"""Non ETL specific s3 utility functions."""

import csv

import boto3


def split_url(url):
    """Split an s3 url to it's parts. Supports https:// and s3:// formats.

    Args:
        url (str): s3 object url.

    Returns:
        dict: labeled parts of the url.
    """
    schema, path = url.split('://', 1)

    # discard the s3.amazonaws.com domain if it's an https:// address
    if schema == 'https':
        path = path.split('/', 1)[1]

    segments = path.split('/', 1)
    return {'bucket': segments[0], 'key': segments[1]}


def get_object(url):
    """Get the s3 object from a full url.

    Args:
        url (str): s3 object url.

    Returns:
        boto.S3.Object: S3 key object.
    """
    s3_parts = split_url(url)

    s3 = boto3.resource('s3')
    return s3.Object(s3_parts['bucket'], s3_parts['key'])


def download_csv(s3_url):
    """Download CSV file from S3 bucket.

    Args:
        s3_url (str): S3 url.

    Returns:
        tuple: header and list of rows.
    """
    data = get_object(s3_url).get()['Body'].read()
    lines = data.decode().splitlines()
    _, *rows = csv.reader(lines)
    return rows