Windows – How to use WinMerge Excel plugin

I have spent the day writing a DLL plugin for Winmerge to do this - it isn't the world's greatest thing, but it gets the job done.

Takes the filename, splits it into directory and class name, then calls javap for each class file. It is rather slow unfortunately (javap's fault I'd say).

I have put the DLL here:

https://www.dropbox.com/s/6hywu64zzj248t1/displayclassfiles.dll?dl=1

Absolutely no licence, warranty, etc. No trojans or virii either. I will get around to cleaning up the source and putting it up somewhere (probably with the winmerge project people).

Create a "MergePlugins" directory where your winmerge executable is and put the dll in there. It is automatically activated on *.class files.

The hardest part was working out how to launch javap, windows programming is just so awful.

It would be nice if winmerge compared the files before unpacking them to see if they are different - that would save a bit of time. Maybe this can be accomplished in the plugin somehow but as I said I really just hacked this together as quick as I could.

Oh I should mention you will need javap on your path.

EDIT: Source code is here: https://www.dropbox.com/s/gberx85pdcdlor2/unpackjavaclassfiles%20source.7z?dl=1 Built using visual studio 2005 - Adapted from DisplayBinaryFiles plugin which is available with winmerge source package.

I found this post because I was also struggling with WinMerge. The line filter documentation goes right at RegEx, so did I. Stupid me; none of the expressions I made did any filtering. Turns out you just put the string in common to left and right lines and it works like I want.

For example, I am comparing 2 XML files for certain differences. Some lines contain these strings:

IntID IntRef

They are preceded by whitespace, and open tag string and followed by a UUID, which is the entity that is causing a difference I would like to ignore.

So I just added two line filters with just the strings and no RegEx stuff. It was a bit slow to process a 12M file, but I more than made up for the time in having to ignore 1000+ hits to see the 4 differences.

I think this is what you are after with the XXXXXXXX XXXXXXXXX on the left and right. I am assuming the data is the same on each side.