On Wed, May 22, 2013 at 3:57 PM, Thomas Thrainer <[email protected]>wrote:

Yeah, but 5000 is _really_ long, isn't it?
>
> BTW, here are all files longer than 2000 lines (which are not a lot, IMHO):
>
>
Ok, let's leave it at 4500 then.


> 2028 ./test/py/ganeti.cmdlib_unittest.py
> 2180 ./lib/rapi/client.py
> 2191 ./lib/hypervisor/hv_kvm.py
> 2204 ./lib/opcodes.py
> 2207 ./lib/objects.py
> 2470 ./lib/constants.py
> 2507 ./lib/cmdlib/instance_storage.py
> 2642 ./lib/config.py
> 2689 ./test/py/ganeti.locking_unittest.py
> 2722 ./lib/jqueue.py
> 2753 ./test/py/ganeti.jqueue_unittest.py
> 2786 ./lib/query.py
> 2874 ./lib/cmdlib/cluster.py
> 3446 ./lib/bdev.py
> 3459 ./lib/cmdlib/instance.py
> 4129 ./lib/cli.py
> 4131 ./lib/backend.py
>
> (this actually tempted me to lower the limit to 3000 and just ignore C0302
> in the longest four files, but then we wouldn't have an upper boundary for
> those four any more...)
>
>
Yeah, let's not do that, for now. Some of these files will need
refactoring/change anyway: let's revisit this in a couple of quarters.

2470 ./lib/constants.py
2204 ./lib/opcodes.py
2207 ./lib/objects.py

These three we'd like to generate from haskell anyway.

Others perhaps we can find some good way to refactor.

Thanks,

Guido

Reply via email to