On 10/24/11 15:10, Danek Duvall wrote:
Shawn Walker wrote:
https://cr.opensolaris.org/action/browse/pkg/dduvall/copytree
Do you plan to putback to 2010.2H branch as well as trunk?
No. I plan to push to trunk, then open a branch "s11sru" and cherry-pick
it over there.
Ok.
src/modules/misc.py:
line 39: we generally put 'from ' imports separate from normal
imports; but that's nitty and i'm not bent over it
I wasn't sure if we had a policy here. It's seemed somewhat haphazard to
me, other than alphabetical order, and non-pkg imports before pkg imports.
We don't, as I said, doesn't matter much to me.
lines 112, 158: a brief docstring would be nice (explaining why we
need this)
Whoops:
Rewrite of shutil.copytree() that can handle special files such as
FIFOs, sockets, and device nodes. It re-creates all symlinks rather
than copying the data behind them, and supports neither the 'symlinks'
nor the 'ignore' keyword arguments of the shutil version.
and
Rewrite of shutil.move() that uses our copy of copytree().
Sure.
line 113: Do you want to use PKG_DIR_MODE here to os.makedirs to
ensure creation isn't affected by umask, etc. ?
Probably using our own copy of makedirs() is best here?
I specifically didn't suggest that because shutil's copytree relies on
makedirs() raising an error if the target already exists whereas our
makedirs assumes that's ok.
-Shawn
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
