If virtualenv is activated:

  $ pip install boto3

Or if we install package globally:

  $ sudo pip install boto3

Two examples for listing products:

Client:

• low-level service access
• generated from service description
• exposes botocore client to the developer
• typically maps 1:1 with the service API
• snake-cased method names (e.g. ListBuckets API => list_buckets method)

import boto3

access_key='anystring'
secret_key='anystring'

host='http://data.cloudferro.com'

s3=boto3.client('s3',aws_access_key_id=access_key,
aws_secret_access_key=secret_key, endpoint_url=host,)

for i in s3.list_objects(Delimiter='/', Bucket="DIAS", Prefix='Sentinel-2/MSI/L1C/2020/01/08/',MaxKeys=30000)['CommonPrefixes']:
        print(i)

Resource:

• higher-level, object-oriented API
• generated from resource description
• uses identifiers and attributes
• has actions (operations on resources)
• exposes subresources and collections

import boto3

access_key='anystring'
secret_key='anystring'

host='http://data.cloudferro.com'


s3=boto3.resource('s3',aws_access_key_id=access_key,
aws_secret_access_key=secret_key, endpoint_url=host,)

bucket=s3.Bucket('DIAS')

for obj in bucket.objects.filter(Prefix='Sentinel-2/MSI/L1C/2016/02/03/S2A_OPER_PRD_MSIL1C_PDMC_20160204T043037_R085_V20160203T203520_20160203T203520.SAFE/'):
     print('{0}:{1}'.format(bucket.name, obj.key))