Description:

TIBCO ActiveMatrix Adapter for Files (Unix/Win)(FA) publication can parse the positional schema based on the 'Record length'. When an input file with a varying length for each record is given to FA, the publisher will only parse the line matching the length value and the rest can not be parsed. Input files need to have each record be of same length in order to be successfully parsed by an FA publisher.

Making TIBCO ActiveMatrix Adapter for Files (Unix/Win) publication service parse variable length record data by using a positional read schema.

You can successfully parse the input file with a varying length using TIBCO ActiveMatrix Adapter for Files (Unix/Win) for each record in the following ways.

1).Use a pre-processing script which would pad the spaces to records and make its length a fixed value (as configured in Positional read schema). The preprocessed file will have records with a fixed length and the TIBCO ActiveMatrix Adapter for Files (Unix/Win) (FA) publisher can then publish without causing any parsing error.

2). Design multiple schemas to parse the records to match record length.

Steps to create and use multiple read schemas with an FA publisher if you have a valid complex schema (positional file record) with a parent and child record which successfully parses the preprocessed file records.

1). Create another child schema (positional file record) for parsing the child records containing other data.
2). Use the display wizard feature of the positional file record to get an exact length of the preprocessed file containing other data.
3). Repeat steps 1-2 for each child record, if needed.
4). Delete the existing schema in the Schema Tab of the FA publisher and re-browse the newly created schema.

For details regarding configuring multiple schemas, refer to the attached sample project (Filename:TESTK.zip).