"""Logic for smart_downloader."""

import datetime
import logging
import subprocess
import time
from typing import Iterable

from config import DEFAULTT_N_DOWNLOAD_THREADS

logger = logging.getLogger(__name__)


def parallel_download_curl_to_s3(
        tasks: Iterable[dict],
        n_threads=DEFAULTT_N_DOWNLOAD_THREADS) -> dict:
    """Download files from URLs to S3 in parallel."""
    if not n_threads:
        n_threads = DEFAULTT_N_DOWNLOAD_THREADS

    bash_commands = generate_bash_commands(
        tasks=tasks,
    )
    n_tasks = len(bash_commands)
    command = ['/usr/bin/xargs', '--delimiter', '\\n',
               f'--max-procs={n_threads}', '-i', 'bash', '-c', '{}',
               ';']
    logger.info(f'Running {n_tasks} commands: {bash_commands}')
    logger.info(f'Wrapped by {command}')
    processing_started_at = datetime.datetime.utcnow()
    started = time.perf_counter()
    completed_process = subprocess.run(
        args=command,
        input=bytearray('\n'.join(bash_commands), 'utf-8')
    )
    finished = time.perf_counter()
    duration = finished - started
    result = dict(
        return_code=completed_process.returncode,
        duration=f'{duration:.2f}',
        n_threads=n_threads,
        n_tasks=n_tasks,
        processing_started_at=processing_started_at.isoformat(),
    )
    return result


def generate_bash_commands(tasks: Iterable[dict]):
    """Generate bash commands for downloading files from URLs to S3."""
    result = []
    for task in tasks:
        destination_url = task['destination_url']
        source_url = task['source_url']
        assert destination_url.startswith('s3://'), \
            f'Destination should be S3 url {destination_url}'
        assert source_url.startswith('https://'), \
            f'Source should be HTTPS url {destination_url}'
        command = f'./curl-to-s3.sh "{source_url}" "{destination_url}"'
        result.append(command)
    return result