Trixy7/plugins/music/sources/soundcloud.py

# -*- coding: utf-8 -*-
"""
SoundCloud-Quelle via yt-dlp.

yt-dlp unterstuetzt SoundCloud nativ — kein zusaetzliches
Paket noetig. Suche und Download laufen komplett ueber yt-dlp.

Vorteile gegenueber YouTube:
    - Viele Indie/DJ/Remix-Tracks die auf YouTube nicht sind
    - Hoehere Audioqualitaet (oft 128-320kbps)
    - Weniger Werbung/Filler in den Suchergebnissen

Abhaengigkeiten:
    - yt-dlp (gleiche wie YouTube — shared dependency)
"""

from __future__ import annotations

import asyncio
import re
from pathlib import Path
from typing import AsyncIterator

from trixy_core.music.sources.base import MusicSource, SourceType, SourceState
from trixy_core.music.track import Track, TrackMetadata, TrackState
from trixy_core.utils.debug import pinfo, pdebug, perror, pwarn

# Optionale Abhaengigkeit (gleiche wie YouTube)
_YT_DLP_AVAILABLE = False
try:
    import yt_dlp
    _YT_DLP_AVAILABLE = True
except ImportError:
    pass


class SoundCloudSource(MusicSource):
    """SoundCloud-Quelle via yt-dlp (optional)."""

    def __init__(
        self,
        name: str = "SoundCloud",
        cache_dir: str = "cache/soundcloud",
        max_cache_size_mb: int = 500,
        max_results: int = 5,
    ) -> None:
        super().__init__(name=name, source_type=SourceType.URL)
        self._cache_dir = Path(cache_dir)
        self._max_cache_size_mb = max_cache_size_mb
        self._max_results = max_results

    async def connect(self) -> bool:
        """Prueft ob yt-dlp installiert ist."""
        if not _YT_DLP_AVAILABLE:
            pwarn("SoundCloudSource: yt-dlp nicht installiert")
            self._state = SourceState.UNAVAILABLE
            return False

        try:
            self._cache_dir.mkdir(parents=True, exist_ok=True)
            self._state = SourceState.AVAILABLE
            pinfo("SoundCloudSource: Verbunden")
            return True
        except OSError as e:
            perror(f"SoundCloudSource: Cache-Verzeichnis Fehler: {e}")
            self._state = SourceState.ERROR
            return False

    async def disconnect(self) -> None:
        self._state = SourceState.UNAVAILABLE

    async def scan(self) -> int:
        return 0

    async def get_track(self, track_id: str) -> Track | None:
        return self._tracks.get(track_id)

    async def get_audio_stream(self, track: Track) -> AsyncIterator[bytes]:
        if not track.path or not track.path.exists():
            return
        chunk_size = 65536
        def read_file():
            with open(track.path, "rb") as f:
                while True:
                    chunk = f.read(chunk_size)
                    if not chunk:
                        break
                    yield chunk
        for chunk in await asyncio.to_thread(lambda: list(read_file())):
            yield chunk

    async def get_metadata(self, track: Track) -> TrackMetadata | None:
        return track.metadata

    async def search(self, query: str, limit: int = 5) -> list[Track]:
        """
        Sucht auf SoundCloud via yt-dlp.

        Nutzt den yt-dlp SoundCloud-Extractor: `scsearch{limit}:query`
        """
        if not _YT_DLP_AVAILABLE or self._state != SourceState.AVAILABLE:
            return []

        effective_limit = min(limit, self._max_results)

        ydl_opts = {
            "quiet": True,
            "no_warnings": True,
            "extract_flat": False,
            "skip_download": True,
            "default_search": f"scsearch{effective_limit}",
        }

        try:
            results = await asyncio.to_thread(self._search_sync, query, ydl_opts)
            return results
        except Exception as e:
            perror(f"SoundCloudSource: Suchfehler: {e}")
            return []

    def _search_sync(self, query: str, ydl_opts: dict) -> list[Track]:
        """Synchrone SoundCloud-Suche."""
        tracks: list[Track] = []

        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            info = ydl.extract_info(query, download=False)
            entries = info.get("entries", []) if info else []
            if not entries and info:
                entries = [info]

            for entry in entries:
                if not entry:
                    continue

                track_id = str(entry.get("id", ""))
                title = entry.get("title", "Unbekannt")
                duration = entry.get("duration", 0)
                uploader = entry.get("uploader", "")
                thumbnail = entry.get("thumbnail", "")
                webpage_url = entry.get("webpage_url", "")

                # Artist/Title: SoundCloud hat oft "Artist - Title" im Titel
                artist, track_title = self._parse_title(title, uploader)

                metadata = TrackMetadata(
                    title=track_title,
                    artist=artist,
                    duration_ms=int(duration * 1000) if duration else 0,
                    artwork_url=thumbnail,
                    genre=entry.get("genre", ""),
                )

                track = Track(
                    uri=f"soundcloud:{track_id}",
                    source_type="soundcloud",
                    source_id=track_id,
                    metadata=metadata,
                    state=TrackState.AVAILABLE,
                    extra={
                        "sc_id": track_id,
                        "url": webpage_url,
                        "like_count": entry.get("like_count", 0),
                        "repost_count": entry.get("repost_count", 0),
                    },
                )
                tracks.append(track)

        return tracks

    def _parse_title(self, title: str, uploader: str) -> tuple[str, str]:
        """Parst SoundCloud-Titel in Artist und Track-Titel."""
        if " - " in title:
            parts = title.split(" - ", 1)
            return parts[0].strip(), parts[1].strip()
        if " | " in title:
            parts = title.split(" | ", 1)
            return parts[0].strip(), parts[1].strip()
        # Suffixe entfernen
        cleaned = re.sub(
            r'\s*[\(\[\{].*?(official|free|download|remix|bootleg).*?[\)\]\}]',
            '', title, flags=re.IGNORECASE
        ).strip()
        return uploader or "Unbekannt", cleaned or title

    async def prepare_track(self, track: Track) -> bool:
        """Laedt Audio herunter via yt-dlp."""
        if not _YT_DLP_AVAILABLE or self._state != SourceState.AVAILABLE:
            return False

        sc_url = track.extra.get("url", "")
        if not sc_url:
            perror("SoundCloudSource: Keine SoundCloud-URL")
            return False

        # Cache pruefen
        cached = self._get_cached_path(track.source_id)
        if cached:
            track.path = cached
            track.state = TrackState.AVAILABLE
            pdebug(f"SoundCloudSource: Cache-Treffer: {cached.name}")
            return True

        track.state = TrackState.LOADING
        try:
            output_path = await asyncio.to_thread(self._download_sync, sc_url, track.source_id)
            if output_path and output_path.exists():
                track.path = output_path
                track.state = TrackState.AVAILABLE
                pinfo(f"SoundCloudSource: Heruntergeladen: {output_path.name}")
                await self.cleanup_cache()
                return True
            else:
                track.state = TrackState.ERROR
                track.error_message = "Download fehlgeschlagen"
                return False
        except Exception as e:
            perror(f"SoundCloudSource: Download-Fehler: {e}")
            track.state = TrackState.ERROR
            track.error_message = str(e)
            return False

    def _download_sync(self, url: str, track_id: str) -> Path | None:
        """Synchroner Download via yt-dlp."""
        output_template = str(self._cache_dir / f"{track_id}.%(ext)s")

        ydl_opts = {
            "format": "bestaudio/best",
            "outtmpl": output_template,
            "quiet": True,
            "no_warnings": True,
            "postprocessors": [{
                "key": "FFmpegExtractAudio",
                "preferredcodec": "mp3",
                "preferredquality": "192",
            }],
        }

        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            ydl.download([url])

        expected = self._cache_dir / f"{track_id}.mp3"
        if expected.exists():
            return expected

        for f in self._cache_dir.iterdir():
            if f.stem == track_id and f.is_file():
                return f
        return None

    def _get_cached_path(self, track_id: str) -> Path | None:
        if not self._cache_dir.exists():
            return None
        for f in self._cache_dir.iterdir():
            if f.stem == track_id and f.is_file():
                return f
        return None

    async def cleanup_cache(self) -> int:
        if not self._cache_dir.exists():
            return 0
        return await asyncio.to_thread(self._cleanup_cache_sync)

    def _cleanup_cache_sync(self) -> int:
        files: list[tuple[Path, float, int]] = []
        total_size = 0
        for f in self._cache_dir.iterdir():
            if f.is_file():
                stat = f.stat()
                files.append((f, stat.st_mtime, stat.st_size))
                total_size += stat.st_size

        max_size = self._max_cache_size_mb * 1024 * 1024
        if total_size <= max_size:
            return 0

        files.sort(key=lambda x: x[1])
        removed = 0
        for file_path, _, size in files:
            if total_size <= max_size:
                break
            try:
                file_path.unlink()
                total_size -= size
                removed += 1
            except OSError:
                pass
        if removed > 0:
            pdebug(f"SoundCloudSource: {removed} Cache-Dateien entfernt")
        return removed