Answer
Not natively as a distinct source-connection type — the practical approach is to first create an Oracle database link from the IFS database to the SQL Server database (using Oracle's standard heterogeneous/gateway database-link mechanism), which lets Oracle query the SQL Server data as if it were a local table or view. From there, the Migration Job can read that linked data the same way it would read any other Oracle-visible source, since the actual cross-database connectivity is handled at the Oracle database-link level rather than by the Migration tool itself.