I have a CFF stage to read a 400 bytes fixed block COBOL file.
The records have some end filler bytes. Example: 230 data bytes and 170 filler bytes, according to the book/file definitions.
When I set the fields on the Records tab, if I define 170 bytes for the filler, the second record "goes into" the first one. But if I define just 1 byte, the reading works correctly.
Anyone knows why it happens?
Other question: in certainly position of the record, a column has 8 character bytes - PIC X(8), but some records seems to have an non display data.
On TSO, the fields is like this ("hex on"):
Code: Select all
The cobol guy said that it's because of non initialized variables or initialized with LOW-VALUES when generating the file.
I could read these records when I set the field as DECIMAL-15 on CFF (layout PIC 9(15) COMP-3), but it will not work with the records that has characters.
Is there some way to handle this cases on DataStage?