agama-project / agama / 19271277503

66%

master: 73%

Build Type

push

github

Committed by web-flow

Commit Message

Adapt network to the new config based HTTP API (backend) (#2862)

## Problem

There is a new config based HTTP API and all the services needs to be
adapted. The aim of this PR is to adapt the network backend service
(replace #2840).

- https://github.com/agama-project/agama/pull/2715

## Solution

In the new architecture all the Agama modules will live in its own
package implementing the actors/service as described in
(https://github.com/agama-project/agama/pull/2715), but as network was
already using a actors model using an enums approach it was decided to
use the current service in the first iteration of changes.

### Changes done

- Added the actions for getting the system, proposal and current network
configuration using the actual actions passing messages system based on
enums.
- All the structs or types that are related to the API (needs to be
serialized) has been moved to agama-utils, it was decided to not depend
on agama-lib anymore. The types will be re-exported to be used by
agama-network when needed.
- As the current merge is concatenating the connections array, the patch
**(updateConfig)** will replace the connections with the given one when
present, but is something we need to revisit and decide how to do it in
a better way.
- Removed the network web service.
- As the proposal, system info and config contains similar collections
and just change the absence or presence of some of them we have
introduced the **NetworkCollection** and
**NetworkConnectionCollections** in order to do the conversion between
them.

### TODO

- Move to the actors/service model using traits instead of enums and
adding the issues and progress monitor to the network service.
- Remove all the actions which are not needed anymore (old network web
service).
- Add unit tests covering the new approach.

Run Details

2695 of 4694 branches covered (57.41%)

Branch coverage included in aggregate %.

108 of 367 new or added lines in 10 files covered. (29.43%)

227 existing lines in 5 files now uncovered.

25358 of 39058 relevant lines covered (64.92%)

151.0 hits per line