"""Digital ETL specific utility functions."""

import csv
import gzip
import json

import boto3

from flows import s3
from flows.digital import config as flow_config


def send_success_notification(correlation_id, date_end, date_start, upcs):
    """Send SNS message after ETL completes.

    Args:
        correlation_id (str): pre-chained correlation id.
        date_end (str): YYYY-MM-DD exclusive date end range of ETL.
        date_start (str): YYYY-MM-DD inclusive date start range of ETL.
        upcs (list): upcs in the ETL.

    Returns:
        dict: with MessageId string.
    """
    payload = {
        'action': flow_config.SNS_ACTION_SUCCESS,
        'correlation_id': correlation_id,
        'date_end': date_end,
        'date_start': date_start,
        'source': flow_config.SNS_SOURCE,
        'upcs': upcs}
    sns = boto3.client('sns', region_name=flow_config.SNS_REGION_NAME)
    return sns.publish(
        TopicArn=flow_config.SNS_TOPIC_ARN,
        Message=json.dumps(payload),
        Subject=payload['action'])


def queue_build_cache(correlation_id, date_end, date_start, upcs):
    """Send the build cache job message to SQS.

    Args:
        correlation_id (str): pre-chained correlation id.
        date_end (str): YYYY-MM-DD exclusive date end range of ETL.
        date_start (str): YYYY-MM-DD inclusive date start range of ETL.
        upcs (list): upcs in the ETL.

    Returns:
        dict: contains message metadata returned from AWS.
    """
    payload = {
        'action': flow_config.SQS_BUILD_CACHE_ACTION,
        'correlation_id': correlation_id,
        'date_end': date_end,
        'date_start': date_start,
        'source': flow_config.SQS_BUILD_CACHE_SOURCE,
        'upcs': upcs}
    sqs = boto3.client(
        'sqs', region_name=flow_config.SQS_BUILD_CACHE_REGION_NAME)
    return sqs.send_message(
        QueueUrl=flow_config.SQS_BUILD_CACHE_URL,
        MessageBody=json.dumps(payload))


def read_csv_from_s3(s3_path):
    """Read gzipped csv file from s3.

    Args:
        s3_path (str): path to csv on s3.

    Returns:
        iterable(list): parsed csv entries: [['v1', 'v2'], ['v3', 'v4']].
    """
    raw = s3.get_object(s3_path).get()['Body'].read()
    preprocessed = filter(
        lambda f: len(f) > 0,
        map(
            lambda m: m.strip(),
            gzip.decompress(raw).decode('utf8').split('\n')
        )
    )
    yield from csv.reader(preprocessed)