[RFC] make 'copy' load quickly

[Robert Collins]

On Tue, 2006-09-12 at 17:45 -0500, John Arbash Meinel wrote:
> We've known for a while that one of the slowest things to import is
> 'copy'. This is because it depends on 'inspect' which depends on
> 'tokenize'. However, the 'copy' only uses 2 functions out of 'inspect',
> and they are practically trivial.
> 
> I put together this patch as part of my lazy_import stuff. On my
> machine, this is the difference:

I have one concern here, which is how will folk that import bzr first
get clued in when they import tonekize and stuff is missing.

Perhaps have a demandload in this module to load the real tokenize if
the functions from it are used ? Or raise an exception with a full
explanation and a note that they should import copy before bzrlib ?

-Rob
-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060913/181e1b56/attachment.pgp