Are you referring to the statistics given by KaVir? I don't know how he did it, but here is what I use for similar purpose:

$ diff -Barwidy -W 2 –left-column Merc22/src Envy22/src |grep '(' |wc -l
23319

Two things I should clarify: this is a GNU diff, I don't know if others can do the same. And it gives you number of shared _text_ lines, not lines of code.

I was thinking of something a little more complicated, that would ignore indentation and coding style (by applying one before any comparaison is made, (sure I could do it myself)) cause this is an obvious way to fool common automatic tools, and something even smarter that could do some refactoring of its own if it detects perfectly similar code but with different variables names.
not to mention reordering methods (I move code around a lot)
I actually don't know of any of those kind of tools, and was wondering if any existed.

indent

I agree that my solution can be fooled, but dude, it is complicated. Read about every single option I used and you'll see that yourself. I know you haven't yet, because you don't know it already ignores whitespace. And yes, indent can help a little bit in this case (unlike when you want it to produce nice looking code). Other than that, I don't know any tools to do what you want. You could try comparing binaries.

There are tools that look at control flow and anonymize variables, and so forth. Universities use them to detect cheating in programming assignments. I don't know of URLs or if they're available for free, but such things exist. Doing text analysis is certainly not a very good way of doing this. I'd start searching for static analysis, program structure comparison, etc.

lol

Why you so mad tho?

MSVC can help a lot with this actually.

Also, and I know this is not ideal but autodoc+diff can wrk to give similar indications (quick 'n dirty)

there are tools. just google for "clone detection".

It was just a diff - as I explained at the time, "my statistics are heavily biased in favour of Medievia - even a tiny change will register as a fully original line of code".

To be honest that just made it all the more incriminating, particularly when you've got files like this, this and this.

Indeed.

And btw, it was just 'for fun' heh, if you do not have a solution like 'use this tool' it is already too much work I want to pour in. I thought some stuff like that would exist made by some GNU ayatollah to look into GPL viloation :p


As I said already whitespace and indent would be the least of my concern as I have code style I can apply, so it would mofify even brace location…I wanted an 'automated tool'…not some poor man solution (I use CVS plugin for eclipse for that…) The hint about universities checking for stuff like that tells you those should exist…just that the tools maybe not free (as in beer)

this is a clone detector for C/C++ assemblies: http://www.swag.uwaterloo.ca/acd/

the wikipedia page for "Duplicate code" links a few tools, too.

I don't know if it's illegal, but I do know that any mud that accepts donations and gives in-game perks in exchange is a game I'm not going to play, and I take a similar view with "free" games that require some method of payment to actually compete. It presents balancing issues between players when these perks are useful, and makes absolutely no sense where a roleplaying-enforced setting is concerned. Either the game is free and you put forth effort and time to reach the goal, or the game is commercial and you do the same. There's no in-between in my opinion.

Taking donations in exchange for goods is pretty clearly against the letter and spirit of the Diku license, not to mention oxymoronic in the first place.