DSXchange

Another solution is to could add a merge stage to append a line to your output, just ensure that the order is set correctly. This line would be created using a row generator stage. I was looking for something similar to finish off a vertical pivot being coded in a transformer (7.5.1 Parallel). I us...

First, yes the mainframe COBOL layout PIC S9(3) will understand the 11F. It will get interpreted as a +116 instead of a -116. But I have a DS parallel job where I imported the cobol layout with a PIC 9(4) -- notice there is no 'S' in the definition. And this uploads back as you wanted (hex F1F1F6). ...

In parallel, on a SF stage, I cannot seem to find where I can set the default value for a NULLable column. Only at the record/file level. When I get warnings only, I then get messages that a target field is NULL and the output record is being dropped. Effectively changes my left outer join to an inn...

Have you tried defining the field as PIC 9(3) in a cobol layout and then importing the layout to DS? I'm doing just that (but with PIC 9(13) and PIC 9(4)) and I'm getting full display numerics in the output from DS as well as when it uploads back to the mainframe.

11F is HEX 'F1F1C6'. The numeric value read should still be 116.

Sorry. missed the BINARY spec. So the entire file should be EBCDIC. Don't convert the field (or any COMP / COMP-3 fields) from EBCDIC to binary. Do that ONLY to the text fields or display numeric fields. The datatype for a binary numeric (PIC S9 COMP) field should be INTEGER (length could be 1 to 3)...

What is the result you are getting? What was the result you were expecting? Or does the job not compile/execute?

When you view the file in a text editor, can you read the text portions? Or are they also in EBCDIC?

Thanks. I tried setting all of the fields to NULLable for the tables on the right of the join. Compiled OK, but on execute complained that I hadn't set a default value for all the NULLable fields (one log entry per field). Set deafult value (SeqFile Stage/Output/Format/FieldDefaults/NullFieldValue) ...

I have 3 files. (A, B and C). A is th driver and I do two outer joins in the join stage. The JOIN stage requires that the join columns be identically named. From the JOIN STAGE, STAG tab, PROPERTIES tab,PROPERTIESIn the columns available for target for the key , the help is "Type: Input Column ...

I have a similar situation. I have 3 files. (A, B an C). A is th driver and I do two outer joins in the join stage. The JOIN stage requires that the join columns be identically named. From the JOIN STAGE, STAG tab, PROPERTIES tab,PROPERTIESIn the columns available for target for the key , the help i...

I'm using a workaround : 1) use a utility on the mainframe (DFSORT, SYNCSORT or FILEAID) to UNPACK the packed and binary data to display numerics. 2) edit the file layout to match the data record with the unpacked numbers 3) transmit the datafile (FTP, Universal Command, IND$FILE) WITH translation t...

1. By "Dumping to a text file", i am assuming that you are ftp'ing to a text file. If your PIC X(?) data (VARCHAR) is readable, then you probably sent with translation to ASCII. Unless your "mainframe" is operating in ASCII mode. Since the data in the COMP fields is BINARY, but F...

Create a TEXT file with your COBOL layout, starting with the COBL 01 level : MAINFRAME LAYOUT 01 FD_STKOUT_REC. 05 FD_STKOUT_REG PIC X (02). 05 FD_STKOUT_OTP PIC X (03). 05 FD_STKOUT_VLN PIC X (04). 05 FD_STKOUT_MYR PIC X (04). 05 FD_STKOUT_PYR PIC X (04). 05 FD_STKOUT_PMN PIC X (02). 05 FD_STKOUT_P...

Have you tried using DOS EXEC (.bat) instead of .sh ?

Are you clearing the hash file before each write?