Currently, I'm working on coding Python services that work with AWS S3 buckets. However, for unit testing parts of these services, I don't want to create a separate AWS S3 bucket just for my unit tests.
It would require managing AWS credentials and it would take a bunch of time to set up for multiple team members. What if there is a way to Mock AWS S3 service on local that does not require any steps from the above?
That's how I found Moto library for Python.
- Run the example below on Replit.
import os
from moto import mock_s3
import boto3
import pprint
BUCKET_NAME = "test_bucket"
os.environ['AWS_ACCESS_KEY_ID'] = 'testing'
os.environ['AWS_SECRET_ACCESS_KEY'] = 'testing'
os.environ['AWS_SECURITY_TOKEN'] = 'testing'
os.environ['AWS_SESSION_TOKEN'] = 'testing'
with mock_s3():
client = boto3.client('s3', region_name='us-east-1')
client.create_bucket(Bucket=BUCKET_NAME)
def write_value_to_s3(key, value):
s3 = boto3.client('s3')
key = key.split("/")
bucket_name = key[0]
key_in_the_bucket = "/".join(key[1:])
s3.put_object(Bucket=bucket_name, Key=key_in_the_bucket, Body=value)
path_1 = f"{BUCKET_NAME}/a/d"
path_2 = f"{BUCKET_NAME}/b/c"
write_value_to_s3(path_1, "")
write_value_to_s3(path_2, "")
client = boto3.client("s3")
object_page = client.list_objects_v2(
Bucket=BUCKET_NAME,
Prefix="a",
Delimiter="/"
)
pprint.pprint(object_page)
object_page = client.list_objects_v2(
Bucket=BUCKET_NAME,
Prefix="",
Delimiter="/"
)
pprint.pprint(object_page)
path_3 = f"{BUCKET_NAME}/a"
write_value_to_s3(path_3, "")
object_page = client.list_objects_v2(
Bucket=BUCKET_NAME,
Prefix="",
Delimiter="/"
)
pprint.pprint(object_page)
path_4 = f"{BUCKET_NAME}/a/b"
write_value_to_s3(path_4, "")
object_page = client.list_objects_v2(
Bucket=BUCKET_NAME,
Prefix="a/",
Delimiter="/"
)
pprint.pprint(object_page)
path_5 = f"{BUCKET_NAME}/a/b/c"
write_value_to_s3(path_5, "")
object_page = client.list_objects_v2(
Bucket=BUCKET_NAME,
Prefix="a/b",
Delimiter="/"
)
pprint.pprint(object_page)
And here you go 💪
Top comments (1)
I have been using moto for mocking a lot of AWS services and it works pretty well. Thanks for sharing this!