|
|
| Author |
Message |
Tom Huntsman
New User
Joined: 17 Oct 2007
Posts: 5
Location: Central Illinois
|
|
|
|
I have 2 copies of one file. File A is yesterday mornings version of the file and File B is current mornings version of the file. Both have LRECL of 224. Key is position 2 for length of 9. I need to keep all records from file B that have a value change in position 71 from an "A" in file A to a "T" in file B. The only time a record should appear in both files is when the value in position 71 has changed to a "T"
File A:
| Code: |
----+----1----+----2----+----3----+----4----+----5----+----6----+----7--
8666666660100 Mickey Mouse 820723060814 A0
8777777770100 Donald Duck 500924060201 A0
8888888880100 HUNTSMAN THOMAS 680731000228 A0
8999999990100 Goofey Dog 620805901015 A9
|
File B :
| Code: |
----+----1----+----2----+----3----+----4----+----5----+----6----+----7--
8666666660100 Minnie Mouse 820723060814 T0
8777777770100 Donald Duck 500924060201 T0
8888888880100 HUNTSMAN THOMAS 680731000228 T0
|
I expect to find 8888888880100 Huntsman Thomas in file. The output should be that the key position 2 thru 9 is in both records and position 71 has changed from an "A" to a "T".
Here is what I have tried that does not give me what I want:
| Code: |
//STEP4 EXEC PGM=ICETOOL
//TOOLMSG DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//*
//IN1 DD DSN=TADM.yesterdays.file,DISP=SHR
//IN2 DD DSN=TADM.todays.file,DISP=SHR
//*
//T1 DD DSN=&&T2,UNIT=SYSDA,SPACE=(CYL,(55,5)),DISP=(MOD,PASS)
//OUT1 DD DSN=TADM.ISIHR.EXPIRE,UNIT=SYSDA,
// DISP=(,CATLG,CATLG),
// SPACE=(CYL,(5,2),RLSE)
//TOOLIN DD *
COPY FROM(IN1) TO(T1) USING(CTL1)
COPY FROM(IN2) TO(T1) USING(CTL2)
SELECT FROM(T1) TO(OUT1) ON(2,9,BI) NODUPS USING(CTL3)
//*SELECT FROM(T1) TO(OUT1) ON(2,9,BI) NODUPS USING(CTL3)
/*
//CTL1CNTL DD *
* IN1 -> T1; T1: ADD '1' ID TO IN1 RECORDS
INREC OVERLAY=(225:C'1')
/*
//CTL2CNTL DD *
* IN2 -> T2; T2: ADD '2' ID TO IN2 RECORDS
INREC OVERLAY=(225:C'2')
/*
//CTL3CNTL DD *
* KEEP ONLY NODUPS RECORDS WITH '2' ID.
* REMOVE ID.
OUTFIL FNAMES=OUT1,INCLUDE=(225,1,CH,EQ,C'2'),
BUILD=(1,224)
//
|
any and all help is appreciated. It has been a couple of years since I have tooled around in DFSORT so it is probably great risk to me that I have things confused. |
|
| Back to top |
|
 |
References
|
|
 |
Tom Huntsman
New User
Joined: 17 Oct 2007
Posts: 5
Location: Central Illinois
|
|
|
|
| I should have said I expect to see two records Donald Duck and Huntsman Thomas in my out put record |
|
| Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 1199
Location: At my desk
|
|
|
|
| Just what did you see in the output file? |
|
| Back to top |
|
|
Skolusu
DFSORT Developer
Joined: 07 Dec 2007
Posts: 357
Location: San Jose
|
|
|
|
Tom Huntsman,
| Quote: |
Key is position 2 for length of 9. I need to keep all records from file B that have a value change in position 71 from an "A" in file A to a "T" in file B. The only time a record should appear in both files is when the value in position 71 has changed to a "T"
I expect to find 8888888880100 Huntsman Thomas in file. The output should be that the key position 2 thru 9 is in both records and position 71 has changed from an "A" to a "T".
|
Why only "8888888880100 Huntsman" Thomas record? Both 8666666660100 MICKEY MOUSE &
8777777770100 DONALD DUCK have a 'T' in file 2 and they have an 'A' in file 1 , so why didn't you pick them?
If your intention is to pick all records from current file which has 'T' and a matching 'A' in yesterday's file, the following DFSORT JCL will give you the desired results.
Concatenate both files to sortin and make sure that the current day file is FIRST in the concatenation list. A brief explanation of the job. We add a +1 in packed decimal format for records initially and we flip that to -1 when the 71st byte is a 'T' and we sort on the key and sum on this additional byte at the end. if there is a matching T record for a record the sum will be zero. We use an Include cond on OUTFIL to inlucde only such records.
| Code: |
//STEP0100 EXEC PGM=ICEMAN
//SYSOUT DD SYSOUT=*
//SORTIN DD DSN=your current day file B,disp=shr
// DSN=Your yesterday day file,disp=shr
//SORTOUT DD DSN=your output file,
// DISP=(NEW,CATLG,DELETE),
// UNIT=SYSDA,
// SPACE=(CYL,(X,Y),RLSE)
//SYSIN DD *
OPTION EQUALS
INREC IFTHEN=(WHEN=INIT,OVERLAY=(225:X'1C')),
IFTHEN=(WHEN=(71,1,CH,EQ,C'T'),OVERLAY=(225:X'1D'))
SORT FIELDS=(2,9,CH,A)
SUM FIELDS=(225,1,PD)
OUTFIL INCLUDE=(225,1,PD,EQ,0),BUILD=(1,224)
/*
|
Hope this helps...
Cheers |
|
| Back to top |
|
|
Tom Huntsman
New User
Joined: 17 Oct 2007
Posts: 5
Location: Central Illinois
|
|
|
|
CICS guy --- with that JCL I posted I get all terminated records even if they did not exist in FILE A.
Skolusu ----- I should see all three records as you mentioned I tried to changed the mouse record from Mickey to Minnie but in fact I did not change the Key of 866666666 so yes in fact I should return all three records. I will try the job stream you provided and let you know how it works.
Thank you both for your time and replies. |
|
| Back to top |
|
|
Frank Yaeger
DFSORT Moderator
Joined: 15 Feb 2005
Posts: 4621
Location: San Jose, CA
|
|
|
|
Just for the record.
| Quote: |
| with that JCL I posted I get all terminated records even if they did not exist in FILE A. |
I don't see how you would get all of the file2 records in OUT1, if that's what you're saying here.
The way you set it up, you are doing a NODUPS on position 2-9. All of the records have matches on 2-9 except for the Goofey Dog record, so that's the only record without a duplicate and the only one SELECT would output with NODUPS. Since that record is from file1, you then remove it with the OUTFIL statement, so you'd end up with no records in OUT1. That's what I got when I used your job and input with DFSORT's ICETOOL. |
|
| Back to top |
|
|
Tom Huntsman
New User
Joined: 17 Oct 2007
Posts: 5
Location: Central Illinois
|
|
|
|
Frank and CICS Guy
I apologize. I have confused my self with having to change data before posting for Confidentiality reasons. I am sure that what Frank states is true the data that I have in my real issue is more convoluted than the four records of examples I came up with and in my haste I had forgotten that I at one time utilized a different file trying to figure out a solution and that is where I got records from both files in my output. That trial was a poor one and will not be shown here. sorry about the misinformation. |
|
| Back to top |
|
|
Tom Huntsman
New User
Joined: 17 Oct 2007
Posts: 5
Location: Central Illinois
|
|
|
|
kolusu
I have ran the files thru your job stream and from all tests I have done it appears to have worked like I needed it to work. Thank you for your assistance. |
|
| Back to top |
|
|
|
|
|
