Files
website/scripts/music-import-history.ts
Stan Runge b757269151
All checks were successful
Deploy production / deploy (push) Successful in 2m37s
music: self-hosted scrobbler (Spotify poller, Last.fm + streaming-history importers, /trackers/music, homepage listening card)
- music schema: artists/albums/tracks (+album art), plays with source + dedupe, now_playing,
  spotify_auth, import_state; plays_deduped view hides Last.fm scrobbles Spotify also has
- in-app Spotify poller (60s, leased) started from hooks init; no new role or env needed
- /api/spotify/connect + callback (owner only), /trackers/music public stats
- scripts: music-poll, music-import-lastfm (resumable), music-import-history
- deploy/music-collector: optional host timer with its own role
2026-10-24 15:57:40 +00:00

70 lines
2.7 KiB
TypeScript

// Import a Spotify data export into music.plays (source 'spotify', origin 'export').
//
// bun scripts/music-import-history.ts <file-or-dir> [...] [--min-ms 30000] [--dry-run]
//
// Accepts the Extended Streaming History (Streaming_History_Audio_*.json, older
// endsong_*.json) and the short account-data export (StreamingHistory_music_*.json).
// Podcasts/audiobooks and plays under --min-ms are skipped. Safe to re-run: a play already
// stored (same artist+title within a minute, also from the live poller) is not added twice.
import { readdirSync, readFileSync, statSync } from 'node:fs';
import { join } from 'node:path';
import { DEFAULT_MIN_MS, parseHistoryEntry, type HistoryPlay } from '../src/lib/music/history';
import { insertPlay, newCache, upsertTrack } from '../src/lib/music/store.server';
import { arg, musicDb } from './music-db';
const minMs = Number(arg('min-ms') ?? DEFAULT_MIN_MS);
const dryRun = process.argv.includes('--dry-run');
const skip = new Set(['--min-ms', arg('min-ms')]);
const inputs = process.argv.slice(2).filter((a) => !a.startsWith('--') && !skip.has(a));
if (!inputs.length) throw new Error('Usage: bun scripts/music-import-history.ts <file-or-dir> ...');
const NAME = /^(Streaming_History_Audio_.*|endsong_?\d*|StreamingHistory(_music)?_?\d*)\.json$/i;
function files(path: string): string[] {
if (statSync(path).isDirectory())
return readdirSync(path).flatMap((name) => {
const child = join(path, name);
return statSync(child).isDirectory() ? files(child) : NAME.test(name) ? [child] : [];
});
return [path];
}
const plays: HistoryPlay[] = [];
let entries = 0;
for (const file of inputs.flatMap(files)) {
const data = JSON.parse(readFileSync(file, 'utf8')) as unknown[];
if (!Array.isArray(data)) continue;
entries += data.length;
for (const entry of data) {
const play = parseHistoryEntry(entry, minMs);
if (play) plays.push(play);
}
console.log(`${file}: ${data.length} entries`);
}
plays.sort((a, b) => a.playedAt.getTime() - b.playedAt.getTime());
console.log(`${plays.length} music plays of ${minMs / 1000}s or more out of ${entries} entries.`);
if (dryRun || !plays.length) process.exit(0);
const sql = musicDb();
try {
const cache = newCache();
let stored = 0;
for (const [i, play] of plays.entries()) {
const trackId = await upsertTrack(sql, play.track, false, cache);
if (
await insertPlay(sql, {
trackId,
playedAt: play.playedAt,
source: 'spotify',
origin: 'export',
msPlayed: play.msPlayed,
platform: play.platform
})
)
stored++;
if ((i + 1) % 5000 === 0) console.log(` ${i + 1}/${plays.length} (${stored} new)`);
}
console.log(`Done: ${stored} new plays stored.`);
} finally {
await sql.end();
}