Improve the separation between the Connect and MM2 operators

[scholzj]

Type of change

• Refactoring

Description

The AbstractConnectOperator, KafkaConnectAssemblyOperator and KafkaMirrorMaker2AssemblyOperator classes are currently badly designed in how they separate the functionality specific to Connect and MM2. The AbstractConnectOperator does not really contain only the abstract and shared parts. But it seems to contain also many parts related only to Connect. This applies specifically for the parts related to the KafkaConnector management and operations -> KafkaConnector resources are used only with Connect as MM2 has the connector configuration directly inside the KafkaMirrorMaker2 CR. These parts are then later overridden in the KafkaMirrorMaker2AssemblyOperator. This makes it very hard to see:

• What methods are really shared
• What should be the access modifiers for the different methods

This PR refactors these classes to improve the separation between them. The AbstractConnectOperator now contains only the parts that are really shared and does not contain any parts unique to either Connect or MM2. The other parts were moved to the corresponding assembly operator, in most cases to KafkaConnectAssemblyOperator. This refactoring should help us to better understand what is shared and what is not shared. And in the future it should allow us to follow up with another refactoring that will try to reuse more code between KafkaConnectAssemblyOperator and KafkaMirrorMaker2AssemblyOperator => these classes have now each its own reconciliation flow. But they really differ only in how the connectors are managed. So there should be space to further improve this. But this is not done in this PR to keep the PR scope under review and keep the reviews easier. Where possible, this PR also makes methods private or static.

Note: While this PR moves a lot of methods around, it usually only changes their placement, access modifiers etc. It doesn't change any content inside these methods with the exception of few places where very simple methods were inlined.

This contributes to #8158.

Checklist

• Make sure all tests pass
• Try your changes from Pod inside your Kubernetes and OpenShift cluster, not just locally

[scholzj]

/azp run regression

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).