"""Tasks for the TikTok OpenEscrow Conflict Report Ingestion Workflow."""

from datetime import datetime
import os

from garcon import task

from feed_ingestion.flows.tiktok_openescrow import config
from feed_ingestion.tasks import bootstrap as reload
from feed_ingestion.util import task_status
from feed_ingestion.util.aws import s3


@task.decorate(timeout=1000)
@reload.reset_dynamodb_status_on_reload(config.feed_name)
def bootstrap(activity, date, reload):
    """Bootstrap the workflow with the configuration for the reporting month.

    Args:
        activity (ActivityWorker): The activity worker.
        date (str): First day of the reporting month (YYYY-MM-DD).
        reload (str or None): If 'True', the feed status is reset upstream.

    Returns:
        dict: Initial context for the workflow.
    """
    date_obj = (
        datetime.strptime(date, '%Y-%m-%d')
        if date
        else datetime.today()
    )

    activity.logger.info('Bootstrap flow: {}'.format(date_obj))

    # Source (orchard FTP share) - file is dropped as a .zip.
    drop_path = config.s3['drop']
    source_file_name = config.filename.format(date=date_obj, format='zip')
    s3_drop_file_path = (
        f's3://{config.drop_bucket}/{drop_path}/{source_file_name}'
    )

    # Archive - file is converted to .gz before being uploaded for loading.
    new_file_name = config.filename.format(date=date_obj, format='gz')
    archive_path = config.s3['archive'].format(
        feed_name=config.feed_name, date=date_obj
    )
    s3_dir_path = f's3://{config.data_bucket}/{archive_path}/'
    s3_archive_file_path = f'{s3_dir_path}{new_file_name}'

    return dict(
        date=date_obj.strftime('%Y-%m-%d'),
        feed_name=config.feed_name,
        drop_path=drop_path,
        archive_path=archive_path,
        source_file_name=source_file_name,
        new_file_name=new_file_name,
        s3_drop_file_path=s3_drop_file_path,
        s3_archive_file_path=s3_archive_file_path,
        s3_dir_path=s3_dir_path,
        secrets_path=config.secrets_path,
        staging_raw_table=config.staging_raw_table
    )


@task.decorate(timeout=3600)
def grab_drop_files(activity, date, s3_drop_file_path, s3_archive_file_path):
    """Archive the dropped .zip report to S3 as .gz.

    Status is checked inline (not via the check_status decorator) so a resumed
    run skips the conversion but still returns source_files_dict for the load.

    Args:
        activity (ActivityWorker): The activity worker.
        date (str): Reporting date (YYYY-MM-DD).
        s3_drop_file_path (str): S3 path of the dropped .zip file.
        s3_archive_file_path (str): S3 path for the archived .gz file.

    Returns:
        dict: source_files_dict for the archived file, or a stop signal if the
            file has not arrived yet.
    """
    if task_status.is_completed_task(
            config.feed_name, date, 'grab_drop_files'):
        activity.logger.info(
            f'grab_drop_files already completed for {date} - reusing '
            f'archived file {s3_archive_file_path}'
        )
    else:
        activity.logger.info(
            f'Checking if there are any files in {s3_drop_file_path} to '
            f'process - '
            f'{s3.get_list_of_files_and_directories(s3_drop_file_path)}'
        )
        if not s3.get_list_of_files_and_directories(s3_drop_file_path):
            activity.logger.info(f'No files found in {s3_drop_file_path}')
            return {'stop': True, 'missing_file': s3_drop_file_path}

        activity.logger.info(
            f'Convert ZIP {s3_drop_file_path} to {s3_archive_file_path}'
        )
        s3.convert_zip_to_gzip_on_s3(
            activity,
            zip_s3_path=s3_drop_file_path,
            gz_s3_path=s3_archive_file_path,
            local_temp_dir='./',
            extract_original_filename=True,
        )
        task_status.mark_completed_task(
            config.feed_name, date, 'grab_drop_files')

    # The zip archive holds a single report file, so build a one-entry
    # source_files_dict pointing at the archived gzip file.
    return {
        'source_files_dict': {
            'files': [{
                'file_name': os.path.basename(s3_archive_file_path),
                'file_size': s3.get_key_size(s3_archive_file_path),
                'found': True,
            }],
        },
    }