So many responses, so many misconceptions: 1) On NUMERIC moves, alignment is along the decimal point (implicit or explicit) and padding is with zeroes not spaces.
2) COBOL does *not* initialize COMP-3 fields to binary zeroes when no VALUE clause is specified. However, the STORAGE LE run-time option may request this. 3) Although COBOL (on IBM mainframes) often uses a ZAP for packed-decimal fields, this MAY depend on various compiler options, most notably the NUMRROC compiler option. (I believe - but haven't tested this - that with NUMPROC(PFD) no ZAP occurs) 4) there is no need to initialize a receiving (numeric) field - only sending fields *** The response with doing an IF NUMERIC class test will certainly avoid the S0C7. However, one might want to "research" how the invalid data got into the packed-field in the first place. (If this is BEFORE the first READ, then that would certainly explain things. If this is after the first read AND the input data is "supposed to be clean" then this probably indicates that the record layout doesn't match the actual input data. "Rashmi Nijaguni Mogali" <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>.. . > Have tried that already...I am getting SOC7 abend > > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
