Resources starting with "io.k8s" being considered builtin prevents Volume Snapshot models generation

[gregth]

I tried to generate models for Volume Snapshots, as follows:

$ bash ./python-crd-cmd.sh -n io.k8s.storage.snapshot < swagger.json

This produced no models.

I inspected the code and it turns out that resources starting with "io.k8s" are considered built:

https://github.com/kubernetes-client/gen/blob/291f9a6ca5004cc91cfc4a744abf129c639d5682/openapi/preprocess_spec.py#L307-L316

This does not hold for (at least) Volume Snapshot resources which are well-known CRDs.

I tweaked a few lines of the script and here is a list of the Resources that were initially ignored, in my case:

io.k8s.storage.snapshot.v1alpha1.VolumeSnapshot
io.k8s.storage.snapshot.v1alpha1.VolumeSnapshotClass
io.k8s.storage.snapshot.v1alpha1.VolumeSnapshotClassList
io.k8s.storage.snapshot.v1alpha1.VolumeSnapshotContent
io.k8s.storage.snapshot.v1alpha1.VolumeSnapshotContentList
io.k8s.storage.snapshot.v1alpha1.VolumeSnapshotList

[gregth]

Let's discuss if anyone else has bumped onto this before, possible solutions, and I proceed with a PR to fix this, if you'd like to.

[brendandburns]

I don't think anyone else has hit this. Feel free to add a flag to optionally disable that check

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues and PRs.

This bot triages issues and PRs according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue or PR as fresh with /remove-lifecycle stale
• Mark this issue or PR as rotten with /lifecycle rotten
• Close this issue or PR with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[gregth]

Instead of adding an extra argument, we can fix this by modifying this if-else statement: https://github.com/kubernetes-client/gen/blob/34ea9c4d4e0c6aeb1c41858b2b230adac3bacfe3/openapi/preprocess_spec.py#L311-L320

so that it does not ignore io.k8s if the user has explicitly specified a KUBERNETES_CRD_GROUP_PREFIX that contains k8s.io:

def filter_api_group(spec):
    models = {}
    for k, v in spec['definitions'].items():
        if k.startswith(os.environ.get('KUBERNETES_CRD_GROUP_PREFIX')):
            models[k] = v
        elif k.startswith("io.k8s"):
            print("Removing builtin Kubernetes Resource %s" %k)
        else:
            print("Ignoring Custom Resource %s" %k)
    spec['definitions'] = models

@brendandburns what do you think? If we agree I can PR this and close this issue.

[gregth]

This issue was resolved by PR https://github.com/kubernetes-client/gen/pull/223.