"""YouTube Asset Report Workflow."""
from datetime import date as date_module
from datetime import datetime
import re

from garcon import task
from garcon_contrib.dynamo_feed_status import garcon_feed_status

from feed_ingestion.flows.youtube_asset import config
from feed_ingestion.util.aws import s3


STOP_RESPONSE = {'stop': True}


@task.decorate(timeout=1000)
def bootstrap(
        activity, date, licensor=None, reload=None,
        skip_grab_reports_files=None):
    """Bootstrap workflow by injecting initial context from config.

    Args:
        activity (ActivityWorker): The Garcon activity worker.
        date (str): Reporting date (YYYY-MM-DD).
        licensor (str): one of config.licensors
        reload (str): If 'True' delete feed status in dynamodb.
        skip_grab_reports_files (str): If 'True' get theorchard files from
            S3 archive location. This is a workaround for API failures when
            the files are available via CMS, but not available via API. We
            must download and place the files into the archive location
            manually.

    Returns:
        dict: Initial context for the workflow.
    """
    if not licensor:
        licensor = 'theorchard'
    assert licensor in config.licensors, f'unsupported licensor "{licensor}"'
    date = date or date_module.today().strftime('%Y-%m-%d')
    parsed_date = datetime.strptime(date, '%Y-%m-%d')

    feed_name = '_'.join([config.feed_name, licensor])
    report_status_name = feed_name

    cms_dict = {
        'theorchard': config.cms_dict
    }

    if reload == 'True':
        activity.logger.info(
            'Delete status for feed: {} {} '.format(report_status_name, date))
        garcon_feed_status.delete_status(report_status_name, date)
    else:
        overall_status = garcon_feed_status.get_overall_status(
            report_status_name, date)
        if overall_status == garcon_feed_status.STATUS_INGESTED:
            return STOP_RESPONSE

    archive_path = config.s3['archive_path'].format(
        date=parsed_date, licensor=licensor)

    credentials_path = config.credentials_paths.get(licensor)

    return {
        'feed_name': feed_name,
        'date': date,
        'licensor': licensor,
        'skip_grab_reports_files': skip_grab_reports_files,
        's3_bucket': config.s3_bucket,
        's3_archive_path': archive_path,
        'credentials_path': credentials_path,
        's3_dir_path': 's3://{}/{}'.format(
            config.s3_bucket, archive_path),
        'cms_dict': cms_dict.get(licensor)
    }


def remove_prefix(text, prefix):
    """Remove prefix from string if string starts with it."""
    return text[text.startswith(prefix) and len(prefix):]


@task.decorate(timeout=1000)
def source_files(activity, licensor, s3_bucket, s3_archive_path):
    """Get list of files to ingest for given date and licensor.

    Args:
        activity (ActivityWorker): The Garcon activity worker.
        licensor (str): one of config.licensors
        s3_bucket (str): s3 bucket containing source files
        s3_archive_path (str): archive path for the date in s3 bucket.

    Returns:
        dict: dict containing 'source_files_dict' with list
        of source files metadata
    """
    s3_path_prefix = s3_archive_path
    s3_full_path = 's3://{}/{}'.format(
        s3_bucket,
        s3_archive_path)

    filenames = s3.get_list_of_files_and_directories(s3_full_path)

    source_files_dict = {'files': []}
    for file_path in filenames:
        file_name = remove_prefix(file_path, s3_path_prefix)

        file_pattern = config.source_file_pattern[licensor]
        if not re.match(file_pattern, file_name):
            activity.logger.warning(
                f'Skipping non-matching source file {file_name}')
            continue

        file_size = s3.get_key_size(
            's3://{}/{}'.format(config.s3_bucket, file_path),
            config.expected_bucket_owner
        )

        source_files_dict['files'].append({
            'file_name': file_name,
            'file_size': file_size})

    if not source_files_dict['files']:
        raise ValueError(f'No source files found in {s3_full_path}')

    return {
        'source_files_dict': source_files_dict,
    }