[BUG] WARN Failed to check if the directory is empty: open data: no such file or directory

[mcarbonneaux]

Expected Behavior

no warning !
and normaly the massage must be ERR and not WARN!

Actual Behavior

when start convertion : WARN Failed to check if the directory is empty: open data: no such file or directory
the directory is not empty, it's containt only one sub directory.

Steps To Reproduce

  myservice:
    image: myimage/myimage
    restart: always
    ports:
      - 5333/tcp
    volumes:
      - /data/mydir:/opt/mydir

WARN Failed to check if the directory is empty: open data: no such file or directory
WARN Skip file in path /data/mydir

the /data/mydir is directory, and contain only one subdir... and /data is directory... and the sub directory containe multiple file...

Kompose Version

v1.33.0 and higher
it's work fine in v1.32.0

Docker-Compose file

myservice:
    image: myimage/myimage
    restart: always
    ports:
      - 5333/tcp
    volumes:
      - /data/mydir:/opt/mydir

Anything else?

i think the problem is in the function checkIsEmptyDir.
they try to recurse the filePath... they found in first a directory data...
but when call recursively checkIsEmptyDir the use relative name data, not the absolute name /data/mydir/data...
they break on first os.ReadDir... but is normal because they try to check data in place of /data/mydir/data.

kompose/pkg/transformer/kubernetes/k8sutils.go

Lines 1265 to 1284 in 9532cee

	// checkIsEmptyDir checks if filepath is empty
	func checkIsEmptyDir(filePath string) (bool, error) {
	files, err := os.ReadDir(filePath)
	if err != nil {
	return false, err
	}
	if len(files) == 0 {
	return true, err
	}
	for _, file := range files {
	if !file.IsDir() {
	return false, nil
	}
	_, err := checkIsEmptyDir(file.Name())
	if err != nil {
	return false, err
	}
	}
	return true, nil
	}

[k8s-triage-robot]

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

This bot triages un-triaged issues 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 as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

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

/lifecycle stale