unzip archive from S3 on AWS Lambda

aws-lambda-unzipper.py

import os
import sys
import re
import boto3
import zipfile

def parse_s3_uri(url):
    match = re.search('^s3://([^/]+)/(.+)', url)
    if match:
        return match.group(1), match.group(2)

def unzip(event, context):

    source_bucket, source_key = parse_s3_uri(event['source'])
    destination_bucket, destination_key = parse_s3_uri(event['destination'])
    
    temp_zip = '/tmp/file.zip'
    
    s3_client = boto3.client('s3')

    print("downloading {}".format(source_key))
    
    s3_client.download_file(source_bucket, source_key, temp_zip)
    zfile = zipfile.ZipFile(temp_zip)
    
    file_list = [( name, 
                   '/tmp/' + os.path.basename(name),
                   destination_key + os.path.basename(name)) 
                for name in zfile.namelist()]
    
    print("got names {}".format("; ".join([n for n,b,d in file_list])))

    for file_name, local_path, s3_key in file_list:
        data = zfile.read(file_name)
        with open(local_path, 'wb') as f:
            f.write(data)
            del(data) # free up some memory

        s3_client.upload_file(local_path, destination_bucket, s3_key)
        os.remove(local_path)

    return {"files": ['s3://' + destination_bucket + '/' + s for f,l,s in file_list]}
    

invoke-unzipper.py

import boto3
import json

client = boto3.client('lambda', region_name='us-east-1')

# unzip an archive
def lambda_unzip(archive, destination):
    pl = {"source": archive,
          "destination": destination}

    response = client.invoke(FunctionName='unzipper', InvocationType='RequestResponse', Payload=json.dumps(pl))

    print(response['Payload'].read()) 
    

archive = 's3://bucket-name/path/to/archive.zip' 
lambda_unzip(archive, 's3://bucket-name/path/to/destination/')