I am trying to fix the problem where downloading xlsx files from s3 bucket, the set up UI using angular and then then backend using python, the front end is all good but the backend is the issue, downloading csv files is not an issue but when downloading xlsx files is where the issue lies, currently using apigateway i am using apiGatewayServiceProxies: as a custom in the yaml file to download and get the objects (xlsx files) whenever I do this the file returns corrupted and double the size, i tried to make a lambda to download instead of using the apiGatewayServiceProxies, i was wondering if there was a way to keep using this?

I tried to use a lambda instead but i ended up with the issue of having to use pandas other external libraries which I rather not use, is there any other solutions?

I've tried

  apiGatewayServiceProxies:
    - s3:
        path: /s3
        method: get
        action: GetObject
        contextHandling: CONVERT_TO_BINARY

and

apiGateway:
  binaryMediaTypes:
    - 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'

I got the error “child “s3” fails because [“contentHandling” is not allowed]”

Trying and creating a new lambda I did:

s3_client = boto3.client('s3')

def handle(event, context):
    logger.debug(f"Received event: {event}")
    logger.debug(f"Context: {context}")

    bucket = os.environ['S3_BUCKET']
    add_keys_to_log_context(bucket=bucket)

    key = extract_key_from_event(event)
    add_keys_to_log_context(key=key)

    if not key:
        error_message = "Missing 'key' query parameter"
        logger.error(error_message)
        raise SkyPilException(HTTPStatus.BAD_REQUEST, error_message)

    try:
        file_content, content_type = get_file_from_s3(bucket, key)

        return {
            'statusCode': HTTPStatus.OK,
            'headers': {
                'Content-Type': content_type,
                'Content-Disposition': f'attachment; filename={os.path.basename(key)}'
            },
            'body': file_content,  # Send the raw binary data
            'isBase64Encoded': False  # Set this to False since it's raw binary data
        }

    except Exception as e:
        msg, title = extract_error_msg(e)
        logger.error(f"Error: {msg}, Title: {title}")
        return {
            "statusCode": HTTPStatus.INTERNAL_SERVER_ERROR,
            "body": f"Error fetching file: {msg}",
            "headers": {
                'Content-Type': 'text/plain'
            }
        }


def extract_key_from_event(event: dict) -> str:
    key = event.get("queryStringParameters", {}).get('key') or event.get("query", {}).get('key')
    return key


def extract_error_msg(e: Exception) -> tuple:
    msg = str(e)
    try:
        error_msg, error_title = msg.split(":", 1)
    except ValueError:
        error_msg = msg
        error_title = None
    return error_msg.strip(), error_title


def get_file_from_s3(bucket_name: str, key: str) -> tuple:
    logger.debug(f"Fetching file from S3 - Bucket: {bucket_name}, Key: {key}")
    try:
        file_stream = io.BytesIO()
        s3_client.download_fileobj(bucket_name, key, file_stream)
        file_stream.seek(0)
        file_content = file_stream.read()  # Read as binary
        content_type = s3_client.head_object(Bucket=bucket_name, Key=key)['ContentType']
        logger.debug(f"Fetched file content of length {len(file_content)} and content type {content_type}")
        return file_content, content_type
    except s3_client.exceptions.NoSuchKey:
        error_message = f"File not found in S3 bucket: {bucket_name}, key: {key}"
        logger.error(error_message)
        raise SkyPilException(HTTPStatus.NOT_FOUND, error_message)
    except s3_client.exceptions.ClientError as e:
        logger.error(f"ClientError fetching file from S3: {e}")
        raise
    except Exception as e:
        logger.error(f"Error fetching file from S3: {e}")
        raise