wrolpi/modules/videos/models.py

import json
from datetime import timedelta, datetime
from pathlib import Path
from typing import Optional

from sqlalchemy import Column, Integer, String, Boolean, JSON, Date, ARRAY, ForeignKey, Computed, types
from sqlalchemy.orm import relationship, Session
from sqlalchemy.orm.collections import InstrumentedList

from wrolpi.common import Base, tsvector, ModelHelper, PathColumn, get_absolute_media_path
from wrolpi.dates import now, today, TZDateTime
from wrolpi.db import get_db_curs
from wrolpi.errors import UnknownVideo, NoFrequency, UnknownFile, UnknownDirectory


class ChannelPath(types.TypeDecorator):
    impl = types.String

    def process_bind_param(self, value, dialect):
        return str(value) if value else None

    def process_result_value(self, value, dialect):
        if value:
            return Path(value)


class Video(ModelHelper, Base):
    __tablename__ = 'video'
    id = Column(Integer, primary_key=True)
    channel_id = Column(Integer, ForeignKey('channel.id'))
    channel = relationship('Channel', primaryjoin='Video.channel_id==Channel.id', back_populates='videos')
    idempotency = Column(String)

    # File paths
    caption_path = Column(ChannelPath)
    description_path = Column(ChannelPath)
    ext = Column(String)
    info_json_path = Column(ChannelPath)
    poster_path = Column(ChannelPath)
    video_path = Column(ChannelPath)

    caption = Column(String)
    duration = Column(Integer)
    favorite = Column(TZDateTime)
    size = Column(Integer)
    source_id = Column(String)
    title = Column(String)
    upload_date = Column(TZDateTime)
    validated_poster = Column(Boolean, default=False)
    viewed = Column(TZDateTime)
    view_count = Column(Integer)
    textsearch = Column(tsvector, Computed('''to_tsvector('english'::regconfig,
                                               ((COALESCE(title, ''::text) || ' '::text) ||
                                                COALESCE(caption, ''::text)))'''))

    def __repr__(self):
        return f'<Video id={self.id}, title={self.title}, path={self.video_path}, channel={self.channel_id} ' \
               f'source_id={self.source_id}>'

    def dict(self) -> dict:
        d = super().dict()
        if self.channel_id:
            d['channel'] = self.channel.dict()
        d['info_json'] = self.get_info_json()
        return d

    def get_minimize(self) -> dict:
        """
        Get a dictionary representation of this Video suitable for sending out the API.
        """
        from .common import minimize_video
        return minimize_video(self.dict())

    def _clear_paths(self):
        self.caption_path = None
        self.description_path = None
        self.ext = None
        self.info_json_path = None
        self.poster_path = None
        self.video_path = None

    def delete(self):
        """
        Remove all files related to this video.  Add it to it's Channel's skip list.

        Raises: UnknownFile if a the video has no files.
        """
        from modules.videos.common import get_absolute_video_files
        video_files = get_absolute_video_files(self)
        for path in video_files:
            try:
                path.unlink()
            except FileNotFoundError:
                pass

        self._clear_paths()

        if not video_files:
            raise UnknownFile('No video files were deleted')

        self.add_to_skip_list()

    def add_to_skip_list(self):
        """
        Add this video to it's Channel's skip list.
        """
        self.channel.add_video_to_skip_list(self.source_id)

    def set_favorite(self, favorite: bool) -> Optional[datetime]:
        self.favorite = now() if favorite else None
        return self.favorite

    def set_viewed(self):
        self.viewed = now()

    def get_info_json(self) -> Optional[JSON]:
        """
        If this Video has an info_json file, return it's contents.  Otherwise, return None.
        """
        if not self.channel or not self.channel.directory:
            return

        try:
            if self.info_json_path:
                path = self.channel.get_relative_path(self.info_json_path)
                with open(path, 'rb') as fh:
                    contents = json.load(fh)
                    return contents
        except UnknownFile:
            pass
        except UnknownDirectory:
            pass

    def get_video_description(self) -> Optional[str]:
        """
        Get the Video description from the file system.
        """
        # First try to get description from info_json file.
        info_json = self.get_info_json()
        if info_json:
            description = info_json.get('description')
            if description:
                return description

        if self.description_path and self.channel_id:
            path = self.channel.get_relative_path(self.description_path)
            if path.exists():
                with open(path, 'rt') as fh:
                    contents = fh.read()
                    return contents

    def get_surrounding_videos(self):
        """
        Get the previous and next videos around this video.  The videos must be in the same channel.

        Example:
            >>> vid1 = Video(id=1, upload_date=10)
            >>> vid2 = Video(id=2, upload_date=20)
            >>> vid3 = Video(id=3, upload_date=30)
            >>> vid4 = Video(id=4)

            >>> vid1.get_surrounding_videos()
            (None, vid2)
            >>> vid2.get_surrounding_videos()
            (vid1, vid3)
            >>> vid3.get_surrounding_videos()
            (vid2, None)
            Video 4 has no upload date, so we can't place it in order.
            >>> vid4.get_surrounding_videos()
            (None, None)
        """
        session = Session.object_session(self)

        if not self.upload_date:
            # We can't place a video that has no upload date.
            return None, None

        with get_db_curs() as curs:
            query = '''
                    WITH numbered_videos AS (
                        SELECT id,
                            ROW_NUMBER() OVER (ORDER BY upload_date ASC) AS row_number
                        FROM video
                        WHERE
                            channel_id = %(channel_id)s
                            AND upload_date IS NOT NULL
                    )

                    SELECT id
                    FROM numbered_videos
                    WHERE row_number IN (
                        SELECT row_number+i
                        FROM numbered_videos
                        CROSS JOIN (SELECT -1 AS i UNION ALL SELECT 0 UNION ALL SELECT 1) n
                        WHERE
                        id = %(video_id)s
                    )
            '''
            curs.execute(query, dict(channel_id=self.channel_id, video_id=self.id))
            results = [i[0] for i in curs.fetchall()]

        # Assign the returned ID's to their respective positions relative to the ID that matches the video_id.
        previous_id = next_id = None
        for index, id_ in enumerate(results):
            if id_ == self.id:
                if index > 0:
                    previous_id = results[index - 1]
                if index + 1 < len(results):
                    next_id = results[index + 1]
                break

        # Fetch the videos by id, if they exist.
        previous_video = session.query(Video).filter_by(id=previous_id).one() if previous_id else None
        next_video = session.query(Video).filter_by(id=next_id).one() if next_id else None

        return previous_video, next_video


class Channel(ModelHelper, Base):
    __tablename__ = 'channel'
    id = Column(Integer, primary_key=True)
    name = Column(String)
    link = Column(String, nullable=False)
    idempotency = Column(String)
    url = Column(String)
    match_regex = Column(String)
    directory = Column(PathColumn)
    skip_download_videos = Column(ARRAY(String))
    generate_posters = Column(Boolean)
    calculate_duration = Column(Boolean)
    download_frequency = Column(Integer)
    next_download = Column(Date)

    info_json = Column(JSON)
    info_date = Column(Date)

    # Will be filled later by directory path relative to the media directory.
    _directory = None

    videos: InstrumentedList = relationship('Video', primaryjoin='Channel.id==Video.channel_id')

    def __repr__(self):
        return f'<Channel(id={self.id}, name={self.name})>'

    def add_video_to_skip_list(self, source_id):
        if not source_id:
            raise UnknownVideo(f'Cannot skip video with empty source id: {source_id}')

        skip_download_videos = {i for i in self.skip_download_videos or [] if i}
        skip_download_videos.add(source_id)
        self.skip_download_videos = skip_download_videos

    def increment_next_download(self):
        """
        Set the next download predictably during the next download iteration.

        For example, two channels that download weekly will need to be downloaded on different days.  We want a channel
        to always be downloaded on it's day.  That may be Monday, or Tuesday, etc.

        This is true for all download frequencies (30 days, 90 days, etc.).

        The order that channels will be downloaded/distributed will be by `link`.
        """
        if not self.download_frequency:
            raise NoFrequency(f'{self} does not have a frequency!')

        session = Session.object_session(self)

        # All the channels that share the my frequency.
        channel_group = session.query(self.__class__).filter_by(download_frequency=self.download_frequency)
        channel_group = list(channel_group.order_by(self.__class__.link).all())

        # My position in the channel group.
        index = channel_group.index(self)

        # The seconds between each download.
        chunk = self.download_frequency // len(channel_group)

        # My next download will be distributed by my frequency and my position.
        position = chunk * (index + 1)
        self.next_download = today() + timedelta(seconds=self.download_frequency + position)

    def delete_with_videos(self):
        """
        Delete all Video records (but not video files) related to this Channel.  Then delete the Channel.
        """
        session = Session.object_session(self)
        # Delete the video records, but not the video files!
        session.query(Video).filter_by(channel_id=self.id).delete()
        session.delete(self)

    def update(self, data: dict):
        for key, value in data.items():
            setattr(self, key, value)

    def config_view(self) -> dict:
        """
        Retrieve the data about this Channel that should be stored in a config file.
        """
        config = dict(
            directory=str(self.directory),
            match_regex=self.match_regex or '',
            name=self.name,
            url=self.url or '',
            generate_posters=self.generate_posters,
            calculate_duration=self.calculate_duration,
            skip_download_videos=self.skip_download_videos or [],
            download_frequency=self.download_frequency,
            next_download=self.next_download,
            favorites={},
        )

        session = Session.object_session(self)
        favorites = session.query(Video).filter(Video.favorite != None, Video.channel_id == self.id).all()
        if favorites:
            config['favorites'] = {i.source_id: {'favorite': i.favorite} for i in favorites}

        return config

    def get_directory(self):
        if self._directory:
            return self._directory

        self._directory = get_absolute_media_path(self.directory)
        return self._directory

    def get_relative_path(self, path: Path, exists: bool = True):
        channel_dir = self._directory or self.get_directory()
        path = channel_dir / path
        if exists and not path.exists():
            raise FileNotFoundError(f'Channel path {path} does not exist!')
        return path