ELIMINATING COLUMNS FROM MID-RECORD FROM COMPARISON

[slindy]

I'm a bit of a newbie to using EDPro, so I may be asking an obvious question, but I have yet to figure out how to do something that I thought would be a no-brainer. Here goes (at the risk of bruising my ego) -

I have a simple text file with thousands of 350+ byte fixed length records. Basically, its an interface file to another system. Records are mainly comprised of digits. In parallel testing of one version of our application with a newer version, we expect this output interface to be exactly the same, excepting for a datestamp in columns 44-47: This reflects when the file was generated. Of course, I am not trying to run the job at exactly the same time each day, so the actual time value varies.

This time-value is insignificant and I'd like to ignore those columns between files BUT compare all other columns on the record against its newer counterpart. Since all records are mostly numbers, and the numbers are not predictable - I am not sure how I'd use regex to eliminate these "parts" of the records. I can't figure out if a clever use of before/after columns ignore would work (my initial attempt to overlap the columns didn't - I didn't think that was the intent?)

Anyway, you get the idea by now. Help!

Thanks a ton to whomever can offer a solution. My ego will get over it, I'll learn something, but I sure need an answer fast.

Otherwise, love the software and its flexibility.

Thanks

[psguru]

Before/after column, unfortunately, will not help -- they cannot remove a middle part of a line. What you may be able to do is use a regex (as you already mentioned) -- if your datestamp can be somehow distinguished form the rest of the line.

[Ed]

No real answer but I'm sure you've considered using something like textpad to delete the columns of dates from the files before comparison either by putting it into block mode and selecting the columns or by searching for regular expression
^(.{43})....
and replace with
\1
(Posix reg exp selected)