Reading non-IOSYS files
Home Up Turning off Warning Messages in Walker Batch Processing Setting CICS terminal to mixed case Template Report Performance Reading non-IOSYS files SelfServe Object Replicator


Tips and Tools : Personalizers : Reportbuilder : Reading non-IOSYS Files

Reading non-IOSYS Files in Walker Reportbuilder

Walker's Reportbuilder includes a handy, but little-known feature that allows you
to use a non-IOSYS file as the primary database for your report.  This article tells how.

Why read non-IOSYS Files?

It is fairly easy to use a dataset that does not reside in your database as the "primary database" for a Walker report.  There are several reasons why you might want to do this:

to write a quick report on a non-Walker dataset.

to combine data from a non-Walker dataset and data from the Walker databases in one report.

to combine data from different Walker regions in one report.

How do you do it?

Choose one of the formats supported by the Reportbuilder Non-IOSYS File I/O program, MRG307S.
  File Type   Max data bytes   Dataset RECFM     Dataset LRECL     DDNAME  
TAPEF080 80 FB 80 TAPF080
TAPEV200 200 VB 204 TAPV200
DISKV452 452 VB 456 DSKV452

(There are 5 additional file types, RGAREB0 through RGAREB4, which are reserved for use with Responsibility Reporting.)

(Alternatively, you could enhance the Walker MRG307S program and TGS FID to include your own formats.) 

Choose a logical filename for your input file.  Set up an FID entry for this logical filename.  Use the File Type from the table above for the format you have chosen.  Use a key displacement and key length of zero, since these are non-keyed files.

Set up DDM entries for your logical filename.  Note that bytes 1-2 will always contain the binary record length, just as with any other Walker file.  (This is even true when reading file type TAPEF080.)  Your first data byte will follow the record length, in record displacement 3.

Use your logical filename on the REPORT statement, and use the field names from the DDM as with any other file.

If necessary, convert your input dataset to your chosen format.  Here is some sample JCL for converting a fixed dataset into the DISKV452 format:

//* -------------------------------------------------------------------
//* -------------------------------------------------------------------
//SYSUT2   DD  DSN=xxx.V452,
//             DISP=(NEW,CATLG,DELETE),
//             DCB=(RECFM=VB,LRECL=456),
//             UNIT=SYSDA,
//             SPACE=(CYL,(10,10),RLSE)

Add a DD statement to the JCL that runs your report.  Use the DDNAME from the table above.  Here's an example for DISKV452:




The file type names are slightly different from the DD names.  Be careful.

The Walker names include a number that refers to the maximum number of data bytes.  With variable length records, the Operating System reserves an additional 4 bytes at the beginning of each record for the record length;  the LRECL is 4 bytes longer than the maximum number of actual data bytes.  Therefore, for variable formats, the LRECL must be 4 bytes more than the number specified in the Walker format name.

The Operating System's 4-byte zoned record length is not returned to the report.  The Reportbuilder always returns the maximum record length in the RECORD-LENGTH field, even when records in the input file are shorter than the maximum.  You should understand the nature of your input data and be careful not to use data past the end of the record. 

You will receive an error message if you try to use a file with an incorrect RECFM or LRECL.

The input files can come from any source (disk, tape, etc.) regardless of the format name.

The input records do not have to be sorted in any specific order.

You cannot use non-IOSYS files in generic or direct reads.  They are only be used with Reportbuilder as the primary database.

Contributed by JC Cunningham