[PATCH 1 of 6] fixing-import-cycle: _storeprovider moved from basestore to localstore

[Yuya Nishihara]

On Sun, 03 Apr 2016 17:03:00 -0500, Pulkit Goyal wrote:
> # HG changeset patch
> # User Pulkit Goyal <7895pulkit at gmail.com>
> # Date 1459720213 -19800
> #      Mon Apr 04 03:20:13 2016 +0530
> # Node ID 68ebd8b8579b03fc98d8e119d82985a3e91862b6
> # Parent  ff0d3b6b287f89594bd8d0308fe2810d2a18ea01
> fixing-import-cycle: _storeprovider moved from basestore to localstore

"largefiles: blah blah..."

> diff --git a/hgext/largefiles/basestore.py b/hgext/largefiles/basestore.py
> --- a/hgext/largefiles/basestore.py
> +++ b/hgext/largefiles/basestore.py
> @@ -160,15 +160,6 @@
>          '''
>          raise NotImplementedError('abstract method')
>  
> -import localstore, wirestore
> -
> -_storeprovider = {
> -    'file':  [localstore.localstore],
> -    'http':  [wirestore.wirestore],
> -    'https': [wirestore.wirestore],
> -    'ssh': [wirestore.wirestore],
> -    }

No test should fail at each changeset.

> --- a/hgext/largefiles/localstore.py
> +++ b/hgext/largefiles/localstore.py
> @@ -68,3 +68,13 @@
>                      % (cset, filename, storepath))
>                  return True             # failed
>          return False
> +
> +import wirestore
> +
> +_storeprovider = {
> +    'file':  [localstore],
> +    'http':  [wirestore.wirestore],
> +    'https': [wirestore.wirestore],
> +    'ssh': [wirestore.wirestore],
> +    }

It doesn't make sense that the localstore keeps the dict of all stores.
To me, the localstore looks just an instance of store implementations, which
isn't a frontend to all store implementations.