The sort programs on the mainframe rely on byte positions and lengths to indicate field locations and a special code to indicate field data types. Connect ETL does this using record layouts.
There are two types of record layouts in Connect ETL: “fixed position fields” and “delimited fields.” Delimited files are extremely rare on the mainframe, so you will want to create a “fixed position fields” record layout for the source file in your task. For a join, you will probably need to create one layout for each of the two source files.
Defining a fixed position layout in Connect ETL is similar to defining fields for use in sorting or comparison with the mainframe sort. The main difference is that Connect ETL only allows you to specify a field length and type. Therefore, you must define all fields, significant or not, before any field you wish to reference in the task. For example, if the first field you wish to reference starts at position 10, you will need to define a 9-byte field before it. Connect ETL knows which fields in a layout you are using and which fields are only present as filler, so there is no performance impact to defining extra fields.
Each Connect ETL field must have a data type. The default is “text,” which can contain any data, including special or non-printing characters. Connect ETL supports several different numeric types, including most of those that are used on the mainframe. Refer to Data Types for a mapping of mainframe data types to Connect ETL data types.
If you are working with variable-length records (DSORG of V, VB, etc.), you will need to adjust your record layout to accommodate the four-byte RDW. Refer to Work with Variable-Length Records for more details on how to make these adjustments and why they are necessary.
If your job step uses the INREC statement, you will need to account for this when defining your record layout. Any fields referenced after an INREC statement, including the sort/merge fields and any fields referenced in INCLUDE, OMIT, OUTREC, and REFORMAT statements refer to positions in the record after it has been altered by the INREC statement. The same is true for a REFORMAT statement that appears after an OUTREC statement. You do not need to create multiple tasks to perform these operations, but you will need to adjust the positions of your fields accordingly. Connect ETL always references the fields as they appear in the source data. Refer to the INREC, OUTREC, and REFORMAT (Reformatting) section below for more information.