boto3_s3utils.py

import boto3
import gzip
from StringIO import StringIO


def get_contents_as_string(bucket, key_prefix,
                           aws_access_key_id=None, aws_secret_access_key=None):
    """Gets the contents as string from a gzipped file on s3
    
    Yeah, it's all in memory, so not a good solution for large objects.

    Args:
        bucket: s3 bucket
        key_prefix: key prefix
        aws_access_key_id: optional aws access key id
        aws_secret_access_key: optional aws secret access key
    Returns:
        string of decompressed contents of the s3 file
    """
    if aws_access_key_id is None or aws_secret_access_key is None:
        cnn = boto3.resource('s3')
    else:
        cnn = boto3.resource('s3',
                             aws_access_key_id=aws_access_key_id,
                             aws_secret_access_key=aws_secret_access_key)

    obj = cnn.Object(bucket, key_prefix).get()
    gz_content = obj["Body"].read()

    return gzip.GzipFile(fileobj=StringIO(gz_content)).read()