Provide validation tooling for plain bundles

[exdx]

The plain bundle format is relatively straightforward and documented in #149. Even though it's very intuitive, it's still useful to be able to provide users of the format with some sort of static validation tooling to ensure that their bundle is valid and can successfully be applied to the cluster. Based on the plain+v0 bundle spec the tooling would do the following:

• Confirm the bundle source is reachable
• Confirm every file in the /manifests directory (image-based bundles) or user-provided directory (git-based bundles) is a valid YAML manifest that can be applied via a Kubernetes client successfully
• Confirm no extra, non-manifest files are included
• Confirm at least one file exists in the directory provided

[exdx]

0.4-0.5 candidate

[asmacdo]

Confirm every file in the /manifests directory (image-based bundles) or user-provided directory (git-based bundles) is a valid YAML manifest that can be applied via a Kubernetes client successfully

We can easily confirm that each file is valid yaml, but confirming that the manifest can be applied to Kubernetes is more complex. I discussed this with folks and @kevinrizza pointed out that at validation time, some kubernetes manifests may NOT be able to be applied because the relevent CRDs are not yet present on cluster.

For initial implementation, I am going to check "valid yaml or json" but not check "applicable via a Kubernetes client".

If we need to check for valid k8s manifests, lets file a new issue.

[github-actions]

This issue has become stale because it has been open 60 days with no activity. The maintainers of this repo will remove this label during issue triage or it will be removed automatically after an update. Adding the lifecycle/frozen label will cause this issue to ignore lifecycle events.

[github-actions]

This issue has become stale because it has been open 60 days with no activity. The maintainers of this repo will remove this label during issue triage or it will be removed automatically after an update. Adding the lifecycle/frozen label will cause this issue to ignore lifecycle events.