( grep -F fileb -v filea ; grep -F filea -v fileb ) | sort | uniq

You can simplify that quite a bit if you only need a one-way comparison. This will only work with text files, although it will work with NCS/locales if you have LANG et al. set correctly.

-Adam

John Lange john@johnlange.ca wrote:

I think I've asked this in the past but try as I might I can't remember what the answer was or figure out how I solved it before.

I have 2 files that are nearly identical and I need to know which lines in the two files do _not_ match.

I can think of some relatively complex loops with grep etc. which would probably work (for every line in file A, grep in file B etc.) solve the problem but I'm hoping for something more elegant.

I thought at first that perhaps "join" would do it but even though it will provide left and right joins, there is no way to prevent it from outputting the matching lines.

Any ideas?

John _______________________________________________ Roundtable mailing list Roundtable@muug.mb.ca http://www.muug.mb.ca/mailman/listinfo/roundtable