improved docker distrubtion invalid repository finder and deleter

main.go

package main

import (
	"context"
	"flag"
	"fmt"
	"log"
	"path"

	"github.com/aws/aws-sdk-go/aws"
	"github.com/aws/aws-sdk-go/aws/credentials"
	"github.com/aws/aws-sdk-go/aws/session"
	"github.com/aws/aws-sdk-go/service/s3"
)

func isEmptyRepository(ctx context.Context, c *s3.S3, bucket, prefix string) (bool, error) {
	res, err := c.ListObjectsV2WithContext(ctx, &s3.ListObjectsV2Input{
		Bucket:  aws.String(bucket),
		Prefix:  aws.String(path.Join(prefix, "_manifests")),
		MaxKeys: aws.Int64(1),
	})
	if err != nil {
		return false, fmt.Errorf("list objects: %w", err)
	}
	return len(res.Contents) == 0, nil
}

func removeEmptyRepository(ctx context.Context, c *s3.S3, bucket, prefix string) error {
	empty, err := isEmptyRepository(ctx, c, bucket, prefix)
	if err != nil {
		return fmt.Errorf("is empty repository: %w", err)
	}

	if !empty {
		return nil
	}

	log.Printf("%s is empty and will be removed", prefix)

	var continuationToken *string
	for {
		res, err := c.ListObjectsV2WithContext(ctx, &s3.ListObjectsV2Input{
			Bucket:            aws.String(bucket),
			Prefix:            aws.String(prefix),
			ContinuationToken: continuationToken,
		})
		if err != nil {
			return fmt.Errorf("list objects: %w", err)
		}

		var objects []*s3.ObjectIdentifier
		for _, o := range res.Contents {
			log.Printf("delete %s", aws.StringValue(o.Key))
			objects = append(objects, &s3.ObjectIdentifier{
				Key: o.Key,
			})
		}

		if _, err := c.DeleteObjectsWithContext(ctx, &s3.DeleteObjectsInput{
			Bucket: aws.String(bucket),
			Delete: &s3.Delete{
				Objects: objects,
			},
		}); err != nil {
			return fmt.Errorf("delete objects: %w", err)
		}

		if res.NextContinuationToken == nil {
			break
		}

		continuationToken = res.NextContinuationToken
	}
	return nil
}

func removeEmptyRepositories(ctx context.Context, c *s3.S3, bucket, prefix string) error {
	var continuationToken *string
	for {
		res, err := c.ListObjectsV2WithContext(ctx, &s3.ListObjectsV2Input{
			Bucket:            aws.String(bucket),
			Prefix:            aws.String(prefix),
			ContinuationToken: continuationToken,
			Delimiter:         aws.String("/"),
		})
		if err != nil {
			return fmt.Errorf("list objects: %w", err)
		}

		for _, p := range res.CommonPrefixes {
			if err := removeEmptyRepository(ctx, c, bucket, aws.StringValue(p.Prefix)); err != nil {
				return fmt.Errorf("remove empty repository: %w", err)
			}
		}

		if res.NextContinuationToken == nil {
			break
		}

		continuationToken = res.NextContinuationToken
	}
	return nil
}

func Main(ctx context.Context) error {
	region := flag.String("region", "auto", "s3 region")
	endpoint := flag.String("endpoint", "", "s3 endpoint, for example https://abc.r2.cloudflarestorage.com")
	bucket := flag.String("bucket", "", "s3 bucket")
	accessKeyID := flag.String("access-key-id", "", "s3 access key id")
	accessKeySecret := flag.String("access-key-secret", "", "s3 access key secret")
	flag.Parse()

	if *bucket == "" {
		return fmt.Errorf("bucket is required")
	}

	sess, err := session.NewSession(&aws.Config{
		Region:      region,
		Endpoint:    endpoint,
		Credentials: credentials.NewStaticCredentials(*accessKeyID, *accessKeySecret, ""),
	})
	if err != nil {
		return fmt.Errorf("new session: %w", err)
	}

	const prefix = "docker/registry/v2/repositories/"
	if err := removeEmptyRepositories(ctx, s3.New(sess), *bucket, prefix); err != nil {
		return fmt.Errorf("remove empty repositories: %w", err)
	}
	return nil
}

func main() {
	if err := Main(context.Background()); err != nil {
		log.Fatal(err)
	}
}

❯ go run . -endpoint=https://.r2.cloudflarestorage.com -access-key-id= -access-key-secret= -bucket=distribution-test
2024/08/27 23:46:22 docker/registry/v2/repositories/caddy/ is empty and will be removed
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/025823b4426f1156a81c3af7b9be4bfc1e1766cb5f8454012b8650af08ebb41e/link
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/04f2b318172a1d2862336f8011c9c221bdfd0ac670d86bee334c4aac33cebcc3/link
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1/link
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/8f119043738166ba92d2b49ec5cf2d796e29e8841253cc96dd51d2bffd5c9440/link
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/adf4aeafa6ae032949b25ab359e9ac5598c4ee93cb4dae3c2235593ec0929a3c/link
2024/08/27 23:46:22 delete docker/registry/v2/repositories/caddy/_layers/sha256/c6a83fedfae6ed8a4f5f7cbb6a7b6f1c1ec3d86fea8cb9e5ba2e5e6673fde9f6/link
2024/08/27 23:48:39 docker/registry/v2/repositories/alpine/ is empty and will be removed
2024/08/27 23:48:40 delete docker/registry/v2/repositories/alpine/_layers/sha256/324bc02ae1231fd9255658c128086395d3fa0aedd5a41ab6b034fd649d1a9260/link
2024/08/27 23:48:40 delete docker/registry/v2/repositories/alpine/_layers/sha256/c6a83fedfae6ed8a4f5f7cbb6a7b6f1c1ec3d86fea8cb9e5ba2e5e6673fde9f6/link