IBM MAINFRAME HELP & SUPPORT FORUMS

vissubhai · Joined: 07 Nov 2007 Posts: 8 Location: Hyderabad

Hi All...this is Viswanath. Nice to see various complex REXX problems and their solutions here.

Coming to my problem...

I have an input file which is a basic EZT macro for a file record layout. That is my input file is a FB file with lrecl=80. And it contains:

FIELD-NAME START-POSN LENGTH DATA-TYPE
FILE1-RECORD 1 100 A
FILE1-COUNTRY 1 10 A
FILE1-GROUP 11 8 N

etc.

Now reading this file sequentially, I am creating a SORT SYMNAMES(symbolic names)

That is my output file should be like:
FILE1-RECORD,1,100,CH
FILE1-COUNTRY,1,10,CH
FILE1-GROUP,11,8,ZD etc.

I am using the WORD function to extract the field-name, starting-position, length, and the data-type like:

varName = word(OUTLOG.i,1)
starPos = word(OUTLOG.i,2)
varLen = word(OUTLOG.i,3)
varType = word(OUTLOG.i,4)

and then concatenating the above variables with commas like:

outrec = varName || ',' || startPos || ',' || varLen || ',' || varType

The problem is for variable length files, we need to add 4 for the starting-position field becaus of RDW(Record descriptor word).

That is I need to add 4 for the starPos field. As this is a string variable, it is giving me "Bad arithmetic conversion" abend.

Can anyone help me out if I can perform this addition on this string field.

Let me know if you require more information. Awaiting for your response.

Thanks in Advance,
Vissubhai.

expat · Posted: Wed Aug 06, 2008 12:05 pm Post subject:

Why not use the TRANSLATE function to translate all of the nasty blank spaces into loveable little commas ?

Pedro · Joined: 01 Sep 2006 Posts: 536 Location: work

Perhaps you have a data record that does not have numeric value there??

Use TRACE to determine what is wrong with data.

dbzTHEdinosauer · Joined: 20 Oct 2006 Posts: 1639 Location: germany

how are you initializing start pos? '0' or 0?