[MERGE] annotate.py code cleanups

[Aaron Bentley]

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

Ian Clatworthy wrote:
> I'm currently looking at how annotations will work and perform with
> packs. Some minor cleanups to annotate.py to get the ball rolling.

bb:approve

This is okay, but please, please clean up 'full' handling in
annotate_file.  Right now, when 'full' is True, _annotate_file detects
annotations that would be repeatitions and forces them to blank, and
then annotate_file detects blank annotations and substitutes the last
non-blank annotation.  I think _annoate_file should take a 'full'
parameter, and the full/non-full handling should take place there.

> BTW, it looks like workingtree.py has all the code necessary to annotate
> an edited file and show uncommitted changes as "current:".

(and all trees support the same method)

> The annotate
> command however ignores this capability and always annotates a committed
> revision tree - the last committed one by default. Is that a historical
> thing or something we want by design?

Historical accident.  When I added that capability, other people were
actively working on the commandline annotate.  I didn't want to
introduce conflicts, so I left it to them.  I just implemented it for
gannotate.  They never saw fit to implement it for the commandline
annotate, and I don't really care about it.

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFG8UUM0F+nu1YWqI0RAvCaAJ4i3S//jedvB3yVJfsQTjbQVvNy+ACeMw/c
PffwJoyWSmhg7y4u19mCTl0=
=ugMu
-----END PGP SIGNATURE-----