 80
WER073I SORTIN : DSNAME=A494472.TICEMAN.JOB58057.D0000101.?
WER110I SORTOUT : RECFM=FB ; LRECL= 80; BLKSIZE= 80
WER074I SORTOUT : DSNAME=NULLFILE
WER410B 5,856K BYTES OF VIRTUAL STORAGE AVAILABLE ABOVE THE 16MEG LINE,
WER410B 0 BYTES RESERVE REQUESTED, 216K BYTES USED
WER211B SYNCSMF CALLED BY SYNCSORT; RC=0000
WER449I SYNCSORT GLOBAL DSM SUBSYSTEM ACTIVE
WER416B BSAM WAS USED FOR SORTIN
WER416B BSAM WAS USED FOR SORTOUT
WER054I RCD IN 1, OUT 1
WER169I RELEASE 1.4 BATCH 0520 TPF LEVEL 0.1
WER052I END SYNCSORT - TICEMAN,S1,,DIAG=8200,400E,AAA0,00E6,CA5E,6CA3,AAE8,EC64 |
|
|
| Back to top |
|
|
Pandora-Box
Global Moderator
Joined: 07 Sep 2006
Posts: 1592
Location: Andromeda Galaxy
|
|
|
|
WER messages belong to Syncsort
ICE messages belong to DFSORT
Syncsort questions should be posted in JCL forum
This forum belongs to only DFSORT queries |
|
| Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007
Posts: 10900
Location: italy
|
|
|
|
do You realize that records 8 and 9 fall under the same case as records 1 and 3
write to file two when an account <group> contains different dates
write to file one otherwise
learn the KISS principle ==> Keep It Simple Sam |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19243
Location: Inside the Matrix
|
|
|
|
Hello,
| Quote: |
My Output file F1 should not contain duplicate accounts (.i.e. There should be only one record for Account number - date combination). Records 8 and 9 meet this requirement.
If there are multiple records with the same Account number and Same date (records 1 through 3 meet this requirement), then that should also be considered .i.e. Records 1 to 3 should be written into Output file F1 |
Well, i'm confused 
Please post what the output should be for the sample input data posted. |
|
| Back to top |
|
|
Sikkandhar
New User
Joined: 03 Dec 2007
Posts: 61
Location: Bangalore
|
|
|
|
Hi All,
I was able to solve the above issue. Used below control card
Assume the below to be the Input File contents
| Code: |
Column 1 - 9 Column 10 - 16
A12345678 2012121
A12345678 2012121
A12345678 2012121
B12345678 2012121
B12345678 2012121
B12345678 2012122
B12345678 2012123
C12345678 2012121
D12345678 2012121 |
The first SELECT statement segregates Duplicate Account numbers (into FIle 3) and Unique Account numbers into File 2
The Second SELECT statement analyzes the File 3 (which contains the Duplicate Account numbers) on Account number and Date. If there are duplicate records on with same Account number and Date, it is moved to Output File F2. Rest (same Account number but different dates) are moved to Output File F1
Since I have used Disposition as MOD for File F2, i am copying contents of File 2 to File 4
So in the Above case
File F1 would contain
| Code: |
Column 1 - 9 Column 10 - 16
B12345678 2012122
B12345678 2012123
|
File 3
Column 1 - 9 Column 10 - 16
| Code: |
A12345678 2012121
A12345678 2012121
A12345678 2012121
B12345678 2012121
B12345678 2012121
B12345678 2012122
B12345678 2012123 |
File 4
| Code: |
Column 1 - 9 Column 10 - 16
A12345678 2012121
A12345678 2012121
A12345678 2012121
B12345678 2012121
B12345678 2012121
C12345678 2012121
D12345678 2012121 |
Tool In control card
| Code: |
SELECT FROM(IN) TO(OUT3) ON(1,9,CH) ALLDUPS -
DISCARD(OUT2) USING(CTL1)
SELECT FROM(OUT3) TO(OUT2) ON(1,9,CH) ON(10,7,CH) ALLDUPS -
DISCARD(OUT1) USING(CTL2)
COPY FROM(OUT2) TO(OUT4) USING(CTL3)
|
CTL1
| Code: |
OUTFIL FNAMES=OUT2,
BUILD=(1,238)
OUTFIL FNAMES=OUT1,
BUILD=(1,238) |
CTL2
| Code: |
OUTFIL FNAMES=OUT3,
BUILD=(1,238)
OUTFIL FNAMES=OUT2,
BUILD=(1,238) |
CTL3
| Code: |
OUTFIL FNAMES=OUT4,
BUILD=(1,238) |
Job
| Code: |
//IN DD DSN=INPUT.FILE,DISP=SHR
//OUT1 DD DSN=OUTPUT.FILE.F1,
// DISP=(,CATLG,DELETE),
// DCB=(LRECL=400,RECFM=FB,BLKSIZE=0),
// SPACE=(CYL,(90,90),RLSE)
//OUT2 DD DSN=OUTPUT.FILE.F2,
// DISP=MOD,
// DCB=(LRECL=238,RECFM=FB,BLKSIZE=0),
// SPACE=(CYL,(90,90),RLSE)
//OUT3 DD DSN=OUTPUT.FILE.F3,
// DISP=(,CATLG,DELETE),
// DCB=(LRECL=238,RECFM=FB,BLKSIZE=0),
// SPACE=(CYL,(90,90),RLSE)
//OUT4 DD DSN=OUTPUT.FILE.F4,
// DISP=(,CATLG,DELETE),
// DCB=(LRECL=238,RECFM=FB,BLKSIZE=0),
// SPACE=(CYL,(90,90),RLSE) |
|
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
| Quote: |
| Code: |
SELECT FROM(IN) TO(OUT3) ON(1,9,CH) ALLDUPS -
DISCARD(OUT2) USING(CTL1)
SELECT FROM(OUT3) TO(OUT2) ON(1,9,CH) ON(10,7,CH) ALLDUPS -
DISCARD(OUT1) USING(CTL2)
COPY FROM(OUT2) TO(OUT4) USING(CTL3)
|
CTL1
| Code: |
OUTFIL FNAMES=OUT2,
BUILD=(1,238)
OUTFIL FNAMES=OUT1,
BUILD=(1,238) |
CTL2
| Code: |
OUTFIL FNAMES=OUT3,
BUILD=(1,238)
OUTFIL FNAMES=OUT2,
BUILD=(1,238) |
CTL3
| Code: |
OUTFIL FNAMES=OUT4,
BUILD=(1,238) |
|
Why the BUILDs? Are you reducing the length of records? If you are, all records should be 238 after the first SELECT. Three outputs from the first SELECT? In the second SELECT OUT3 as input and output?
Did you try the two-step approach you were considering? How does it do in terms of resources compared to this? |
|
| Back to top |
|
|
|
|
  |
All times are GMT + 6 Hours |
|
