The UMF file conversion utility is a utility that you can use to
prepare incoming data files for processing by a pipeline.
The UMF file conversion utility can add required UMF tags to source files
and can also add custom tags that you created for your implementation. It
can also perform simple data transformations, such as formatting dates in
a uniform style or combining separate address fields together into one new
field.
Input files are transformed into UMF records by the UMF file conversion
utility according to script files, which are identified by a .fp extension.
After an input file has been transformed, the results are written to stdout
or to a file that you specify.
Considerations
The UMF file conversion utility operates
under the following assumptions:
- All input files must have fixed width fields. Each field within a record
must have the same number of characters as the same field in every other record.
For example if the Address field has 255 characters, the Address field in
each record must contain 255 characters. If the data contains fewer characters
than the fixed width requirements, you must include spaces after the data.
- The utility cannot accept non-printable characters.
Restriction: On AIX systems, the UMF file conversion utility
cannot process input files that are larger than 2 gigabytes. The error states
that the file cannot be found, and you must split larger files into files
smaller than 2 gigabytes. The smaller files must contain complete UMF records.
You cannot only use the UNIX split command to truncate files to less than
2 gigabytes. You need to look at the file content and ensure that the files
are split correctly.