Problem:

Isn't there a Mars report that reported on fields which are compared to numeric as a potential migration issue. For example:  

IF CT-CMPRSN-FLG (CT-CP-NBR) > 0  

where  

15 CT-CMPRSN-FLG            PIC X.                          

In this case, the program worked on the mainframe because in EBCDIC collating sequence the compare was false but in ASCII the compare was true so we were falling through the wrong path.

Isn't this part of the the Mars reports or maybe it was a different report.

Resolution:

The comparisons that usually involve literals, use the "Data items used in comparisons" the most likely analysis tool to find these items.