IBM Mainframe Forum

by **nikesh_rai** » Wed Mar 13, 2013 11:28 pm

Hi Guys,

I want to format certain string into a different format. My Input file will have the strings given below:

S9
S9V9
S99
S99V9
S99V99
S999
S999V9
S999V99
S999V999
S9(X)V9(X)

X= NUMBER OF OCCURANCE OF 9

I think you understand that its a COBOL data format.. and I have to modify these strings into SAS equivalent data format... and the output should be:

S9 - PD1.
S9V9 - PD1.1
S99 - PD2.
S99V9 - PD2.1
S99V99 - PD2.2
S999 - PD3.
S999V9 - PD3.1
S999V99 - PD3.2
S999V999 - PD3.3
S9(X)V9(X) - PDX.X

Please suggest me some ideas.. I have done it with Parse.. but the code became too huge for this.. is there any other way to do this.. like using ICETOOL
I have gone through a link.. which seems me a good one..

http://ibmmainframes.com/about56572.html

but still want yours opinion

Thanks
Nikesh

by **skolusu** » Wed Mar 13, 2013 11:39 pm

nikesh_rai wrote:I think you understand that its a COBOL data format.. and I have to modify these strings into SAS equivalent data format... and the output should be:

Please suggest me some ideas.. I have done it with Parse.. but the code became too huge for this.. is there any other way to do this.. like using ICETOOL
I have gone through a link.. which seems me a good one..

http://ibmmainframes.com/about56572.html

but still want yours opinion

Thanks
Nikesh

Nikesh all your variables defined as COMP-3? What happens if any of fields are Comp, Comp-1, Comp-2 or comp-5?. Don't get hung up using parse as there are other simple ways to do what you want.

by **nikesh_rai** » Thu Mar 14, 2013 12:04 am

Thanks Kolusu..

Yes.. it will have only comp-3.. just for your information.. please don't be confused with data and data declaration

.. I have nothing to do with data at this step.. the data declaration is only the data as input.. and I need to format it into PDX.X format as I have mention above..

My input file looks like

Select all

WXXX-BAA-SAA-NAA S9(4)
WXXX-CAA-IA S9(10)
WXXX-RAAA-SAAAA-VAA-NAA S999
WXXX-GAA-AAA S9(9)V9(4)
WXXX-PAAA-IAAAA S9(4)V999

and output should looks like

Select all

WXXX-BAA-SAA-NAA PD4.
WXXX-CAA-IA PD10.
WXXX-RAAA-SAAAA-VAA-NAA PD3.
WXXX-GAA-AAA PD9.4
WXXX-PAAA-IAAAA PD4.3

by **enrico-sorichetti** » Thu Mar 14, 2013 1:07 am

just curious ... here is a POC on how to do it in REXX

Select all

/* REXX */
Trace "O"

pic.1 = "S9(4)"
pic.2 = "S9(10)"
pic.3 = "S999"
pic.4 = "S9(9)V9(4)"
pic.5 = "S9(4)V999"
pic.0 = 5

do p = 1 to pic.0
format = "PD"
/* isolate integer and decimal part */
parse var pic.p ipic "V" dpic

/* try xx(yy) format */
parse var ipic . "(" digits ")"
if digits = "" then ,
digits = length(ipic) - 1
format = format || digits

if dpic = "" then do
format = format || "."
end
else do
parse var dpic . "(" digits ")"
if digits = "" then ,
digits = length(dpic)
format = format || "." || digits
end
say left(pic.p, 10) "..." format

end

exit

and here is the result

Select all

S9(4) ... PD4.
S9(10) ... PD10.
S999 ... PD3.
S9(9)V9(4) ... PD9.4
S9(4)V999 ... PD4.3

by **enrico-sorichetti** » Thu Mar 14, 2013 1:38 am

hit enter too quickly
change

Select all

parse var ipic . "(" digits ")"

to

Select all

parse var ipic . "(" digits ")" .

and

Select all

parse var dpic . "(" digits ")"

to

Select all

parse var dpic . "(" digits ")" .

the before works, but the after is more compliant with a proper parsing logic

by **skolusu** » Thu Mar 14, 2013 4:32 am

Here is a DFSORT JCL which will give you the desired results. I assumed that you field declarations start at pos 36 and you can have S9999... upto 20 and v9999... upto 20.

Select all

//STEP0100 EXEC PGM=SORT
//SYSOUT DD SYSOUT=*
//SORTIN DD *
----+----1----+----2----+----3----+----4----+----5----+----6----+----7
WXXX-BAA-SAA-NAA S9(4)
WXXX-CAA-IA S9(10)
WXXX-RAAA-SAAAA-VAA-NAA S999
WXXX-GAA-AAA S9(9)V9(4)
WXXX-PAAA-IAAAA S9(4)V999
WXXX-PAAA-IBBBB S9V99999
//SORTOUT DD SYSOUT=*
//SYSIN DD *
OPTION COPY
INREC IFOUTLEN=120,IFTHEN=(WHEN=INIT,
PARSE=(%01=(ABSPOS=36,STARTAFT=C'S',ENDBEFR=C'V',ENDBEFR=C' ',
FIXLEN=20),
%02=(FIXLEN=20)),OVERLAY=(81:%01,%02)),

IFTHEN=(WHEN=(81,20,SS,NE,C'('),
OVERLAY=(081:81,20,TRAN=ALTSEQ,121:C'PD',
081,1,ZD,ADD,082,1,ZD,ADD,083,1,ZD,ADD,084,1,ZD,ADD,
085,1,ZD,ADD,086,1,ZD,ADD,087,1,ZD,ADD,088,1,ZD,ADD,
089,1,ZD,ADD,090,1,ZD,ADD,091,1,ZD,ADD,092,1,ZD,ADD,
093,1,ZD,ADD,094,1,ZD,ADD,095,1,ZD,ADD,096,1,ZD,ADD,
097,1,ZD,ADD,098,1,ZD,ADD,099,1,ZD,ADD,100,1,ZD,
EDIT=(IT),C'.',81:121,5,15X),HIT=NEXT),

IFTHEN=(WHEN=(101,20,SS,NE,C'(',AND,101,20,CH,GT,C' '),
OVERLAY=(101:101,20,TRAN=ALTSEQ,
101:101,1,ZD,ADD,102,1,ZD,ADD,103,1,ZD,ADD,104,1,ZD,ADD,
105,1,ZD,ADD,106,1,ZD,ADD,107,1,ZD,ADD,108,1,ZD,ADD,
109,1,ZD,ADD,110,1,ZD,ADD,111,1,ZD,ADD,112,1,ZD,ADD,
113,1,ZD,ADD,114,1,ZD,ADD,115,1,ZD,ADD,116,1,ZD,ADD,
117,1,ZD,ADD,118,1,ZD,ADD,119,1,ZD,ADD,120,1,ZD,
EDIT=(II),18X))
ALTSEQ CODE=(F9F1,40F0)

OUTREC IFTHEN=(WHEN=INIT,FINDREP=(STARTPOS=081,ENDPOS=100,
INOUT=(C'9(',C'PD',C') ',C'. '))),
IFTHEN=(WHEN=INIT,FINDREP=(STARTPOS=101,
INOUT=(C'9(',C' ',C') ',C' '))),
IFTHEN=(WHEN=INIT,BUILD=(1,35,81,40,SQZ=(SHIFT=LEFT)))

//*

by **nikesh_rai** » Sat Mar 16, 2013 12:08 pm

Thanks a lot Kolusu... Its working..

Thanks Enrico...

IBM Mainframe Forum

String formatting using DFSORT/ICETOOL

String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL

Re: String formatting using DFSORT/ICETOOL