Files
musicmouse/python-backend/musicmouse/library/cache.py
2026-08-27 23:46:19 +02:00

205 lines
7.3 KiB
Python

"""Where scan results live between runs.
A directory rather than a single file, because the three kinds of content cost wildly
different amounts to produce::
<cache_dir>/
├── index.json cheap: tags and structure. Thrown away freely.
├── covers/<album_id>.jpg medium: art pulled out of an ID3 APIC frame
└── analysis/<track_key>.json expensive: minutes of DSP per track
analysis/<track_key>.beats.json
That split is the whole point. A rescan must be free to rebuild ``index.json`` without
destroying analysis, so everything expensive is keyed by a *content* key (see
:func:`~musicmouse.library.models.track_key`) rather than by album id - renaming a
folder or re-sorting a section then costs nothing.
"""
from __future__ import annotations
import json
import logging
import os
from dataclasses import dataclass
from pathlib import Path
from typing import Any, cast
from musicmouse.library.analysis import BeatGrid, TrackAnalysis
from musicmouse.library.models import Album, LibraryTrack
from musicmouse.library.sections import SECTIONS, AlbumKind
_log = logging.getLogger(__name__)
__all__ = ["Fingerprint", "LibraryCache"]
#: Bump whenever the scanner's *derivation* changes - how a title, artist or series is
#: worked out - not just when the JSON shape does. A cached entry is reused whenever its
#: files are untouched, so otherwise a change to that logic is invisible until somebody
#: edits their music folder.
_INDEX_VERSION = 4
@dataclass(frozen=True, slots=True)
class Fingerprint:
"""What makes a folder's cached entry still valid: its files, sizes and mtimes."""
files: tuple[tuple[str, int, int], ...]
@classmethod
def of(cls, paths: list[Path]) -> Fingerprint:
entries: list[tuple[str, int, int]] = []
for path in paths:
stat = path.stat()
entries.append((path.name, stat.st_size, int(stat.st_mtime)))
return cls(tuple(entries))
def to_json(self) -> list[list[Any]]:
return [list(entry) for entry in self.files]
@classmethod
def from_json(cls, data: list[list[Any]]) -> Fingerprint:
return cls(tuple((str(n), int(s), int(m)) for n, s, m in data))
def _write_atomic(path: Path, payload: str) -> None:
"""Write via a sibling temp file so a crash never leaves a half-written cache."""
temp = path.with_name(f"{path.name}.tmp{os.getpid()}")
temp.write_text(payload, encoding="utf-8")
temp.replace(path)
class LibraryCache:
def __init__(self, directory: Path) -> None:
self.directory = directory
self.covers = directory / "covers"
self.analysis = directory / "analysis"
def prepare(self) -> None:
for folder in (self.directory, self.covers, self.analysis):
folder.mkdir(parents=True, exist_ok=True)
# -------------------------------------------------------------------- covers
def cover_path(self, album_id: str) -> Path:
return self.covers / f"{album_id}.jpg"
def store_cover(self, album_id: str, data: bytes) -> Path:
path = self.cover_path(album_id)
path.write_bytes(data)
return path
# ------------------------------------------------------------------ analysis
def load_analysis(self, key: str) -> TrackAnalysis | None:
path = self.analysis / f"{key}.json"
try:
return TrackAnalysis.from_json(json.loads(path.read_text(encoding="utf-8")))
except (OSError, ValueError):
return None
def store_analysis(self, key: str, analysis: TrackAnalysis) -> None:
_write_atomic(self.analysis / f"{key}.json", json.dumps(analysis.to_json()))
def load_beats(self, key: str) -> BeatGrid | None:
path = self.analysis / f"{key}.beats.json"
try:
return BeatGrid.from_json(json.loads(path.read_text(encoding="utf-8")))
except (OSError, ValueError, KeyError):
return None
def store_beats(self, key: str, grid: BeatGrid) -> None:
_write_atomic(self.analysis / f"{key}.beats.json", json.dumps(grid.to_json()))
# --------------------------------------------------------------------- index
def load_index(self) -> dict[str, tuple[Album, Fingerprint]]:
path = self.directory / "index.json"
try:
raw = json.loads(path.read_text(encoding="utf-8"))
except (OSError, ValueError):
return {}
if raw.get("version") != _INDEX_VERSION:
_log.info("Library index is from an older version; rescanning from scratch")
return {}
out: dict[str, tuple[Album, Fingerprint]] = {}
for entry in raw.get("albums", []):
try:
out[entry["id"]] = (_album_from_json(entry), Fingerprint.from_json(entry["files"]))
except (KeyError, TypeError, ValueError):
_log.debug("Dropping unreadable index entry %r", entry.get("id"))
return out
def store_index(self, albums: dict[str, tuple[Album, Fingerprint]]) -> None:
payload = {
"version": _INDEX_VERSION,
"albums": [
_album_to_json(album) | {"files": fingerprint.to_json()}
for album, fingerprint in albums.values()
],
}
self.prepare()
_write_atomic(self.directory / "index.json", json.dumps(payload, ensure_ascii=False))
# ------------------------------------------------------------------ serialisation
def _album_to_json(album: Album) -> dict[str, Any]:
return {
"id": album.id,
"section": album.section,
"kind": album.kind,
"title": album.title,
"artist": album.artist,
"series": album.series,
"figure": album.figure,
"colors": list(album.colors),
"folder": str(album.folder),
"cover": str(album.cover) if album.cover else None,
"tracks": [
{
"path": str(track.path),
"title": track.title,
"duration": track.duration,
"analysis": track.analysis.to_json() if track.analysis else None,
}
for track in album.tracks
],
}
def _album_from_json(data: dict[str, Any]) -> Album:
section = str(data["section"])
if section not in SECTIONS:
raise ValueError(f"unknown section {section!r}")
# Read back what was stored rather than re-deriving it from the section: a figure's
# kind comes from the config, so the section cannot answer for it.
kind = str(data["kind"])
if kind not in ("music", "book"):
raise ValueError(f"unknown kind {kind!r}")
red, green, blue = data["colors"]
return Album(
id=str(data["id"]),
section=section,
kind=cast("AlbumKind", kind),
title=str(data["title"]),
artist=str(data["artist"]),
series=data["series"],
figure=data["figure"],
colors=(str(red), str(green), str(blue)),
folder=Path(data["folder"]),
cover=Path(data["cover"]) if data["cover"] else None,
tracks=tuple(
LibraryTrack(
path=Path(track["path"]),
title=str(track["title"]),
duration=float(track["duration"]),
analysis=(
TrackAnalysis.from_json(track["analysis"]) if track.get("analysis") else None
),
)
for track in data["tracks"]
),
)