Udev methods consolidation #410
Conversation
mchf
force-pushed
the
udev_methods_consolidation
branch
from
1ea3476
to
e760f1a
Compare
mchf
force-pushed
the
udev_methods_consolidation
branch
from
e760f1a
to
f636806
Compare
| # | ||
| # @return [string] value corresponding to the key or empty string | ||
| def udev_key_value(rule, key) | ||
| raise ArgumentError if rule.nil? |
There was a problem hiding this comment.
it is good practice to print also which argument have which problem so something like raise ArgumentError, "rule must not be nil" if rule.nil?
mchf
force-pushed
the
udev_methods_consolidation
branch
from
f636806
to
e9b02fa
Compare
|
|
||
| SetModified() if current_rule != new_rule | ||
| Items()[@current]["udev"] = { "net" => {} } if !Items()[@current]["udev"] |
There was a problem hiding this comment.
I would add a log here maybe explaining that we are adding a new udev rule instead of replacing a new one, but this is also one the things i don't like about calling ReplateItemUdev for Items without udev rules.
|
|
|
What i mean is that we could add the default one in case GetItemUdevRule(@current).empty? directly |
|
Well I tried to split udev methods to basic ones and to some with additional complexity. Generating fallback / default rules belongs to those more complex in my POV. And I'm not sure if it should be generated immediately when reading udev rules or when user asks for device renaming. E.g. It requires some more investigation in udev but I wouldn't be surprised if udev "expects" (in modern linux) to each device having own udev rule - in such case we can use first approach. |
|
So now i think is time to remove the [WIP] from it an update bump version |
|
LGTM |
Purpose is to consolidate all udev rule handling methods at one place and design a shining new class for handling that.