IBM Mainframe Forum

by **Swapnilkumar** » Mon Aug 27, 2012 11:19 am

There were 42 different headers having a common footer text "COND CODE" and LRECL=110. We had to split this concatenated data into different files as per report titles(Headers), so that we can process it. say

Select all

0HEADER CONFIRMATION
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE9
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00001
WORK OF DATE: 10 30 2008
COND CODE: ABCD
0HEDER LETTERS
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00002
WORK OF DATE: 10 30 2008
COND CODE: EFGH
0HEADR NEXT LEVEL
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00003
WORK OF DATE: 10 30 2008
COND CODE: IKJK
0HEADER CONFIRMATION
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE9
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00004
WORK OF DATE: 10 30 2008
COND CODE: LMN
0HEADR NEXT LEVEL
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00005
WORK OF DATE: 10 30 2008
COND CODE: OPQR

In this data we needed to split the data as per the header along with footer line and data lines into the respective header file details as

File 1 should contain for header "HEADER CONFIRMATION"

Select all

0HEADER CONFIRMATION
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE9
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00001
WORK OF DATE: 10 30 2008
COND CODE: ABCD

0HEADER CONFIRMATION
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE9
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00004
WORK OF DATE: 10 30 2008
COND CODE: LMN

File 2 should contain for header "HEDER LETTERS"

Select all

0HEDER LETTERS
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00002
WORK OF DATE: 10 30 2008
COND CODE: EFGH

File 3 should contain for header "HEADR NEXT LEVEL"

Select all

0HEADR NEXT LEVEL
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00003
WORK OF DATE: 10 30 2008
COND CODE: IKJK
0HEADR NEXT LEVEL
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE1
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE2
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE3
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE4
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE5
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE6
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE7
DATA LINE FOR THE DIFFERNT TYPS OF HEADER LINE8
0-------------------------------------------------------------------------------
1 REPORT: DAILY REPORTS PAGE 00005
WORK OF DATE: 10 30 2008
COND CODE: OPQR

How we approached for this in single step as mentioend below.

Select all

//STEP01 EXEC PGM=ICETOOL
//SYSOUT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//TOOLMSG DD SYSOUT=*
//SYSUDUMP DD SYSOUT=*
//SORTIN DD DSN=Input data,DISP=SHR
//SORTO01 DD DSN=output file for first header, <<HEADER CONFIRMATION
// DISP=(,CATLG,DELETE),
// UNIT=SYSDA,
// DCB=*.SORTIN,
// SPACE=(CYL,(50,20),RLSE)
//*
//SORTO02 DD DSN=output file for second header, <<HEDER LETTERS
// DISP=(,CATLG,DELETE),
// UNIT=SYSDA,
// DCB=*.SORTIN,
// SPACE=(CYL,(50,20),RLSE)
//*
//SORTO03 DD DSN=output file for third header, <<HEADR NEXT LEVEL
// DISP=(,CATLG,DELETE),
// UNIT=SYSDA,
// DCB=*.SORTIN,
// SPACE=(CYL,(50,20),RLSE)
.
.
.
//*
//TOOLIN DD *
COPY FROM(SORTIN) TO(SORTO01) USING(CP01)
COPY FROM(SORTIN) TO(SORTO02) USING(CP02)
COPY FROM(SORTIN) TO(SORTO03) USING(CP03)
.
.
.
COPY FROM(SORTIN) TO(SORTO03) USING(CP42)
/*
//*
//CP01CNTL DD *
OPTION COPY
INREC IFTHEN=(WHEN=GROUP,BEGIN=(2,19,CH,EQ,C'HEADER CONFIRMATION'),
END=(23,10,CH,EQ,C'COND CODE:'),PUSH=(111:ID=1))
OUTFIL INCLUDE=(111,1,CH,NE,C' '),BUILD=(1,110)
/*
//CP02CNTL DD *
OPTION COPY
INREC IFTHEN=(WHEN=GROUP,BEGIN=(2,13,CH,EQ,C'HEDER LETTERS'),
END=(23,10,CH,EQ,C'COND CODE:'),PUSH=(111:ID=1))
OUTFIL INCLUDE=(111,1,CH,NE,C' '),BUILD=(1,110)
/*
//CP03CNTL DD *
OPTION COPY
INREC IFTHEN=(WHEN=GROUP,BEGIN=(2,16,CH,EQ,C'HEADR NEXT LEVEL'),
END=(23,10,CH,EQ,C'COND CODE:'),PUSH=(111:ID=1))
OUTFIL INCLUDE=(111,1,CH,NE,C' '),BUILD=(1,110)
/*
.
.
.

This information is for sharing, please let us know if you have any different method to carry this out

by **enrico-sorichetti** » Mon Aug 27, 2012 11:44 am

Kolusu will certainly give better explanation,
I might be wrong , but why 42 passes ( reads ) ove the input data.

try something along the lines of

Select all

OPTION COPY
INREC IFTHEN=(WHEN=GROUP,BEGIN=(2,6,CH,EQ,C'HEADER'),
END=(23,10,CH,EQ,C'COND CODE:'),PUSH=(111:ID=2)) <== works for 99 groups
OUTFIL FNAMES=OUT01,INCLUDE=(111,2,CH,EQ,C'01') BUILD=(1,110)
OUTFIL FNAMES=OUT02,INCLUDE=(111,2,CH,EQ,C'02') BUILD=(1,110)
...
OUTFIL FNAMES=OUT42,INCLUDE=(111,2,CH,EQ,C'42') BUILD=(1,110)

by **Swapnilkumar** » Mon Aug 27, 2012 11:52 am

Enrico,
its a good suggestion from your end; thank you for the reply. What you suggested certainly work for the first 6 characters matching with "HEADER" string

Please have a look at the headers again, those are unique and hence the line

Select all

WHEN=GROUP,BEGIN=(2,6,CH,EQ,C'HEADER'),

can not be a generic one for all the 42 different headers

by **enrico-sorichetti** » Mon Aug 27, 2012 12:05 pm

did You try it ?
from Your description nothing prevent using just HEADER as group separator

tested and working

by **enrico-sorichetti** » Mon Aug 27, 2012 2:20 pm

Just curious... for what reason the different spelling for HEADER, .... .....
anyway You can achieve the same result just with the END clause

Select all

INREC IFTHEN=(WHEN=GROUP,
END=(23,10,CH,EQ,C'COND CODE:'),PUSH=(111:ID=2)) <== works for 99 groups

the rest is the same as my previous suggestion

by **Swapnilkumar** » Mon Aug 27, 2012 2:37 pm

Enrico,

Actually headers are unique, the mentioned headers viz,
1) HEADER CONFIRMATION
2) HEDER LETTERS
3) HEADR NEXT LEVEL
had actually different values :!:

The above mentioend solution is executed well and then posted on to publish

by **enrico-sorichetti** » Mon Aug 27, 2012 3:08 pm

The above mentioend solution is executed well and then posted on to publish

looks like there is a language barrier here ...
You posted a slotion for some problem... it worked but
with 42 passes over input data

did You try my last suggestion with the GROUP on the ending <separator>
ONE PASS OF THE DATA

by **skolusu** » Mon Aug 27, 2012 10:01 pm

Swapnilkumar wrote:The above mentioend solution is executed well and then posted on to publish

If you are presenting DFSORT Solutions , then you need to present an optimized solution and not one that does a 42 passes of data. As enrico pointed it can be DONE in a single pass of data, even when you have varying headers. Please try the solution enrico posted. Alternatively you can actually push the header(longest header is 20 bytes) on to every record and then validate that string to your different header value to split on OUTFIL like this

Select all

INREC IFTHEN=(WHEN=GROUP,BEGIN=(2,6,CH,EQ,C'HEADER'),PUSH=(111:2,20))
OUTFIL BUILD=(1,110),INCLUDE=(111,20,CH,EQ,C'HEADER CONFIRMATION')
OUTFIL BUILD=(1,110),INCLUDE=(111,20,CH,EQ,C'HEDER LETTERS')
OUTFIL BUILD=(1,110),INCLUDE=(111,20,CH,EQ,C'HEADER CONFIRMATION')

IBM Mainframe Forum

Data split according to header's and footer's details

Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details

Re: Data split according to header's and footer's details