ytsm/app/YtManagerApp/management/jobs/download_video.py

import os
import re
from string import Template
from threading import Lock

import youtube_dl

from YtManagerApp.models import Video
from YtManagerApp.scheduler import Job, scheduler


class DownloadVideoJob(Job):
    name = "DownloadVideoJob"
    __lock = Lock()

    def __init__(self, job_execution, video: Video, attempt: int = 1):
        super().__init__(job_execution)
        self.__video = video
        self.__attempt = attempt
        self.__log_youtube_dl = self.log.getChild('youtube_dl')

    def get_description(self):
        ret = "Downloading video " + self.__video.name
        if self.__attempt > 1:
            ret += f" (attempt {self.__attempt})"
        return ret

    def run(self):
        # Issue: if multiple videos are downloaded at the same time, a race condition appears in the mkdirs() call that
        # youtube-dl makes, which causes it to fail with the error 'Cannot create folder - file already exists'.
        # For now, allow a single download instance.
        self.__lock.acquire()

        try:
            user = self.__video.subscription.user
            max_attempts = user.preferences['max_download_attempts']

            youtube_dl_params, output_path = self.__build_youtube_dl_params(self.__video)
            with youtube_dl.YoutubeDL(youtube_dl_params) as yt:
                ret = yt.download(["https://www.youtube.com/watch?v=" + self.__video.video_id])

            self.log.info('Download finished with code %d', ret)

            if ret == 0:
                self.__video.downloaded_path = output_path
                self.__video.save()
                self.log.info('Video %d [%s %s] downloaded successfully!', self.__video.id, self.__video.video_id, self.__video.name)

            elif self.__attempt <= max_attempts:
                self.log.warning('Re-enqueueing video (attempt %d/%d)', self.__attempt, max_attempts)
                DownloadVideoJob.schedule(self.__video, self.__attempt + 1)

            else:
                self.log.error('Multiple attempts to download video %d [%s %s] failed!', self.__video.id, self.__video.video_id,
                          self.__video.name)
                self.__video.downloaded_path = ''
                self.__video.save()

        finally:
            self.__lock.release()

    def __build_youtube_dl_params(self, video: Video):

        sub = video.subscription
        user = sub.user

        # resolve path
        download_path = user.preferences['download_path']

        template_dict = self.__build_template_dict(video)
        output_pattern = Template(user.preferences['download_file_pattern']).safe_substitute(template_dict)

        output_path = os.path.join(download_path, output_pattern)
        output_path = os.path.normpath(output_path)

        youtube_dl_params = {
            'logger': self.__log_youtube_dl,
            'format': user.preferences['download_format'],
            'outtmpl': output_path,
            'writethumbnail': True,
            'writedescription': True,
            'writesubtitles': user.preferences['download_subtitles'],
            'writeautomaticsub': user.preferences['download_autogenerated_subtitles'],
            'allsubtitles': user.preferences['download_subtitles_all'],
            'postprocessors': [
                {
                    'key': 'FFmpegMetadata'
                },
            ]
        }

        sub_langs = user.preferences['download_subtitles_langs'].split(',')
        sub_langs = [i.strip() for i in sub_langs]
        if len(sub_langs) > 0:
            youtube_dl_params['subtitleslangs'] = sub_langs

        sub_format = user.preferences['download_subtitles_format']
        if len(sub_format) > 0:
            youtube_dl_params['subtitlesformat'] = sub_format

        return youtube_dl_params, output_path

    def __build_template_dict(self, video: Video):
        return {
            'channel': video.subscription.channel_name,
            'channel_id': video.subscription.channel_id,
            'playlist': video.subscription.name,
            'playlist_id': video.subscription.playlist_id,
            'playlist_index': "{:03d}".format(1 + video.playlist_index),
            'title': video.name,
            'id': video.video_id,
        }

    def __get_valid_path(self, path):
        """
        Normalizes string, converts to lowercase, removes non-alpha characters,
        and converts spaces to hyphens.
        """
        import unicodedata
        value = unicodedata.normalize('NFKD', path).encode('ascii', 'ignore').decode('ascii')
        value = re.sub('[:"*]', '', value).strip()
        value = re.sub('[?<>|]', '#', value)
        return value

    @staticmethod
    def schedule(video: Video, attempt: int = 1):
        """
        Schedules to download video immediately
        :param video:
        :param attempt:
        :return:
        """
        scheduler.add_job(DownloadVideoJob, args=[video, attempt])