write is not using the area from the FROM area

xsray · New User Joined: 16 Sep 2008 Posts: 19 Location: illinois

i have 2 input records that get reformated and then written to a flat file.
the write command is:

WRITE ECOE-TRAN-RECORD FROM ECOE-CB006-RECORD.

ECOE-CB006-RECORD contains the correct data from input record 1 but when the WRITE is executed a low value record is written, when the next input record is formated, ECOE-CB006-RECORD contains the info from input record 2, but when the write is exected, the data from input record 1 is written.

what can be causing this?

thanks

Bill Woodger · Posted: Thu Jan 08, 2015 6:05 am

Program error.

For a more detailed explanation, we'd need to see all the relevant code, including knowledge of where data-definitions are located for each name that you mention.

The simplest problem would be that the WRITE is being executed before the data is put into ECOE-CB006-RECORD.

xsray · New User Joined: 16 Sep 2008 Posts: 19 Location: illinois

this shows the buffer before the first write

Robert Sample · Posted: Thu Jan 08, 2015 8:10 am

Bill Woodger · Posted: Thu Jan 08, 2015 5:51 pm

Please use the Code tags for information requiring preservation of spacing.

Still guessing as you have not supplied sufficient code.

It looks like you have variable-length records and are using APPLY WRITE ONLY for the file, or compiler option AWO.

Possibly you have no BLOCK CONTAINS on the FD (so defaults to one record per block) and only one buffer defined for the file, but this is unlikely.

Otherwise, as Robert has indicated, the address under the FD would have been "moved on" to the next available part of the buffer, and you would not see the data you expect.

Assuming the above, are you looking at the correct file when you are checking it? Look at the third file in the spool for the step, and check that there is nothing unexpected for that dataset. Check the dataset name is the one you are looking at.

If, during your initial test, you had the WRITEs out of sync, your file would look like that. If you or the system have failed to delete the original dataset and it is specified as NEW,CATLG, you would not be seeing the latest output from your program.

There can be other reasons, so I'll stop guessing there.

xsray · New User Joined: 16 Sep 2008 Posts: 19 Location: illinois

new update - after going over the program with the programmer, the problem only happens when after the output file is closed in the program it is opened again for I-O for a rewrite to the first record. when the open I-O is commented out, the writes work just fine, recompile with the i-o in and the low-value record appears during the initial write when open for output only.

Bill Woodger · Posted: Thu Jan 08, 2015 11:32 pm

Still none of the code that's been requested? Should be even more of it now.

I have no guesses for closing and re-opening a file. I've never done that.

Are you saying that the mere inclusion of the OPEN for input-output somehow causes the original WRITE not to work? Or is the COBOL code after the second OPEN buggy, and is trashing the file and you'd somehow prefer to blame the original WRITE statement?

How did you confirm that the file was OK before if you do do the second OPEN?

xsray · New User Joined: 16 Sep 2008 Posts: 19 Location: illinois

so the question is, how do i update the first record in file that was just created in the same program?

xsray · New User Joined: 16 Sep 2008 Posts: 19 Location: illinois

Bill, solved the problem. Your thought about the rewrite trashing the file was right. I added a second SELECT and FD for the file in the program. then opened for output to create the file, then close when done. After that i opened the file again for i-o this time using a different SELECT and FD that corresponded to a new DD in the JCL but pointing to the same file just created.

thanks for your help.

new JCL
//*
//ECOETRA2 DD DSN=QCPJNT.ECOETRAN,DISP=OLD
//*
//ECOETRAN DD DSN=QCPJNT.ECOETRAN,DISP=OLD

new select and fd for ECOETRA2 (same file as ECOETRAN)
SELECT ECOE-TRAN-FILE
ASSIGN TO ECOETRAN
FILE STATUS IS ECOETRAN-STATUS
ORGANIZATION IS SEQUENTIAL
ACCESS MODE IS SEQUENTIAL.

SELECT ECOE-TRAN-FILE2
ASSIGN TO ECOETRA2
FILE STATUS IS ECOETRA2-STATUS
ORGANIZATION IS SEQUENTIAL
ACCESS MODE IS SEQUENTIAL.

FD ECOE-TRAN-FILE
RECORDING MODE IS F
BLOCK CONTAINS 0 RECORDS
RECORD CONTAINS 3000 CHARACTERS
LABEL RECORDS ARE STANDARD.

01 ECOE-TRAN-RECORD PIC X(3000).

FD ECOE-TRAN-FILE2
RECORDING MODE IS F
BLOCK CONTAINS 0 RECORDS
RECORD CONTAINS 3000 CHARACTERS
LABEL RECORDS ARE STANDARD.

01 ECOE-TRAN-RECORD2 PIC X(3000).

new rewrite code
--------------------------------------------------------
UPDATE FIRST RECORD OF FILE WITH THE HIST COPY ON OFF SW
--------------------------------------------------------
OPEN I-O ECOE-TRAN-FILE2
READ ECOE-TRAN-FILE2 INTO WS-X3000
MOVE WS-COPY-HIST-ON-OFF-SW TO WSX-HIST-IND
REWRITE ECOE-TRAN-RECORD2 FROM WS-X3000
CLOSE ECOE-TRAN-FILE2

by splitting the file into 2 DDs and 2 FDs - it works!

chandan.inst · Posted: Fri Jan 09, 2015 8:18 am

Hi,

Do you really need a second DD statement here?

After closeing the file opened in OUTPUT mode, we can open the same file using same DD statement in I-O mode as well.

Correct me if I am missing anything here

Regards,
Chandan