MVSFORUMS.com

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

I have a requirement where in i have to write a program to achive the results.
INPUT
-----

Code:

MGR HRD ABC 1
MGR HRD ABC 2
MGR HRD PQR 001 10
MGR HRD PQR 001 20
MGR HRD PQR 002 5
MGR HRD PQR 002 6
MGR HRD XYZ 002 7
MGR HRD XYZ 003 1
MGR HRD XYZ 003 2

OUTPUT
-------

Code:

MGR HRD ABC 1 2
PQR 001 10 20
002 5 6
XYZ 002 7
003 1 2

PROGRAM
-------

Code:

01 A .
05 A1 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 A2 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 A3 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 A4 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 A5 PIC 9(2).
05 FILLER PIC X(62) VALUE SPACES.
01 B .
05 B1 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 B2 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 B3 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 B4 PIC X(3).
05 FILLER PIC X VALUE SPACE.
05 B5 PIC 9(2).
05 FILLER PIC X(4) VALUE SPACES.
05 B6 PIC 9(2).
05 FILLER PIC X(56) VALUE SPACES.

My assumptions:
--------------
1.Read the file record by record assuming record to be 80 bytes long and Input record structure is A and OUTPUT record structure is B.
2.

Code:

IF A(1:3) = 'MGR'
IF A(5:3) = 'HRD'
IF A(9:3) = 'ABC'
MOVE A(11:3) TO CORRESPONDING B RECORD STRUCTURE
ELSE
IF A(9:3) = 'PQR'
MOVE A(11:3) TO CORRESPONDING B RECORD STRUCTURE
ELSE
MOVE A(11:3) TO CORRESPONDING B RECORD STRUCTURE
END-IF
END-IF
END-IF

Can anybody guide me ahead further as how do i get my results as i am not able to proceed ahead?

SureshKumar · Intermediate Joined: 23 Jan 2003 Posts: 211 Topics: 21

venkats,
This should not be difficult, however, I don't think I will be able to do the cobol code. This depends on how flexible you need the code to be. For the code to be generic you need to think of "contro break" logic. "MGR HRD ABC " MGR would be your first break, HRD second break and so on. Moreover, you are moving values to the column format - what would be the limitation ? if you have 100's of rows for a particulr type how would you fit it - i mean the LRECL. Avoid hardcoding the values unless you are sure it's going to be static. Sorry, could not be much more help.

kolusu · Posted: Mon Jan 09, 2006 3:03 pm Post subject:

Venkats,

I would suggest to declare a 3 dimensional table and load it and work with it to get the desired results.

Hope this helps...

Cheers

Kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu

semigeezer · Posted: Mon Jan 09, 2006 3:48 pm Post subject:

This sounds like a homework problem, and I doubt that the correct answer would involve hardcoding MGR, HDR, ABC, XYZ, etc. You just need to remember the last value in each column and when, scanning left to right, a column changes, you start a new line printing the changed columns (changed and all to the right of the change) with the obvious different printing for the last column.

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Thanks Experts for the views on my query.

My input datset can contain other values in the column for HRD also ,and the output will be accordingly as shown if it exists,please guide me ahead as i am unable to write the code.Kolusu why do we need three dimensional table ,can i have a sample program where in i can understand the steps.Please guide me ahead further to complete the task ?

Code:

MGR HRD ABC 1
MGR HRD ABC 2
MGR HRD PQR 001 10
MGR HRD PQR 001 20
MGR HRD PQR 002 5
MGR HRD PQR 002 6
MGR HRD XYZ 002 7
MGR HRD XYZ 003 1
MGR HRD XYZ 003 2

kolusu · Posted: Tue Jan 10, 2006 5:13 am Post subject:

venkats,

Check this link for table handling

http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/IGY3PG10/1.4?DT=20020923143836

Hope this helps...

Cheers

kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Kolusu,

Basically the input dataset is derived from the unload of a dataset and i have the query whether we can use 3 dimension table to get the same results and the link provided by you is very useful for me.Can you brief the steps to follow for getting the output and can i have a sample program using 3 dimensional table where in a similar kind of requirement is used ?

It would be helpful for me if you guide me the steps to follow to get the results.

kolusu · Posted: Tue Jan 10, 2006 12:16 pm Post subject:

Venkats,

Check this link for an example

http://www.mvsforums.com/helpboards/viewtopic.php?t=3399&highlight=search+table

Hope this helps...

Cheers

Kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Kolusu,

Thanks for a very good example which has given me understanding of storing and searching values in a table.I need a slight modification in the desired output by vlady as

Instead of the result showing summation only ,if i need the values of fields as well as SUM how can it be achived assuming the sum to be placed purposefully in ( ).

Code:

EXAMPLE OF DESIRED RESULT SHOWING MONTH OF FEBRUARY ONLY:

02 005 11282003 11282004 02282004 333.00 24.98 50.73

My desired output:

Code:

EXAMPLE OF DESIRED RESULT SHOWING MONTH OF FEBRUARY ONLY:

02 005 11282003 11282004 02282004 166.50 166.50 (333.00) 12.49 12.49 (24.98) 50.73 (50.73)

kolusu · Posted: Tue Jan 10, 2006 3:18 pm Post subject:

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Hai Kolusu,

Basically my requirement goes this way:

1.The input dataset which is derived from the unload of a dataset is below and the query used is :

Code:

SELECT DIV, HOME, DEPT,NO,STATUS,TEST,
SUM(MARKS)
FROM USERID.EMPLOYEE
WHERE DIV = 'AB' AND NONHOME = '100'
AND HOME IN ('ABCD','EFGH','PQRS','WXYZ')
GROUP BY DIV, HOME, DEPT, NO, STATUS, TEST
ORDER BY DIV, HOME, DEPT, NO, STATUS, TEST

DIV HOME DEPT NO STATUS TEST MARKS
---- ----- ---- ----- ---- -------- ------------------
AB PQRS PAL F 1 10
AB PQRS PAL F 2 20
AB PQRS PAL F 3 30
AB PQRS PAL 001 M 1 30
AB PQRS PAL 001 M 2 40
AB PQRS PAL 002 M 1 99
AB PQRS PAL 002 M 2 25
AB PQRS PAL 002 M 3 12
AB PQRS PAL 003 M 1 11
AB PQRS PAL 003 M 2 30
AB PQRS PAL 011 M 1 96
AB PQRS PAL 011 M 2 65
AB PQRS PAL 012 M 1 22
AB PQRS PAL 012 M 2 44
AB PQRS PAL 015 M 1 36
AB PQRS PAL 015 M 2 20
AB PQRS PAL 016 M 1 20
AB PQRS PAL 016 M 2 15
AB PQRS PAL 020 M 1 90
AB PQRS PAL 020 M 2 40
AB PQRS PAL 021 M 1 30
AB PQRS PAL 021 M 2 5
AB PQRS PAL 022 M 1 80
AB PQRS PAL 022 M 2 88
AB PQRS PAL 070 M 1 70
AB PQRS PAL 080 M 2 35
AB PQRS PAL 085 M 1 70
AB PQRS PAL 085 M 2 50

2.I have to code a program where in the report appears like this which basically groups as per the given order:

Code:

DIV HOME DEPT NO STATUS COL1 COL2 COL3 TOTAL
--- ---- ----- ---- ----- ---- ---- ---- ----
12 PQRS PAL F 10 20 30 60
001 M 30 40 70
002 M 99 25 12 136
003 M 11 30 41
011 M 96 65 161
012 M 22 44 66
015 M 36 20 56
016 M 20 15 35
020 M 90 40 130
021 M 30 5 35
022 M 80 88 168
070 M 70 70
080 M 35 35
085 M 70 50 120
---- ---- ---- ----
TOTAL
664 477 42 1183

Basically the program needs to take care of other HOME values,DEPT values,STATUS values also.

Pre-requisites:

1.The HOME values derived from the unload of the query can contain other values aslo ,this has to be taken care by the program.
2.The DEPT values derived from the unload of the query can contain other values also, this has also to be taken care by the program.
3.The STATUS values derived from the unload of the query can contain other values also, this has also to be taken care by the program.
4.The nos remain the same as derived from the unload of the query.

Please guide me how to go about this requirement.

For example there might be situations where the unloaded dataset might look like that is why i mentioned HOME,DEPT ,STATUS can have other values

Code:

DIV HOME DEPT NO STATUS TEST MARKS
---- ----- ---- ----- ---- -------- ------------------
AB ABCD ABC D 1 10
AB ABCD IJK E 2 20
AB ABCD PAL F 3 30
AB ABCD PAL 001 M 1 30
AB EFGH ABC 001 M 2 40
AB EFGH PAL 002 M 1 99
AB EFGH PAL 002 M 2 25
AB EFGH PAL 002 M 3 12
AB EFGH UVW 003 M 1 11
AB PQRS ABC 003 M 2 30
AB PQRS ABC 011 M 1 96
AB PQRS ABC 011 M 2 65
AB PQRS ABC 012 M 1 22
AB PQRS ABC 012 M 2 44
AB PQRS IJK 015 M 1 36
AB PQRS IJK 015 M 2 20
AB PQRS IJK 016 M 1 20
AB PQRS PAL 016 M 2 15
AB PQRS PAL 020 M 1 90
AB PQRS PAL 020 M 2 40
AB PQRS PAL 021 M 1 30
AB PQRS PAL 021 M 2 5
AB PQRS PAL 022 M 1 80
AB PQRS PAL 022 M 2 88
AB PQRS UVW 070 M 1 70
AB PQRS UYW 080 M 2 35
AB WXYZ PAL 085 M 1 70
AB WXYZ PAL 085 M 2 50

kolusu · Posted: Wed Jan 11, 2006 7:37 am Post subject:

venkats,

I am Sorry , I cannot code the entire pgm for you. I already have provided samples as to how it can be done.

Thanks

Kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Kolusu,

Please guide me the steps to follow as i have gone through the sample programs provided by you but my requirement is not particularly based on table key as i am having multiple values in my where clause as shown in the examples.
Please help me out in just highlighting the major steps to follow as there are many kinds of situations which need to be taken care and i will make out the program from my side after your guideance and i will come out with the coding for the program and i understand that coding is not simple for this kind of requirement.

kolusu · Posted: Wed Jan 11, 2006 12:11 pm Post subject:

venkats · Beginner Joined: 09 Jan 2006 Posts: 9 Topics: 1

Kolusu thanks for your early reply despite of me coming up with many queries,

My assumption:
--------------

Code:

read-para
read input-file
   move the fields in the input-key
end-read

perform until input-eof
if input-key = ws-prev-key
   populate the marks in the array
else
   perform the summation of the table values
   load the table with new key
   move input-key to ws-prev-key
end-if
perform read-para
end-perform

I have few queries in the pseudo-code provided by you.

If i am reading record starting from the first and storing the key in input-key,so the key becomes ABPQRSPAL and we are checking the input-key with ws-prev-key which does not have any value and the condition will fail so,i am not able to get what exactly summation of table values means.

Next load the table with the new key means ,we will be storing the fields formed by input-key to a table which has the key defination and move the input-key to ws-prev-key ,so in this case ws-prev-key will have ABPQRSPAL .

Again we read the next record and move the fields in the input-key ,at present the condition will satisfy and we will populate the value of marks 20 .I am not clear with the concept of ws-prev-key ,how is the previous key brought up,can u please brief me the sequence of processing with some sample input.