Download a folder from S3 using Boto3

quick and dirty but it works:

import boto3
import os 

def downloadDirectoryFroms3(bucketName, remoteDirectoryName):
    s3_resource = boto3.resource('s3')
    bucket = s3_resource.Bucket(bucketName) 
    for obj in bucket.objects.filter(Prefix = remoteDirectoryName):
        if not os.path.exists(os.path.dirname(obj.key)):
            os.makedirs(os.path.dirname(obj.key))
        bucket.download_file(obj.key, obj.key) # save to same path

Assuming you want to download the directory foo/bar from s3 then the for-loop will iterate all the files whose path starts with the Prefix=foo/bar.

A slightly less dirty modification of the accepted answer by Konstantinos Katsantonis:

import boto3
s3 = boto3.resource('s3') # assumes credentials & configuration are handled outside python in .aws directory or environment variables

def download_s3_folder(bucket_name, s3_folder, local_dir=None):
    """
    Download the contents of a folder directory
    Args:
        bucket_name: the name of the s3 bucket
        s3_folder: the folder path in the s3 bucket
        local_dir: a relative or absolute directory path in the local file system
    """
    bucket = s3.Bucket(bucket_name)
    for obj in bucket.objects.filter(Prefix=s3_folder):
        target = obj.key if local_dir is None \
            else os.path.join(local_dir, os.path.relpath(obj.key, s3_folder))
        if not os.path.exists(os.path.dirname(target)):
            os.makedirs(os.path.dirname(target))
        if obj.key[-1] == '/':
            continue
        bucket.download_file(obj.key, target)

This downloads nested subdirectories, too. I was able to download a directory with over 3000 files in it. You'll find other solutions at Boto3 to download all files from a S3 Bucket, but I don't know if they're any better.

You could also use cloudpathlib which, for S3, wraps boto3. For your use case, it's pretty simple:

from cloudpathlib import CloudPath

cp = CloudPath("s3://bucket/folder/folder2/")
cp.download_to("local_folder")

Using boto3 you can set aws credentials and download dataset from S3

import boto3
import os 

# set aws credentials 
s3r = boto3.resource('s3', aws_access_key_id='xxxxxxxxxxxxxxxxx',
    aws_secret_access_key='xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx')
bucket = s3r.Bucket('bucket_name')

# downloading folder 
prefix = 'dirname'
for object in bucket.objects.filter(Prefix = 'dirname'):
    if object.key == prefix:
        os.makedirs(os.path.dirname(object.key), exist_ok=True)
        continue;
    bucket.download_file(object.key, object.key)

If you cannot find ur access_key and secret_access_key, refer to this page
I hope it will helps.
thank you.

Another approach building on the answer from @bjc that leverages the built in Path library and parses the s3 uri for you:

import boto3
from pathlib import Path
from urllib.parse import urlparse

def download_s3_folder(s3_uri, local_dir=None):
    """
    Download the contents of a folder directory
    Args:
        s3_uri: the s3 uri to the top level of the files you wish to download
        local_dir: a relative or absolute directory path in the local file system
    """
    s3 = boto3.resource("s3")
    bucket = s3.Bucket(urlparse(s3_uri).hostname)
    s3_path = urlparse(s3_uri).path.lstrip('/')
    if local_dir is not None:
        local_dir = Path(local_dir)
    for obj in bucket.objects.filter(Prefix=s3_path):
        target = obj.key if local_dir is None else local_dir / Path(obj.key).relative_to(s3_path)
        target.parent.mkdir(parents=True, exist_ok=True)
        if obj.key[-1] == '/':
            continue
        bucket.download_file(obj.key, str(target))

Download a folder from S3 using Boto3

Related

Recent Posts