brisbane: groupcompress cleanups - part1

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ian Clatworthy wrote:
> I've spent the day doing a code review of groupcompress,
> well the first 3rd of it at least. Attached are my proposed
> code clean-ups. All tests (that passed before) still pass so
> I'm pretty sure I haven't broken anything. OTOH, there's some
> unanswered questions in XXX comments I've added that make it
> worthy of a quick review by jam or lifeless before landing
> these changes.
> 
> I'll continue reviewing the rest of groupcompress tomorrow.
> After that, I'll work my way up the stack (chk_map, inventory)
> getting the core pieces of brisbane-core acceptable (IMO) for
> landing into bzr.dev.
> 
> A start,
> Ian C.
> 

I'm a bit concerned that some of this is going to conflict with the
recent changes Vincent and I did for implementing a pure-python version
of GroupCompress.

For example, I completely got rid of the "_z_header" parts of the code,
because we aren't using labels for now, and if we bring something back,
they are going to be different.

I guess there are some more bits to clean up, since allowing
"self._content_length" to be None was just a backwards compatibility
thing (which I got rid of with the new disk format).

I'm sorry that part didn't land earlier, since your cleanups do look
nice for that area. I think they landed today, so I'll try to resolve
the two patches.

Thanks for doing some cleanup,
John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAknSIjYACgkQJdeBCYSNAANFNgCfXvLjcpXBvG58hL90ffewyFBT
HlUAn3ehTTkyd+Mo1x/P3Ls38ZcJrxUA
=cpdQ
-----END PGP SIGNATURE-----