Statistical distinguisher for compression algorithms ?

["Halvar Flake" <halvar () gmx de>]

Hey all,

I ran into a small problem today: Given a stream of data that I suspect to be compressed with
a well-known lossless compression algorithm, is there a way of deciding which algorithm that
might be ?

It appears to be such a common problem that somebody else (who understands something about
statistics, which I don't) would've run into -- can somebody give me pointers to existing work on
this perhabs ?

Cheers,
Halvar