How to delete multiple files and specific pattern in S3 boto3

Viewed 6156

Can Python delete specific multiple files in S3?

I want to delete multiple files with specific extensions.

This script removes all files.

These are the various specific files that I want to delete:

XXX.tar.gz
XXX.txt

** Current code: ** (all files deleted)

import boto3

accesskey = "123"
secretkey = "123"
region = "ap-northeast-1"

s3 = boto3.resource ('s3', aws_access_key_id = accesskey, aws_secret_access_key = secretkey, region_name = region)

bucket = s3.Bucket ('test')
files = [os.key for os in bucket.objects.filter (Prefix = "myfolder / test /")]
tar_files = [file to file in files if file.endswith ('tar.gz')]

#print (f'All files: {files} ')
#print (f'CSV files: {csv_files} ')

objects_to_delete = s3.meta.client.list_objects (Bucket = "test", Prefix = "myfolder / test /")

delete_keys = {'Objects': []}
delete_keys ['Objects'] = [{'Key': tar_files} for tar_files in [obj ['Key'] for obj in objects_to_delete.get ('Content', [])]]

s3.meta.client.delete_objects (Bucket = "test", Delete = delete_keys)

If anyone knows, please let me know.

2 Answers

Presuming that you want to delete *.tar.gz and *.txt files from the given bucket and prefix, this would work:

import boto3

s3_resource = boto3.resource('s3')

bucket = s3_resource.Bucket('my-bucket')
objects = bucket.objects.filter(Prefix = 'myfolder/')

objects_to_delete = [{'Key': o.key} for o in objects if o.key.endswith('.tar.gz') or o.key.endswith('.txt')]

if len(objects_to_delete):
    s3_resource.meta.client.delete_objects(Bucket='my-bucket', Delete={'Objects': objects_to_delete})
  1. Iterate over your S3 buckets
  2. For each bucket, iterate over the files
  3. Delete the requested file types
import boto3
s3 = boto3.resource('s3')
   
   for bucket in s3.meta.client.list_buckets()['Buckets']:
       for count, obj in enumerate(s3.Bucket(bucket['Name']).objects.filter()):
           if obj.key.endswith('.tar.gz') or obj.key.endswith('.txt'):
               print("{}: deleting: {} from: {}".format(count, obj.key, bucket['Name']))
               s3.meta.client.delete_object(Bucket=bucket['Name'], Key=obj.key)

I used this method since, for large buckets creating the complete list and then deleting it can take some time. This way, you see the progress.

Related