Files
ytplayer/server/server.js

602 lines
24 KiB
JavaScript

/* ============================================================================
* server.js — YT Player PWA backend
*
* Runtime: Bun (https://bun.sh)
* Framework: Hono v4
* DB: libsql (concurrent SQLite fork, embedded file mode)
*
* Endpoints:
* GET /api/search?q=<query> yt-dlp search → slim card array
* GET /api/channel?c=<channel> yt-dlp channel uploads → slim card array
* GET /api/streams?v=<videoId> yt-dlp stream info → {meta, audioUrl, qualities}
* GET /api/download/:videoId proxy best progressive stream → binary
* GET /api/version { version }
* POST /api/user/sync upsert user playlists + last-seen version
* GET /api/user/data?fp=<fp> retrieve stored playlists + history
* GET /* serve frontend/public static files
*
* JSON shapes mirror the Tauri (Rust) bridge exactly so the existing app.js
* UI code works without modification in WEB mode.
* ========================================================================== */
import { Hono } from 'hono';
import { serveStatic } from 'hono/bun';
import { logger } from 'hono/logger';
import { spawn } from 'node:child_process';
import { createServer } from 'node:http';
import { readFileSync, readdirSync, statSync, openSync, unlinkSync, createReadStream } from 'node:fs';
import { Readable } from 'node:stream';
import { tmpdir } from 'node:os';
import { createHash } from 'node:crypto';
import { initDb, upsertUser, recordVideoAccess, getUserData, createProfile, getProfile, saveProfile } from './db.js';
const PORT = parseInt(process.env.PORT || '3000', 10);
const APP_VERSION = process.env.APP_VERSION || '1.0.0';
const YTDLP = process.env.YTDLP_PATH || 'yt-dlp';
// ----------------------------------------------------------------------------
// BUILD_TAG — must be DETERMINISTIC across restarts of identical code.
//
// Previously this was `Date.now().toString(36)`, which changes every time the
// process starts even if nothing was deployed (crash-loop, healthcheck
// restart, container reschedule). The frontend's checkBuildTag() polls
// /api/version and re-shows the "Update available" modal the instant the tag
// drifts — so a restarting-but-unchanged server kept re-announcing an update
// that never actually happened, and clicking "Refresh UI" (which itself
// reloads the page and re-polls) never made the prompt go away for good.
//
// Fix: hash the actual served frontend files. Identical code → identical
// hash → identical tag, no matter how many times the process restarts. A
// real deploy (changed files) still produces a new tag as intended.
// process.env.BUILD_TAG still wins if a CI pipeline already injects a git
// SHA — that's an even better source of truth than a content hash.
// ----------------------------------------------------------------------------
function computeBuildTag() {
try {
// Hash EVERY served frontend file (recursively, in sorted order), not a
// hand-picked subset — a change to any shell file (e.g. sw-update.js or
// opfs.js) must produce a new tag, or clients keep their old SW cache
// and never receive the change.
const hash = createHash('sha256');
const walk = (dir) => {
for (const name of readdirSync(dir).sort()) {
const path = `${dir}/${name}`;
if (statSync(path).isDirectory()) walk(path);
else { hash.update(path); hash.update(readFileSync(path)); }
}
};
walk('./public');
return hash.digest('hex').slice(0, 12);
} catch {
// Frontend files not readable (e.g. unit tests run outside ./public) —
// fall back to a fixed tag rather than Date.now(), so it still never
// drifts spuriously between restarts.
return 'dev-build';
}
}
const BUILD_TAG = process.env.BUILD_TAG || computeBuildTag();
// BUILD_TIME — human-readable "when was this image built". Written by the
// Dockerfile at image build time (never at container start, so restarts
// don't drift it). Kept OUTSIDE ./public so it can't perturb BUILD_TAG.
const BUILD_TIME = process.env.BUILD_TIME || (() => {
try { return readFileSync('./build-time.txt', 'utf8').trim(); }
catch { return null; }
})();
const SEARCH_LIMIT = 25;
const CHANNEL_LIMIT = 60;
// ============================================================================
// yt-dlp helpers
// ============================================================================
// Run yt-dlp asynchronously and resolve stdout as a string.
// MUST stay async (spawn, not spawnSync): a sync child process blocks Bun's
// event loop for the full yt-dlp runtime (~2-3s per call), which stalls every
// concurrent request — including in-flight /api/download proxy streams, which
// Bun then kills at its idle timeout ("fetch failed" mid-download on clients).
// Rejects on non-zero exit.
function runYtdlp(args) {
return new Promise((resolve, reject) => {
const child = spawn(YTDLP, args, { stdio: ['ignore', 'pipe', 'pipe'] });
let out = '';
let err = '';
child.stdout.setEncoding('utf8');
child.stderr.setEncoding('utf8');
child.stdout.on('data', (d) => { out += d; });
child.stderr.on('data', (d) => { err += d; });
child.on('error', (e) => reject(new Error('yt-dlp not found: ' + e.message)));
child.on('close', (code) => {
if (code !== 0) reject(new Error(err.trim() || 'yt-dlp exited with code ' + code));
else resolve(out);
});
});
}
// Helpers to pick the right field from a yt-dlp JSON record
function pick(obj, ...keys) {
for (const k of keys) {
const v = obj[k];
if (v && typeof v === 'string' && v.trim()) return v.trim();
}
return '';
}
function pickChannel(obj) { return pick(obj, 'channel', 'uploader'); }
function pickChannelUrl(obj) { return pick(obj, 'channel_url', 'uploader_url'); }
function pickChannelId(obj) { return pick(obj, 'channel_id', 'uploader_id'); }
// Normalise a flat-playlist yt-dlp record into the slim UI card shape
function slimEntry(j) {
const id = pick(j, 'id');
if (!id) return null;
return {
id,
title: pick(j, 'title') || '(untitled)',
channel: pickChannel(j),
channelId: pickChannelId(j),
channelUrl: pickChannelUrl(j),
duration: typeof j.duration === 'number' ? j.duration : 0,
thumbnail: `https://i.ytimg.com/vi/${id}/mqdefault.jpg`,
};
}
// Parse multi-line JSON output from yt-dlp --dump-json --flat-playlist
function parseCards(output) {
const results = [];
for (const line of output.split('\n')) {
const t = line.trim();
if (!t) continue;
try {
const j = JSON.parse(t);
const card = slimEntry(j);
if (card) results.push(card);
} catch { /* skip malformed lines */ }
}
return results;
}
// Resolve a channel identifier to a /videos URL yt-dlp can fetch
function channelToUrl(c) {
let base = c.trim();
if (!base.startsWith('http')) {
base = base.startsWith('@') ? `https://www.youtube.com/${base}`
: base.startsWith('UC') ? `https://www.youtube.com/channel/${base}`
: `https://www.youtube.com/@${base}`;
}
base = base.replace(/\/$/, '');
return base.endsWith('/videos') ? base : base + '/videos';
}
// ============================================================================
// App setup
// ============================================================================
const app = new Hono();
app.use('*', logger());
// ============================================================================
// API routes
// ============================================================================
// GET /api/version
// Returns version string + a build tag that changes on every server restart/deploy.
// Clients poll this to detect when a new build is live and prompt a reload.
app.get('/api/version', (c) =>
c.json(
{ version: APP_VERSION, buildTag: BUILD_TAG, buildTime: BUILD_TIME },
200,
{ 'Cache-Control': 'no-store, no-cache, must-revalidate' }
)
);
// GET /api/search?q=<query>
app.get('/api/search', async (c) => {
const q = (c.req.query('q') || '').trim();
if (!q) return c.json({ ok: false, error: 'empty query' }, 400);
try {
const out = await runYtdlp([
`ytsearch${SEARCH_LIMIT}:${q}`,
'--dump-json', '--flat-playlist',
'--no-warnings', '--ignore-errors',
]);
return c.json({ ok: true, results: parseCards(out) });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// GET /api/channel?c=<channel>
app.get('/api/channel', async (c) => {
const chan = (c.req.query('c') || '').trim();
if (!chan) return c.json({ ok: false, error: 'missing channel' }, 400);
try {
const url = channelToUrl(chan);
const out = await runYtdlp([
url,
'--dump-json', '--flat-playlist',
'--no-warnings', '--ignore-errors',
'--playlist-end', String(CHANNEL_LIMIT),
]);
const results = parseCards(out);
// Extract channel name + URL from the first record
const first = results[0];
let channelName = '', channelUrl = '';
for (const line of out.split('\n')) {
const t = line.trim();
if (!t) continue;
try {
const j = JSON.parse(t);
channelName = channelName || pickChannel(j);
channelUrl = channelUrl || pickChannelUrl(j);
if (channelName && channelUrl) break;
} catch { /* skip */ }
}
return c.json({ ok: true, channel: channelName || (first?.channel || ''), channelUrl, results });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// GET /api/streams?v=<videoId>
app.get('/api/streams', async (c) => {
const videoId = (c.req.query('v') || '').replace(/[/\\:?<>|*"]/g, '').trim();
if (!videoId) return c.json({ ok: false, error: 'missing videoId' }, 400);
try {
const url = `https://www.youtube.com/watch?v=${videoId}`;
const out = await runYtdlp(['-J', '--no-warnings', url]);
const info = JSON.parse(out);
const formats = Array.isArray(info.formats) ? info.formats : [];
// Best audio-only stream (prefer mp4a/m4a by bitrate)
let bestAudioUrl = null;
let bestAudioScore = -1;
for (const f of formats) {
if (!f.url) continue;
const hasVideo = f.vcodec && f.vcodec !== 'none';
const hasAudio = f.acodec && f.acodec !== 'none';
if (hasVideo || !hasAudio) continue;
let score = f.abr || 0;
if (f.acodec && f.acodec.includes('mp4a')) score += 1000;
if (score > bestAudioScore) { bestAudioScore = score; bestAudioUrl = f.url; }
}
// Quality list — progressive (single-file, hasAudio) first, then adaptive
const qualities = [];
const seen = new Set();
for (const wantProg of [true, false]) {
for (const f of formats) {
if (!f.url) continue;
const hasVideo = f.vcodec && f.vcodec !== 'none';
const hasAudio = f.acodec && f.acodec !== 'none';
if (!hasVideo) continue;
if ((hasAudio) !== wantProg) continue;
const h = f.height || 0;
if (h <= 0 || seen.has(h)) continue;
seen.add(h);
qualities.push({ label: h + 'p', height: h, hasAudio: !!hasAudio, url: f.url, ext: f.ext || '' });
}
}
qualities.sort((a, b) => b.height - a.height);
return c.json({
ok: true,
data: {
meta: {
id: videoId,
title: pick(info, 'title') || '(untitled)',
channel: pickChannel(info),
channelId: pickChannelId(info),
channelUrl: pickChannelUrl(info),
duration: typeof info.duration === 'number' ? info.duration : 0,
thumbnail: `https://i.ytimg.com/vi/${videoId}/hqdefault.jpg`,
},
audioUrl: bestAudioUrl,
qualities,
},
});
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// GET /api/download/:videoId
// Resolves the best progressive (audio+video single-file) stream URL via
// yt-dlp and proxies the binary to the browser so OPFS can store it.
// The browser never contacts YouTube CDN directly (CORS would block it).
app.get('/api/download/:videoId', async (c) => {
const videoId = (c.req.param('videoId') || '').replace(/[/\\:?<>|*"]/g, '').trim();
if (!videoId) return c.json({ ok: false, error: 'missing videoId' }, 400);
try {
const url = `https://www.youtube.com/watch?v=${videoId}`;
// ?mux=1 — "Save before playing" path: let yt-dlp download bestvideo up
// to 720p PLUS bestaudio and compile them into one mp4 with ffmpeg on
// the server, then stream the finished file. Default saves (no mux)
// keep the progressive single-stream behavior below, unchanged.
if (c.req.query('mux') === '1') {
const tmp = `${tmpdir()}/ytp-mux-${videoId}-${Date.now()}.mp4`;
try {
await runYtdlp([
url,
'--no-warnings', '--no-playlist',
'-f', 'bv*[height<=720][ext=mp4]+ba[ext=m4a]/bv*[height<=720]+ba/b[ext=mp4]/b',
'--merge-output-format', 'mp4',
'-N', '4',
'-o', tmp,
]);
const size = statSync(tmp).size;
// Open the fd first, then unlink: on Linux the data stays readable
// until the fd closes, so the temp file cleans itself up even if the
// client disconnects mid-transfer.
const fd = openSync(tmp, 'r');
unlinkSync(tmp);
const stream = createReadStream('', { fd });
const fp = c.req.query('fp');
if (fp) recordVideoAccess(fp, { id: videoId }).catch(() => {});
return new Response(Readable.toWeb(stream), {
status: 200,
headers: {
'Content-Type': 'video/mp4',
'Content-Length': String(size),
'Content-Disposition': `attachment; filename="${videoId}.mp4"`,
'Cache-Control': 'no-store',
'Access-Control-Allow-Origin': '*',
},
});
} catch (err) {
// ffmpeg missing or extraction failed — clean up any partial output
// and fall through to the progressive proxy below.
for (const leftover of [tmp, tmp + '.part']) {
try { unlinkSync(leftover); } catch { /* not created */ }
}
console.warn(`[ytplayer] mux download failed for ${videoId}, falling back to progressive:`, err.message);
}
}
// --get-url with bestvideo+bestaudio/best format is not what we want
// here — we need a SINGLE FILE so no ffmpeg muxing is required in the
// browser. Use -f "bestvideo[ext=mp4][acodec!=none]/best[ext=mp4]/best"
const out = await runYtdlp([
url,
'--no-warnings',
'-f', 'bestvideo[ext=mp4][acodec!=none]/bestvideo[acodec!=none]/best[ext=mp4]/best',
'--get-url',
]);
const streamUrl = out.trim();
if (!streamUrl) throw new Error('No stream URL returned');
// Fetch from YouTube and pipe to the client
const upstream = await fetch(streamUrl, {
headers: {
// Mimic a browser to avoid 403s from YouTube CDN
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.36',
'Referer': 'https://www.youtube.com/',
},
});
if (!upstream.ok) throw new Error(`Upstream error ${upstream.status}`);
const contentType = upstream.headers.get('content-type') || 'video/mp4';
const contentLength = upstream.headers.get('content-length');
const headers = new Headers({
'Content-Type': contentType,
'Content-Disposition': `attachment; filename="${videoId}.mp4"`,
'Cache-Control': 'no-store',
'Access-Control-Allow-Origin': '*',
});
if (contentLength) headers.set('Content-Length', contentLength);
// Log the access if a fingerprint was supplied (fire-and-forget)
const fp = c.req.query('fp');
if (fp) {
recordVideoAccess(fp, { id: videoId }).catch(() => {});
}
return new Response(upstream.body, { status: 200, headers });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// ============================================================================
// Online profiles — named cross-device sync. The NAME IS THE PASSKEY:
// anyone who knows it can load and overwrite that profile, so clients are
// encouraged to use the random generator. No other auth by design.
// ============================================================================
const PROFILE_NAME_RE = /^[A-Za-z0-9][A-Za-z0-9_-]{2,39}$/;
const PROFILE_MAX_BYTES = 2_000_000; // full data blob; typical payloads are ~KBs
const RAND_ADJ = ['amber', 'brave', 'calm', 'coral', 'crimson', 'dusty', 'gentle', 'golden',
'hidden', 'ivory', 'jade', 'lunar', 'mellow', 'misty', 'noble', 'quiet',
'rapid', 'silver', 'solar', 'stormy', 'swift', 'velvet', 'wild', 'zesty'];
const RAND_NOUN = ['falcon', 'harbor', 'willow', 'ember', 'canyon', 'meadow', 'otter', 'pine',
'raven', 'reef', 'sparrow', 'summit', 'thicket', 'tundra', 'brook', 'cedar',
'dune', 'fjord', 'glade', 'heron', 'lagoon', 'maple', 'prairie', 'wren'];
function randomProfileName() {
const a = RAND_ADJ[Math.floor(Math.random() * RAND_ADJ.length)];
const n = RAND_NOUN[Math.floor(Math.random() * RAND_NOUN.length)];
return `${a}-${n}-${1000 + Math.floor(Math.random() * 9000)}`;
}
// POST /api/profile/create
// Body: { name?, data? } — empty/absent name asks the server to generate a
// unique random one. Fails with 409 when the requested name is taken.
app.post('/api/profile/create', async (c) => {
let body;
try { body = await c.req.json(); } catch { return c.json({ ok: false, error: 'invalid JSON' }, 400); }
let name = (body.name || '').trim().toLowerCase();
const dataJson = JSON.stringify(body.data && typeof body.data === 'object' ? body.data : {});
if (dataJson.length > PROFILE_MAX_BYTES) return c.json({ ok: false, error: 'profile data too large' }, 413);
try {
if (name) {
if (!PROFILE_NAME_RE.test(name)) {
return c.json({ ok: false, error: 'invalid name — 3-40 characters: letters, digits, - or _' }, 400);
}
if (!(await createProfile(name, dataJson))) {
return c.json({ ok: false, error: `${name}” is already taken — pick another name` }, 409);
}
} else {
let created = false;
for (let tries = 0; tries < 20 && !created; tries++) {
name = randomProfileName();
created = await createProfile(name, dataJson);
}
if (!created) return c.json({ ok: false, error: 'could not generate a unique name — try again' }, 500);
}
const row = await getProfile(name);
return c.json({ ok: true, name, updatedAt: row ? row.updatedAt : 0 });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// GET /api/profile/load?name=<name>
app.get('/api/profile/load', async (c) => {
const name = (c.req.query('name') || '').trim().toLowerCase();
if (!name) return c.json({ ok: false, error: 'missing name' }, 400);
try {
const row = await getProfile(name);
if (!row) return c.json({ ok: false, error: 'profile not found' }, 404);
let data = {};
try { data = JSON.parse(row.data || '{}'); } catch { /* corrupt blob — hand back empty */ }
return c.json({ ok: true, name, data, updatedAt: row.updatedAt });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// POST /api/profile/save
// Body: { name, data } — updates an EXISTING profile only (404 otherwise).
app.post('/api/profile/save', async (c) => {
let body;
try { body = await c.req.json(); } catch { return c.json({ ok: false, error: 'invalid JSON' }, 400); }
const name = (body.name || '').trim().toLowerCase();
if (!name) return c.json({ ok: false, error: 'missing name' }, 400);
const dataJson = JSON.stringify(body.data && typeof body.data === 'object' ? body.data : {});
if (dataJson.length > PROFILE_MAX_BYTES) return c.json({ ok: false, error: 'profile data too large' }, 413);
try {
if (!(await saveProfile(name, dataJson))) {
return c.json({ ok: false, error: 'profile not found' }, 404);
}
const row = await getProfile(name);
return c.json({ ok: true, updatedAt: row ? row.updatedAt : 0 });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// POST /api/user/sync
// Body: { fingerprint, playlists?, recentVideo?, appVersion? }
app.post('/api/user/sync', async (c) => {
let body;
try { body = await c.req.json(); } catch { return c.json({ ok: false, error: 'invalid JSON' }, 400); }
const fp = (body.fingerprint || '').trim();
if (!fp) return c.json({ ok: false, error: 'missing fingerprint' }, 400);
try {
await upsertUser({ fingerprint: fp, appVersion: body.appVersion, playlists: body.playlists });
if (body.recentVideo && body.recentVideo.id) {
await recordVideoAccess(fp, body.recentVideo);
}
return c.json({ ok: true });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// GET /api/user/data?fp=<fingerprint>
app.get('/api/user/data', async (c) => {
const fp = (c.req.query('fp') || '').trim();
if (!fp) return c.json({ ok: false, error: 'missing fingerprint' }, 400);
try {
const userData = await getUserData(fp);
return c.json({ ok: true, ...userData, version: APP_VERSION });
} catch (err) {
return c.json({ ok: false, error: err.message }, 500);
}
});
// ============================================================================
// GET /sw.js — serve the service worker with BUILD_TAG injected
//
// The raw sw.js file contains the placeholder `__BUILD_TAG__` which is
// replaced here with the actual BUILD_TAG string so the SW's cache name
// tracks the deployment automatically — no manual version bump needed.
// Served with no-store cache headers so browsers always re-fetch it and
// pick up the substituted value rather than a browser-cached stale copy.
// ============================================================================
let _swSource = null;
app.get('/sw.js', (c) => {
if (!_swSource) {
try {
_swSource = readFileSync('./public/sw.js', 'utf8');
} catch {
return c.text('Service worker not found', 404);
}
}
// Inject the build tag: replace the whole fallback expression with the
// real value. Matched by REGEX, not an exact string — an exact match broke
// the moment the fallback literal in sw.js was bumped ('v1.0.3' → 'v1.0.4'),
// after which the replacement silently did nothing, the SW version froze,
// and clients never saw another update no matter how many times we deployed.
const src = _swSource.replace(
/typeof __BUILD_TAG__ !== 'undefined' \? __BUILD_TAG__ : '[^']*'/,
JSON.stringify(BUILD_TAG)
);
if (src === _swSource) {
console.error('[sw] BUILD_TAG injection failed — placeholder not found in sw.js');
}
return c.text(src, 200, {
'Content-Type': 'application/javascript; charset=utf-8',
'Cache-Control': 'no-store, no-cache, must-revalidate',
});
});
// ============================================================================
// Static files — serve the frontend/public directory
// Must come AFTER all /api routes so API takes priority
// ============================================================================
app.use('/*', serveStatic({ root: './public' }));
// SPA fallback — return index.html for any unmatched path
app.get('/*', serveStatic({ path: './public/index.html' }));
// ============================================================================
// Boot
// ============================================================================
async function main() {
await initDb();
console.log(`[ytplayer] DB ready`);
console.log(`[ytplayer] Starting on port ${PORT}`);
// Bun.serve is the native Bun HTTP server
Bun.serve({
port: PORT,
fetch: app.fetch,
// Default is 10s, which killed /api/download proxy streams whenever the
// connection went idle mid-transfer. 240s covers slow saves; Bun caps
// this field at 255.
idleTimeout: 240,
});
console.log(`[ytplayer] Listening → http://localhost:${PORT}`);
}
main().catch((err) => {
console.error('[ytplayer] fatal:', err);
process.exit(1);
});