fixing broken .Z files?

[Eric Smith]

Jules wrote:
> I've got a 40MB compressed tar archive here, but uncompress barfs after
> the first 23MB - it'd be nice if there was a way of skipping over the
> duff bits if possible and reading *something* from the last 17MB!

I'm not sure what algorithm is used for ".Z" files.  For ".gz", the
answer would probably be no, because the dictionary is continuously
updated and a corrupt area would get completely out of sync, with no
obvious means to guess what the intervening dictionary updates should
be.

If ".Z" uses a simpler algorithm with a fixed dictionary (either
completely fixed, or fixed for the duration of one file), recovery
may well be possible.

Eric