IBM Mainframe Forum

Hi,
I have a two files and I need to create a new file with records from file1 and file2. I will try to explain the problem and what I want with a example:

File1 (ordered descending by KEY and DATE )


File2 (ordered descending by KEY ) this file has 365 records, one by day of the 2010 year.

I need to generate a new file with 365 records for every KEY of file1 with the records from file1 and inserting new records from file2 who the date don’t exist into file1. The problem is the amount that must be some time zeros and sometime the same as the last processed record from file1.

Example:

AAAAA 2010-01-01 000000 (As 2010-01-01 exist into file2 and not in file1, generate the record from key with date from file2 and amount = 0)
AAAAA 2010-01-02 000000 (same as previous explanation)ç
“ “ “
“ “ “
AAAAA 2010-02-15 000100 (as 2010-02-15 exist into file2 we generate the same record as file1)
AAAAA 2010-02-16 000100 (as 2010-02-16 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
AAAAA 2010-02-17 000100 (same as previous explanation)
“ “ “
“ “ “
AAAAA 2010-07-25 001230 (as 2010-07-25 exist into file2 we generate the same record as file1)
AAAAA 2010-07-26 001230 (as 2010-02-16 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
AAAAA 2010-07-28 001230 (same as previous explanation)
“ “ “
“ “ “
AAAAA 2010-12-11 000075
AAAAA 2010-12-12 000075 (as 2010-12-12 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
AAAAA 2010-12-13 000075 (same as previous explanation)
“ “ “
“ “ “
AAAAA 2010-12-31 000075 (same as previous explanation)
BBBBB 2010-01-01 000000 (THE KEY CHANGE, As 2010-01-01 exist into file2 and not in file1, generate the record from key with date from file2 and amount = 0)
BBBBB 2010-01-02 000000 (same as previous explanation)ç
“ “ “
“ “ “
BBBBB 2010-05-20 000250 (as 2010-05-20 exist into file2 we generate the same record as file1)
BBBBB 2010-05-21 000250 (as 2010-05-21 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
BBBBB 2010-05-22 000250 (same as previous explanation)
“ “ “
“ “ “
BBBBB 2010-07-12 000400 (as 2010-07-12 exist into file2 we generate the same record as file1)
BBBBB 2010-07-13 000400 (as 2010-02-13 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
BBBBB 2010-07-14 000400 (same as previous explanation)
“ “ “
“ “ “
BBBBB 2010-08-11 000000 (as 2010-08-11 exist into file2 we generate the same record as file1)
BBBBB 2010-08-12 000000 (as 2010-08-12 exist into file2 and not in file1, generate the record from key with date from file2 and amount = to the last file1 record processed)
BBBBB 2010-08-13 000000 (same as previous explanation)
“ “ “
“ “ “
BBBBB 2010-12-31 000075 (same as previous explanation)

LRECL? RECFM? What to do if a duplicate should occur in either file (yes- it may be impossible but the impossible has a habit of occurring)

Hi NicC
lrecl=80
Recfm=FB
in this case, no duplicates in the files

Thanks

You have the following mistakes in your posts:

File1 (ordered descending by KEY and DATE )
File2 (ordered descending by KEY )

Your example shows ascending order, not descending order.

BBBBB 2010-12-31 000075 (same as previous explanation)

Since you don't have a BBBBB 2010-12-31 record in file1, the amount in this output record should be 000000 from the BBBBB 2010-08-11 000000 record in file1, not 000075.

At any rate, I believe this DFSORT/ICETOOL job will do what you want:

Thanks (you have right there were 2 mistakes...

I'll try your sort and I will post the result.

Thanks again....

Hi Frank

Thanks by your solution, it work perfectly and will use it in our real proyect.

Dfsort and Icetool are absolutely incredibles tools ......

TVM