Re: [Distutils] PEP 376 up-to-date
At 11:01 AM 5/26/2009 +0200, Tarek Ziadé wrote:
Just one point though:
- do we want to scan egg-info directories that are *directly* added in sys.path ?
No. Such a directory does not represent an importable egg on sys.path. If you mean an .egg file or directory that's directly on sys.path, then that's a bit different. Are you planning to support those? In that event you would need another EggInfo subclass anyway, and you could have the directory (or "PathEntry", perhaps?) constructor simply wrap a single such EggInfo in that event. Or, you can handle it the way pkg_resources does, which is to convert target paths into PEP 302 importer objects, and then use the importer's type to determine what sort of handler should be used for it.
On Tue, May 26, 2009 at 3:51 PM, P.J. Eby <pje@telecommunity.com> wrote:
No. Such a directory does not represent an importable egg on sys.path.
ok
If you mean an .egg file or directory that's directly on sys.path, then that's a bit different.
No I was talking about the .egg-info ones.
Are you planning to support those? In that event you would need another EggInfo subclass anyway, and you could have the directory (or "PathEntry", perhaps?) constructor simply wrap a single such EggInfo in that event.
I didn't plan to support other formats because I am in favor of a unique format. I do think having several formats lead to confusion. Tarek -- Tarek Ziadé | http://ziade.org
participants (2)
-
P.J. Eby -
Tarek Ziadé