IBM Mainframe Forum

by **chandu_rck** » Mon Apr 21, 2014 11:47 pm

Hi
I have input file of rec lenght 1107.Each group starts with record type 1 and ends with rec type 9(first character) .

I want to extract the first group 1 into a separate file.
Inputfile:

Select all

clientname1
address1
address2
Account1
account2
clientname2
address1
Account 3
 

Select all

 //SYSIN    DD *                                                    
   SORT FIELDS=COPY,                                                
   SKIPREC=570                                                      
   OUTREC IFTHEN=(WHEN=GROUP,END=(1,2,CH,EQ,C'9|'),PUSH(1108:ID=8)) 
   OUTFIL BUILD=(1,1107),INCLUDE=(1108,8,ZD,EQ,1)  
 

with the above sync sort. I am missing the second instance of 9| record.
Actual output

Select all

clientname1
address1
address2
Account1
 

Expected output

Select all

clientname1
address1
address2
Account1
account2<-This record is missing.
 

Appreciate your help!

by **chandu_rck** » Tue Apr 22, 2014 12:04 am

HI adding to the above question.I cannot use the below Syncsort as I don't have fixed BEGIN record type. BEGIN record can be anynumber between 1 to 8.The only condition I have is I can have multiple end records of type 9.

//SYSIN DD *
SORT FIELDS=COPY,
SKIPREC=558
OUTREC IFTHEN=(WHEN=GROUP,BEGIN=(1,2,CH,EQ,C'1|'),PUSH(1108:ID=8))
OUTFIL BUILD=(1,1107),INCLUDE=(1108,8,ZD,EQ,1)
//*

Thanks

by **BillyBoyo** » Tue Apr 22, 2014 1:59 am

What is the SKIPREC for?

by **chandu_rck** » Tue Apr 22, 2014 2:58 am

Sorry please ignore the skiprec.That can be removed

by **BillyBoyo** » Tue Apr 22, 2014 5:21 am

This:

Select all

OPTION COPY

INREC IFTHEN=(WHEN=GROUP,END=(1,2,CH,EQ,C'9|'),PUSH(40:ID=8)),
IFTHEN=(WHEN=GROUP,
BEGIN=(1,2,CH,EQ,C'9|',
AND,40,8,CH,EQ,C'00000001'),
END=(1,2,CH,LT,C'9|'),
PUSH=(50:40,8))

With this:

Select all

Gives this:

Select all

You should be able to amend your INCLUDE= so that it gets 1,2, of '9|' and the second extended field being C'00000001' in addition to your existing condition.

There's quite a lot of repetition. Have you considered using SORT symbols? They have many advantages, one of which is only having to type positions and values once (after that you use a name).

by **Surabhi** » Wed Nov 23, 2016 1:35 pm

Hi Billy Boyo,
I have a similar requirement.
And as you said this seems to be quite repetative.
So i am thinking of using SORT symbols the way you suggested.I tried finding some info on this but no luck.
Could you give me any link which i can refer to for this.

Thank you.

by **NicC** » Wed Nov 23, 2016 4:56 pm

Do not address your post to a specific person.

What have you read in the manual that you do not understand?

by **Surabhi** » Wed Nov 23, 2016 5:15 pm

Hi,
Apologies for addressing it to a specific person.
I dint get any Link or manunal to refer for Specific Sort Symbols to be used in When=Group.
So, can anyone provide me the link which i can go through for this.
Any help appreciated.

Thank you.

by **NicC** » Wed Nov 23, 2016 11:39 pm

How about trying the Syncsort manual as it is a Syncsort feature?

IBM Mainframe Forum

SYNCSORT when=group end condition with multiple end records

SYNCSORT when=group end condition with multiple end records

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec

Re: SYNCSORT when=group end condition with multiple end rec