[MERGE] groupcompress (brisbane)

[Robert Collins]

On Mon, 2009-04-06 at 09:00 +0200, Vincent Ladeuil wrote:
> >>>>> "Ian" == Ian Clatworthy <ian.clatworthy at internode.on.net> writes:
> 
> <snip/>
> 
>     Ian> So the areas needing attention IMO by a reviewer are:
> 
>     Ian> 1. the pyrex implementation of groupcompress  it's
>     Ian>    pretty new IIUIC so it hasn't necessarily had the
>     Ian>    workout the core groupcompress code has had
> 
> Of course you mean *python* version here ?

Well Ian may not have; but in fact I'm happy with unit tests :).

Ian - the evolution was rather roundabout for groupcompress - I wrote
it[line based], Jam and I did a bunch of optimisation, it got migrated
to character based, then we found the xdiff3 code which provided a very
fast core for the find-unchanged-regions logic, and came with a paired
deserialiser that was more pithy than the python one; so the while the
idea was the same we [John] changed the very inner core to this new
matcher/deserialiser, and only most recently has a pure python version
been readded.

That said:
 - the name could be changed even after landing in bzr.dev, lets not
block on that. We could go with ljcompress or something if you wanted,
but as groupcompress isn't part of the UI for brisbane-core I'm not sure
it matters at all.
 - Lets do it.

-Rob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20090406/323a5513/attachment.pgp