Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
07da89bf07 | ||
|
|
99f6372526 | ||
|
|
0aa5920ebb | ||
|
|
f5e28b0d4e | ||
|
|
e4fcb2bfde | ||
|
|
f28f7d8594 |
@@ -0,0 +1,82 @@
|
||||
# Gitea twin of .github/workflows/docker.yml. Gitea runs only .gitea/workflows/ once this
|
||||
# folder exists, and GitHub never reads it.
|
||||
#
|
||||
# Publishes to IPCR on watch.nsl.sh instead of ghcr.io:
|
||||
# push ipcr-push-85-17-246-67.nip.io/metadec.eth/<repo>:<tag> docker login as metadec.eth with the
|
||||
# IPCR_PUSH_TOKEN secret (the metadec.eth namespace token from the IPCR dashboard)
|
||||
# pull ipcr-watch.nsl.sh/ipns/metadec.eth/metadec.eth/<repo>:<tag> anonymous
|
||||
# The push goes to ipcr-push-85-17-246-67.nip.io, the box's own address, not ipcr-push-watch.nsl.sh: Cloudflare fronts
|
||||
# *.watch.nsl.sh and refuses any request over 100 MB (413), which a large image layer is.
|
||||
# That host's certificate is signed by the box's private Mesh Router CA, given to buildkit from
|
||||
# the IPCR_CA_PEM repo variable. docker login would make the job's Docker daemon check the
|
||||
# certificate without that CA, so the login is written straight into ~/.docker/config.json.
|
||||
# The runner's job image (node:22-bookworm) has no docker CLI, so the static one is
|
||||
# installed first; buildx installs itself. No `type=gha` cache: the runner has none.
|
||||
|
||||
name: Build and Push Docker Image
|
||||
|
||||
on:
|
||||
push:
|
||||
branches: [main, master]
|
||||
tags: ['v*']
|
||||
pull_request:
|
||||
branches: [main, master]
|
||||
|
||||
env:
|
||||
REGISTRY: ipcr-push-85-17-246-67.nip.io
|
||||
IMAGE_NAME: metadec.eth/${{ github.event.repository.name }}
|
||||
|
||||
jobs:
|
||||
build:
|
||||
runs-on: ubuntu-latest
|
||||
permissions:
|
||||
contents: read
|
||||
packages: write
|
||||
|
||||
steps:
|
||||
- name: Checkout repository
|
||||
uses: actions/checkout@v4
|
||||
|
||||
- name: Install Docker CLI
|
||||
run: curl -fsSL https://download.docker.com/linux/static/stable/x86_64/docker-28.5.1.tgz | tar -xz -C /usr/local/bin --strip-components=1 docker/docker
|
||||
|
||||
- name: Trust the Mesh Router CA
|
||||
run: printf '%s\n' "$IPCR_CA_PEM" > /tmp/ipcr-ca.pem
|
||||
env:
|
||||
IPCR_CA_PEM: ${{ vars.IPCR_CA_PEM }}
|
||||
|
||||
- name: Set up Docker Buildx
|
||||
uses: docker/setup-buildx-action@v3
|
||||
with:
|
||||
version: latest
|
||||
buildkitd-config-inline: |
|
||||
[registry."${{ env.REGISTRY }}"]
|
||||
ca = ["/tmp/ipcr-ca.pem"]
|
||||
|
||||
- name: Log in to IPCR
|
||||
if: github.event_name != 'pull_request'
|
||||
env:
|
||||
IPCR_PUSH_TOKEN: ${{ secrets.IPCR_PUSH_TOKEN }}
|
||||
run: |
|
||||
mkdir -p ~/.docker
|
||||
printf '{"auths":{"%s":{"auth":"%s"}}}\n' "$REGISTRY" "$(printf 'metadec.eth:%s' "$IPCR_PUSH_TOKEN" | base64 -w0)" > ~/.docker/config.json
|
||||
|
||||
- name: Extract metadata
|
||||
id: meta
|
||||
uses: docker/metadata-action@v5
|
||||
with:
|
||||
images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
|
||||
tags: |
|
||||
type=ref,event=branch
|
||||
type=ref,event=pr
|
||||
type=semver,pattern={{version}}
|
||||
type=semver,pattern={{major}}.{{minor}}
|
||||
type=sha
|
||||
|
||||
- name: Build and push
|
||||
uses: docker/build-push-action@v5
|
||||
with:
|
||||
context: .
|
||||
push: ${{ github.event_name != 'pull_request' }}
|
||||
tags: ${{ steps.meta.outputs.tags }}
|
||||
labels: ${{ steps.meta.outputs.labels }}
|
||||
@@ -8,7 +8,8 @@ This plugin enriches video metadata by querying the TMDB API:
|
||||
|
||||
- **Search by IMDB ID**: Uses existing IMDB ID for precise matching
|
||||
- **Search by title**: Falls back to title/year search
|
||||
- **Image download**: Downloads poster and backdrop images to `/files/poster/`
|
||||
- **Image download**: Downloads poster and backdrop images to `/files/plugin/tmdb/`
|
||||
- **Alternative posters**: Lists up to 10 posters (the primary plus the best-voted alternatives, as `url` locator CIDs — nothing extra is downloaded)
|
||||
- **Rich metadata**: Fetches plot, rating, genres, studios
|
||||
|
||||
**Note**: This runs on the background queue as it makes external API calls.
|
||||
@@ -26,9 +27,8 @@ This plugin enriches video metadata by querying the TMDB API:
|
||||
| `rating` | Vote average |
|
||||
| `plot/eng` | English plot synopsis |
|
||||
| `poster` | Poster image CID |
|
||||
| `posterPath` | Poster file path |
|
||||
| `posters/{lang3}/{cid}` | Alternative posters key-set, `poster` among them (METADATA_KEYS.md §6) |
|
||||
| `backdrop` | Backdrop image CID |
|
||||
| `backdropPath` | Backdrop file path |
|
||||
| `genres` | Genre set |
|
||||
| `studio` | Production company set |
|
||||
|
||||
|
||||
+2
-2
@@ -5,7 +5,7 @@
|
||||
"description": "MetaMesh plugin: Fetches metadata from TMDB API",
|
||||
"type": "module",
|
||||
"main": "dist/index.js",
|
||||
"scripts": { "build": "tsc", "start": "node dist/index.js" },
|
||||
"scripts": { "build": "tsc", "start": "node dist/index.js", "test": "vitest run", "test:watch": "vitest" },
|
||||
"dependencies": { "fastify": "^5.0.0", "axios": "^1.6.0" },
|
||||
"devDependencies": { "@types/node": "^20.0.0", "typescript": "^5.3.0" }
|
||||
"devDependencies": { "@types/node": "^20.0.0", "typescript": "^5.3.0", "vitest": "^1.6.1" }
|
||||
}
|
||||
|
||||
Generated
+1088
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,188 @@
|
||||
/**
|
||||
* Beacon v2 advertiser — makes this plugin findable by meta-sort's Plugins
|
||||
* page (Scan → Add) without anyone typing its URL.
|
||||
*
|
||||
* Advertises one resource, cap `metamesh.enrich/<manifest id>`, on the beacon
|
||||
* group (239.255.99.1:9099): unsolicited every 10 s, and unicast in reply to
|
||||
* any probe whose `want` matches. Says `bye` on shutdown. It keeps no view of
|
||||
* other nodes — a plugin only needs to be found. Spec:
|
||||
* meta-root docs/project-architecture/beacon-v2.md.
|
||||
*
|
||||
* Env:
|
||||
* ENABLE_UDP_DISCOVERY false/0/no/off disables
|
||||
* BEACON_ADVERTISE_URL endpoints.http (default http://$HOSTNAME:$PORT)
|
||||
* BEACON_BINDS the one meta-sort instance allowed to list it
|
||||
* (meta-sort sets this on containers it spawns)
|
||||
* BEACON_GROUP / BEACON_PORT / BEACON_INTERVAL_MS
|
||||
*
|
||||
* ⚠ MIRRORED FILE. Byte-identical copies live in every TypeScript
|
||||
* metamesh-plugin-*; see meta-root `scripts/check-mirrors.sh`. Never fatal: a
|
||||
* host that blocks multicast logs a warning and the plugin serves on.
|
||||
*/
|
||||
|
||||
import dgram from 'dgram';
|
||||
import { hostname, networkInterfaces } from 'os';
|
||||
|
||||
const PROTO = 'beacon';
|
||||
const VERSION = 2;
|
||||
|
||||
interface Resource {
|
||||
id: string;
|
||||
caps: string[];
|
||||
endpoints: Record<string, string>;
|
||||
binds?: string;
|
||||
}
|
||||
|
||||
interface NodeInfo {
|
||||
name: string;
|
||||
instance: string;
|
||||
version?: string;
|
||||
}
|
||||
|
||||
function envNum(key: string): number | undefined {
|
||||
const n = Number(process.env[key]);
|
||||
return Number.isFinite(n) && n > 0 ? n : undefined;
|
||||
}
|
||||
|
||||
function enabled(): boolean {
|
||||
const v = (process.env.ENABLE_UDP_DISCOVERY ?? '').trim().toLowerCase();
|
||||
return !['false', '0', 'no', 'off'].includes(v);
|
||||
}
|
||||
|
||||
/** IPv4 addresses of every up, non-internal interface. */
|
||||
function interfaceAddresses(): string[] {
|
||||
const out: string[] = [];
|
||||
for (const list of Object.values(networkInterfaces())) {
|
||||
for (const ni of list ?? []) {
|
||||
const v4 = (ni as { family: string | number }).family === 'IPv4' || (ni as { family: string | number }).family === 4;
|
||||
if (v4 && !ni.internal && ni.address) out.push(ni.address);
|
||||
}
|
||||
}
|
||||
return out;
|
||||
}
|
||||
|
||||
/** Same matching rules as the spec: `*`, `x/*`, optional `@contract`. */
|
||||
function capMatches(pattern: string, cap: string): boolean {
|
||||
if (pattern === '*') return true;
|
||||
const split = (s: string): [string, string | null] => {
|
||||
const i = s.lastIndexOf('@');
|
||||
return i >= 0 ? [s.slice(0, i), s.slice(i + 1)] : [s, null];
|
||||
};
|
||||
const [pb, pc] = split(pattern);
|
||||
const [cb, cc] = split(cap);
|
||||
if (pc !== null && cc !== pc) return false;
|
||||
if (pb.endsWith('/*')) {
|
||||
const prefix = pb.slice(0, -2);
|
||||
const rest = cb.startsWith(prefix) ? cb.slice(prefix.length) : '';
|
||||
return rest.startsWith('/') && rest.length > 1;
|
||||
}
|
||||
return pb === cb;
|
||||
}
|
||||
|
||||
let socket: dgram.Socket | null = null;
|
||||
let timer: NodeJS.Timeout | null = null;
|
||||
let node: NodeInfo | null = null;
|
||||
let resource: Resource | null = null;
|
||||
let ifaces: string[] = [];
|
||||
let group = '239.255.99.1';
|
||||
let udpPort = 9099;
|
||||
|
||||
function send(msg: object, dest?: { address: string; port: number }): void {
|
||||
const s = socket;
|
||||
if (!s) return;
|
||||
const body = Buffer.from(JSON.stringify(msg));
|
||||
if (dest) {
|
||||
s.send(body, dest.port, dest.address, () => {});
|
||||
return;
|
||||
}
|
||||
// Once per interface: the default route alone reaches one network.
|
||||
for (const addr of ifaces) {
|
||||
try {
|
||||
s.setMulticastInterface(addr);
|
||||
s.send(body, udpPort, group, () => {});
|
||||
} catch (e) {
|
||||
console.warn(`[beacon] send on ${addr} failed:`, e);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
function advertise(dest?: { address: string; port: number }): void {
|
||||
if (!node || !resource) return;
|
||||
send({ proto: PROTO, v: VERSION, type: 'advertise', node, resources: [resource] }, dest);
|
||||
}
|
||||
|
||||
/**
|
||||
* Start advertising this plugin. Call once the HTTP server is listening, so a
|
||||
* consumer that hears it can reach it. Idempotent.
|
||||
*/
|
||||
export function startBeacon(pluginId: string, version: string, httpPort: number): void {
|
||||
if (socket || !enabled()) return;
|
||||
group = process.env.BEACON_GROUP || group;
|
||||
udpPort = envNum('BEACON_PORT') ?? udpPort;
|
||||
const intervalMs = envNum('BEACON_INTERVAL_MS') ?? 10_000;
|
||||
const instance = process.env.HOSTNAME || hostname();
|
||||
const url = (process.env.BEACON_ADVERTISE_URL || `http://${instance}:${httpPort}`).replace(/\/+$/, '');
|
||||
const binds = (process.env.BEACON_BINDS || '').trim();
|
||||
|
||||
node = { name: `metamesh-plugin-${pluginId}`, instance, ...(version ? { version } : {}) };
|
||||
resource = {
|
||||
id: pluginId,
|
||||
caps: [`metamesh.enrich/${pluginId}`],
|
||||
endpoints: { http: url, manifest: '/manifest' },
|
||||
...(binds ? { binds } : {}),
|
||||
};
|
||||
|
||||
const s = dgram.createSocket({ type: 'udp4', reuseAddr: true });
|
||||
socket = s;
|
||||
s.on('error', (err) => console.warn('[beacon] socket error:', err.message));
|
||||
s.on('message', (buf, rinfo) => {
|
||||
let m: { proto?: string; v?: number; type?: string; from?: string; want?: string[] };
|
||||
try {
|
||||
m = JSON.parse(buf.toString('utf8'));
|
||||
} catch {
|
||||
return; // foreign traffic on the shared group
|
||||
}
|
||||
if (m?.proto !== PROTO || m.v !== VERSION || m.type !== 'probe' || m.from === instance) return;
|
||||
const want = Array.isArray(m.want) ? m.want : [];
|
||||
if (want.length === 0 || resource!.caps.some((c) => want.some((p) => capMatches(p, c)))) {
|
||||
advertise({ address: rinfo.address, port: rinfo.port });
|
||||
}
|
||||
});
|
||||
s.bind(udpPort, '0.0.0.0', () => {
|
||||
try {
|
||||
s.setMulticastTTL(1);
|
||||
} catch {
|
||||
/* best effort */
|
||||
}
|
||||
ifaces = interfaceAddresses();
|
||||
for (const addr of ifaces) {
|
||||
try {
|
||||
s.addMembership(group, addr);
|
||||
} catch (e) {
|
||||
console.warn(`[beacon] join ${group} on ${addr} failed:`, e);
|
||||
}
|
||||
}
|
||||
console.log(`[beacon] advertising metamesh.enrich/${pluginId} at ${url} on ${group}:${udpPort}`);
|
||||
advertise();
|
||||
timer = setInterval(() => advertise(), intervalMs);
|
||||
timer.unref?.();
|
||||
});
|
||||
}
|
||||
|
||||
/** Say `bye` so consumers drop this plugin at once. */
|
||||
export async function stopBeacon(): Promise<void> {
|
||||
if (!socket || !node) return;
|
||||
if (timer) clearInterval(timer);
|
||||
send({ proto: PROTO, v: VERSION, type: 'bye', node });
|
||||
// Let the bye datagrams leave before the socket closes under them.
|
||||
await new Promise((r) => setTimeout(r, 50));
|
||||
const s = socket;
|
||||
socket = null;
|
||||
await new Promise<void>((resolve) => {
|
||||
try {
|
||||
s.close(() => resolve());
|
||||
} catch {
|
||||
resolve();
|
||||
}
|
||||
});
|
||||
}
|
||||
+58
@@ -0,0 +1,58 @@
|
||||
/**
|
||||
* CID encoding helpers (METADATA_KEYS.md §2).
|
||||
*/
|
||||
|
||||
const BASE32_ALPHABET = 'abcdefghijklmnopqrstuvwxyz234567';
|
||||
|
||||
/** RFC 4648 base32, lowercase, no padding, with the `b` multibase prefix. */
|
||||
export function base32LowerMultibase(bytes: Uint8Array): string {
|
||||
let out = 'b';
|
||||
let bits = 0;
|
||||
let value = 0;
|
||||
for (const byte of bytes) {
|
||||
value = (value << 8) | byte;
|
||||
bits += 8;
|
||||
while (bits >= 5) {
|
||||
bits -= 5;
|
||||
out += BASE32_ALPHABET[(value >> bits) & 0x1f];
|
||||
}
|
||||
}
|
||||
if (bits > 0) {
|
||||
out += BASE32_ALPHABET[(value << (5 - bits)) & 0x1f];
|
||||
}
|
||||
return out;
|
||||
}
|
||||
|
||||
/** Unsigned LEB128 varint, as multiformats encodes codecs and lengths. */
|
||||
function varint(value: number): number[] {
|
||||
const out: number[] = [];
|
||||
let n = value;
|
||||
while (n >= 0x80) {
|
||||
out.push((n & 0x7f) | 0x80);
|
||||
n = Math.floor(n / 0x80);
|
||||
}
|
||||
out.push(n);
|
||||
return out;
|
||||
}
|
||||
|
||||
/** Custom multicodec of the `url` locator (METADATA_KEYS.md §2.1). */
|
||||
export const URL_LOCATOR_CODEC = 0x1006;
|
||||
|
||||
/**
|
||||
* Encode an absolute http(s) URL as a `0x1006` identity-multihash CIDv1:
|
||||
* `[0x01][varint codec][0x00][varint len][utf-8 url]`, base32lower. The digest
|
||||
* is the URL itself, so minting one costs no I/O. `null` for anything that is
|
||||
* not an absolute http(s) URL.
|
||||
*
|
||||
* CROSS-BINARY CONTRACT: byte-identical to meta-feeder-sdk's `compute_url_cid`
|
||||
* (`hash.rs`); both test suites pin the same vectors.
|
||||
*/
|
||||
export function encodeUrlCid(url: string): string | null {
|
||||
const trimmed = url.trim();
|
||||
if (!trimmed.startsWith('http://') && !trimmed.startsWith('https://')) {
|
||||
return null;
|
||||
}
|
||||
const digest = Buffer.from(trimmed, 'utf8');
|
||||
const header = Buffer.from([0x01, ...varint(URL_LOCATOR_CODEC), 0x00, ...varint(digest.length)]);
|
||||
return base32LowerMultibase(Buffer.concat([header, digest]));
|
||||
}
|
||||
+3
-2
@@ -27,6 +27,7 @@
|
||||
import Fastify from 'fastify';
|
||||
import type { HealthResponse, ProcessRequest, ProcessResponse, CallbackPayload, ConfigureRequest, ConfigureResponse } from './types.js';
|
||||
import { manifest, process as processFile, configure } from './plugin.js';
|
||||
import { startBeacon, stopBeacon } from './beacon-advertise.js';
|
||||
|
||||
const app = Fastify({ logger: true });
|
||||
let ready = false;
|
||||
@@ -64,5 +65,5 @@ app.post<{ Body: ProcessRequest }>('/process', async (request, reply) => {
|
||||
});
|
||||
|
||||
const port = parseInt(process.env.PORT || '8080', 10);
|
||||
app.listen({ port, host: '0.0.0.0' }).then(() => { ready = true; console.log(`[${manifest.id}] Listening on port ${port}`); });
|
||||
process.on('SIGTERM', async () => { ready = false; await app.close(); process.exit(0); });
|
||||
app.listen({ port, host: '0.0.0.0' }).then(() => { startBeacon(manifest.id, manifest.version, port); ready = true; console.log(`[${manifest.id}] Listening on port ${port}`); });
|
||||
process.on('SIGTERM', async () => { await stopBeacon(); ready = false; await app.close(); process.exit(0); });
|
||||
|
||||
+98
-25
@@ -25,10 +25,12 @@
|
||||
*
|
||||
* Matches old TMDBProcessor output:
|
||||
* - tmdbid, imdbid
|
||||
* - originalTitle, movieYear, releasedate
|
||||
* - title, originalTitle, movieYear, releasedate
|
||||
* - titles/<lang3>/<name> key-set: localized + original title + every AKA
|
||||
* - plot/eng, rating
|
||||
* - genres (add), studio (add), tags (add)
|
||||
* - poster, backdrop (CID hashes of downloaded images)
|
||||
* - posters/{lang3}/{cid} (alternative posters as url locators, `poster` among them)
|
||||
*/
|
||||
|
||||
import axios from 'axios';
|
||||
@@ -38,6 +40,8 @@ import type { PluginManifest, ProcessRequest, CallbackPayload } from './types.js
|
||||
import { MetaCoreClient } from './meta-core-client.js';
|
||||
import { readJson, writeJson } from './cache.js';
|
||||
import { getWebDAVClient, WebDAVClient } from './webdav-client.js';
|
||||
import { base32LowerMultibase } from './cid.js';
|
||||
import { posterMembers } from './posters.js';
|
||||
|
||||
const IMAGE_BASE_URL = 'https://image.tmdb.org/t/p/original';
|
||||
|
||||
@@ -104,23 +108,7 @@ function computeMidHash256FromData(fileSize: number, sampleData: Buffer): string
|
||||
]);
|
||||
|
||||
// Encode as base32lower with 'b' prefix (multibase)
|
||||
const base32Chars = 'abcdefghijklmnopqrstuvwxyz234567';
|
||||
let cid = 'b';
|
||||
let bits = 0;
|
||||
let value = 0;
|
||||
for (const byte of cidBytes) {
|
||||
value = (value << 8) | byte;
|
||||
bits += 8;
|
||||
while (bits >= 5) {
|
||||
bits -= 5;
|
||||
cid += base32Chars[(value >> bits) & 0x1f];
|
||||
}
|
||||
}
|
||||
if (bits > 0) {
|
||||
cid += base32Chars[(value << (5 - bits)) & 0x1f];
|
||||
}
|
||||
|
||||
return cid;
|
||||
return base32LowerMultibase(cidBytes);
|
||||
}
|
||||
|
||||
export const manifest: PluginManifest = {
|
||||
@@ -168,6 +156,51 @@ let forceRecompute = false;
|
||||
|
||||
const BASE_URL = 'https://api.themoviedb.org/3';
|
||||
|
||||
// ISO 639-1 → ISO 639-3 (`lang3`, the 639-2/T form where B/T differ) for the
|
||||
// languages TMDB commonly returns. Unknown codes map to nothing.
|
||||
const ISO639_1_TO_3: Record<string, string> = {
|
||||
en: 'eng', ja: 'jpn', fr: 'fra', de: 'deu', es: 'spa', it: 'ita', ru: 'rus',
|
||||
ko: 'kor', zh: 'zho', pt: 'por', nl: 'nld', sv: 'swe', no: 'nor', da: 'dan',
|
||||
fi: 'fin', pl: 'pol', tr: 'tur', cs: 'ces', hu: 'hun', el: 'ell', he: 'heb',
|
||||
th: 'tha', vi: 'vie', id: 'ind', uk: 'ukr', ro: 'ron', ar: 'ara', fa: 'fas',
|
||||
hi: 'hin', ta: 'tam', te: 'tel', ms: 'msa', tl: 'tgl', ca: 'cat', bg: 'bul',
|
||||
hr: 'hrv', sr: 'srp', sk: 'slk', sl: 'slv', et: 'est', lv: 'lav', lt: 'lit',
|
||||
is: 'isl', ga: 'gle', cn: 'zho',
|
||||
};
|
||||
|
||||
function iso639_1to3(code: string | undefined | null): string | undefined {
|
||||
return code ? ISO639_1_TO_3[code.trim().toLowerCase()] : undefined;
|
||||
}
|
||||
|
||||
// TMDB keys AKAs by MARKET (`iso_3166_1`), not language. Only markets that
|
||||
// clearly imply one language are mapped; a multilingual or unknown market
|
||||
// (CA, CH, BE, IN, HK, SG, …, or empty) files its AKA under `und` rather than
|
||||
// guessing (METADATA_KEYS.md §3).
|
||||
const MARKET_TO_LANG3: Record<string, string> = {
|
||||
US: 'eng', GB: 'eng', AU: 'eng', NZ: 'eng', IE: 'eng',
|
||||
FR: 'fra', DE: 'deu', AT: 'deu', ES: 'spa', MX: 'spa', AR: 'spa', CO: 'spa',
|
||||
CL: 'spa', PE: 'spa', VE: 'spa', IT: 'ita', BR: 'por', PT: 'por', JP: 'jpn',
|
||||
KR: 'kor', CN: 'zho', TW: 'zho', RU: 'rus', NL: 'nld', SE: 'swe', NO: 'nor',
|
||||
DK: 'dan', FI: 'fin', PL: 'pol', TR: 'tur', CZ: 'ces', HU: 'hun', GR: 'ell',
|
||||
IL: 'heb', TH: 'tha', VN: 'vie', ID: 'ind', UA: 'ukr', RO: 'ron', IR: 'fas',
|
||||
SA: 'ara', EG: 'ara', AE: 'ara', BG: 'bul', HR: 'hrv', RS: 'srp', SK: 'slk',
|
||||
SI: 'slv', EE: 'est', LV: 'lav', LT: 'lit', IS: 'isl',
|
||||
};
|
||||
|
||||
function marketToLang3(market: string | undefined | null): string {
|
||||
return (market && MARKET_TO_LANG3[market.trim().toUpperCase()]) || 'und';
|
||||
}
|
||||
|
||||
/**
|
||||
* `titles/<lang3>/<name>` key-set member key (METADATA_KEYS.md §3): trimmed,
|
||||
* whitespace collapsed, `/` (the key-set separator) written as U+2215 `∕`.
|
||||
* `undefined` when nothing is left to name.
|
||||
*/
|
||||
function titleMemberKey(lang3: string, name: string | undefined | null): string | undefined {
|
||||
const clean = (name ?? '').trim().replace(/\s+/g, ' ').replace(/\//g, '\u2215');
|
||||
return clean ? `titles/${lang3}/${clean}` : undefined;
|
||||
}
|
||||
|
||||
let apiKey: string | null = null;
|
||||
let isV4Token = false;
|
||||
let metadataLanguage = 'en-US';
|
||||
@@ -218,7 +251,17 @@ async function findByImdbId(imdbId: string): Promise<any> {
|
||||
|
||||
async function getByTmdbId(tmdbId: string, mediaType: string): Promise<any> {
|
||||
try {
|
||||
const config = getAxiosConfig({ language: metadataLanguage });
|
||||
// `alternative_titles` rides along so the AKAs land in `titles/*` (§3).
|
||||
// `images` rides the same call too: the poster candidates for the
|
||||
// `posters/{lang3}/{cid}` key-set (§6). `language` would otherwise filter
|
||||
// them to one language, so widen it to the configured one, English and
|
||||
// textless (`null`).
|
||||
const lang1 = metadataLanguage.split('-')[0];
|
||||
const config = getAxiosConfig({
|
||||
language: metadataLanguage,
|
||||
append_to_response: 'alternative_titles,images',
|
||||
include_image_language: [...new Set([lang1, 'en'])].join(',') + ',null',
|
||||
});
|
||||
const response = await axios.get(`${BASE_URL}/${mediaType}/${tmdbId}`, config);
|
||||
return response.data;
|
||||
} catch {
|
||||
@@ -406,7 +449,7 @@ export async function process(
|
||||
// entry now, but tolerates the legacy raw-payload format from older builds.
|
||||
const midhash = existingMeta?.['cid_midhash256'];
|
||||
if (midhash && !forceRecompute) {
|
||||
const cached = await readJson<any>(`${midhash}_tmdb.json`);
|
||||
const cached = await readJson<any>(`${midhash}_aka_tmdb.json`);
|
||||
if (cached) {
|
||||
console.log(`[tmdb] Using byte-hash cached TMDB data for ${label}`);
|
||||
const isEntry = cached.data !== undefined;
|
||||
@@ -485,7 +528,7 @@ export async function process(
|
||||
// Keep the exact-file cache warm too, so a future re-enrichment of
|
||||
// this same file skips straight to the byte-hash hit.
|
||||
if (midhash) {
|
||||
await writeJson(`${midhash}_tmdb.json`, entry);
|
||||
await writeJson(`${midhash}_aka_tmdb.json`, entry);
|
||||
}
|
||||
console.log(`[tmdb] Enriched ${label} (tmdbid=${entry.data.id})`);
|
||||
} else {
|
||||
@@ -525,7 +568,9 @@ const tmdbEntryInflight = new Map<string, Promise<TmdbCacheEntry | null>>();
|
||||
// Persistent cache filename for a resolved show/movie. Includes mediaType +
|
||||
// language so a tv/movie id clash or a language switch can't return stale text.
|
||||
function tmdbCacheKey(tmdbId: string, mediaType: string): string {
|
||||
return `tmdbid_${mediaType}_${tmdbId}_${metadataLanguage}_tmdb.json`;
|
||||
// `_aka`: entries written before `alternative_titles` was appended carry no
|
||||
// AKAs, so the name moved to make them refetch once.
|
||||
return `tmdbid_${mediaType}_${tmdbId}_${metadataLanguage}_aka_tmdb.json`;
|
||||
}
|
||||
|
||||
// Media type from a raw TMDB payload's own shape (tv payloads carry
|
||||
@@ -632,6 +677,10 @@ async function applyTmdbData(
|
||||
data.original_name !== undefined;
|
||||
metadata.videoType = isTv ? 'tvshow' : 'movie';
|
||||
metadata.contentKind = isTv ? 'episode' : 'movie';
|
||||
// `domain` is written in the same breath as `contentKind`
|
||||
// (METADATA_KEYS.md §1) — it is the key meta-watch filters its wall on,
|
||||
// and TMDB is precisely the identity graph that decides film vs tv here.
|
||||
metadata.domain = isTv ? 'tv' : 'film';
|
||||
|
||||
// Title - store both original and localized
|
||||
const originalTitle = data.original_title || data.original_name;
|
||||
@@ -648,6 +697,23 @@ async function applyTmdbData(
|
||||
metadata.originalTitle = originalTitle;
|
||||
}
|
||||
|
||||
// Every clean name as a `titles/<lang3>/<name>` key-set member
|
||||
// (METADATA_KEYS.md §3): the localized title under the configured metadata
|
||||
// language, the original title under TMDB's `original_language`, and each
|
||||
// AKA under the language its market implies (`und` when it doesn't).
|
||||
const addTitleMember = (lang3: string, name: string | undefined) => {
|
||||
const key = titleMemberKey(lang3, name);
|
||||
if (key) metadata[key] = 'true';
|
||||
};
|
||||
addTitleMember(iso639_1to3(metadataLanguage.split('-')[0]) ?? 'und', localizedTitle);
|
||||
addTitleMember(iso639_1to3(data.original_language) ?? 'und', originalTitle);
|
||||
const akas = data.alternative_titles?.titles ?? data.alternative_titles?.results;
|
||||
if (Array.isArray(akas)) {
|
||||
for (const aka of akas) {
|
||||
addTitleMember(marketToLang3(aka?.iso_3166_1), aka?.title);
|
||||
}
|
||||
}
|
||||
|
||||
// Dates
|
||||
const releaseDate = data.release_date || data.first_air_date;
|
||||
let year: string | undefined;
|
||||
@@ -662,9 +728,8 @@ async function applyTmdbData(
|
||||
// Plot (in configured language)
|
||||
if (data.overview) {
|
||||
// Extract language code from metadataLanguage (e.g., 'en-US' -> 'eng')
|
||||
const langCode = metadataLanguage.split('-')[0];
|
||||
const langKey = langCode === 'en' ? 'eng' : langCode;
|
||||
metadata[`plot/${langKey}`] = data.overview;
|
||||
const langKey = iso639_1to3(metadataLanguage.split('-')[0]);
|
||||
if (langKey) metadata[`plot/${langKey}`] = data.overview;
|
||||
// Also set as primary plot for compatibility
|
||||
metadata['plot/eng'] = data.overview;
|
||||
}
|
||||
@@ -706,6 +771,14 @@ async function applyTmdbData(
|
||||
if (posterCid) {
|
||||
await metaCore.setProperty(cid, 'poster', posterCid);
|
||||
console.log(`[tmdb] Set poster CID: ${posterCid}`);
|
||||
// The alternatives, as the `posters/{lang3}/{cid}` key-set with `poster`
|
||||
// among them (METADATA_KEYS.md §6). Leaf keys through PATCH — never
|
||||
// addToSet, which writes the legacy comma-joined csv-set. Empty for a
|
||||
// cached payload from before the `images` append: no backfill by design.
|
||||
const members = posterMembers(data, posterCid);
|
||||
if (Object.keys(members).length > 0) {
|
||||
await metaCore.mergeMetadata(cid, members);
|
||||
}
|
||||
}
|
||||
|
||||
let backdropCid = images ? images.backdropCid : null;
|
||||
|
||||
+102
@@ -0,0 +1,102 @@
|
||||
/**
|
||||
* The `posters/{lang3}/{cid}` key-set for a TMDB-enriched video record
|
||||
* (METADATA_KEYS.md §6): the primary poster plus the best-voted alternatives.
|
||||
*
|
||||
* CROSS-BINARY CONTRACT: same ranking, cap, CDN size and language rule as the
|
||||
* tmdb card feeder's `poster_member_keys`
|
||||
* (`meta-feeder-tmdb/feeder-plugin/card-feeder/src/card.rs`).
|
||||
*/
|
||||
|
||||
import { encodeUrlCid } from './cid.js';
|
||||
|
||||
/** Most members a record's `posters/*` key-set may carry, the primary included. */
|
||||
export const MAX_POSTER_MEMBERS = 10;
|
||||
|
||||
/**
|
||||
* CDN root + size the alternatives are referenced at. `w500`, as the card
|
||||
* feeder's `poster_url`, so both writers mint the same locator for one image.
|
||||
*/
|
||||
export const ALT_POSTER_BASE_URL = 'https://image.tmdb.org/t/p/w500';
|
||||
|
||||
/** One entry of a TMDB details payload's `images.posters` (`append_to_response=images`). */
|
||||
export interface TmdbImage {
|
||||
file_path?: string;
|
||||
iso_639_1?: string | null;
|
||||
vote_average?: number;
|
||||
vote_count?: number;
|
||||
}
|
||||
|
||||
/**
|
||||
* ISO 639-1 → ISO 639-3 (`lang3`), the 639-2/T form where B and T differ
|
||||
* (`fra`, `deu`, `zho` — never `fre`/`ger`/`chi`). Mirrors the card feeder's
|
||||
* `iso639_1_to_3` table.
|
||||
*/
|
||||
const ISO639_1_TO_3: Record<string, string> = {
|
||||
en: 'eng', ja: 'jpn', fr: 'fra', de: 'deu', es: 'spa', it: 'ita', ru: 'rus',
|
||||
ko: 'kor', zh: 'zho', pt: 'por', nl: 'nld', sv: 'swe', no: 'nor', da: 'dan',
|
||||
fi: 'fin', pl: 'pol', tr: 'tur', ar: 'ara', hi: 'hin', th: 'tha', vi: 'vie',
|
||||
id: 'ind', cs: 'ces', el: 'ell', he: 'heb', hu: 'hun', ro: 'ron', uk: 'ukr',
|
||||
fa: 'fas', ms: 'msa', tl: 'tgl', ca: 'cat', nb: 'nob',
|
||||
};
|
||||
|
||||
/**
|
||||
* The language segment for a poster: that of the text printed on it. `zxx`
|
||||
* (no linguistic content) when TMDB gives none — a textless poster — and
|
||||
* `und` for a code outside the table.
|
||||
*/
|
||||
export function posterLang3(code: string | null | undefined): string {
|
||||
const trimmed = (code ?? '').trim().toLowerCase();
|
||||
if (!trimmed) return 'zxx';
|
||||
return ISO639_1_TO_3[trimmed] ?? 'und';
|
||||
}
|
||||
|
||||
function compareCandidates(a: TmdbImage, b: TmdbImage): number {
|
||||
const byAverage = (b.vote_average ?? 0) - (a.vote_average ?? 0);
|
||||
if (byAverage !== 0) return byAverage;
|
||||
const byCount = (b.vote_count ?? 0) - (a.vote_count ?? 0);
|
||||
if (byCount !== 0) return byCount;
|
||||
const pa = a.file_path ?? '';
|
||||
const pb = b.file_path ?? '';
|
||||
return pa < pb ? -1 : pa > pb ? 1 : 0;
|
||||
}
|
||||
|
||||
/**
|
||||
* The `posters/{lang3}/{cid}` members to merge onto a record whose `poster` is
|
||||
* `primaryPosterCid`. The primary is always a member, filed under the language
|
||||
* of its own `images.posters` entry (`und` if TMDB did not list it); the
|
||||
* alternatives follow as `url` locators, ranked by `vote_average`, then
|
||||
* `vote_count`, then `file_path` (so equal payloads give equal keys), capped at
|
||||
* {@link MAX_POSTER_MEMBERS} in total.
|
||||
*
|
||||
* Returns `{}` when the payload has no image list — a cached entry from before
|
||||
* the `images` append. Without one the primary's language is unknown, and a set
|
||||
* is better absent than mislabelled.
|
||||
*/
|
||||
export function posterMembers(data: any, primaryPosterCid: string): Record<string, 'true'> {
|
||||
const candidates: TmdbImage[] = Array.isArray(data?.images?.posters) ? data.images.posters : [];
|
||||
if (candidates.length === 0 || !primaryPosterCid) {
|
||||
return {};
|
||||
}
|
||||
const primaryPath: string | undefined = data?.poster_path;
|
||||
const members: Record<string, 'true'> = {};
|
||||
|
||||
const primary = candidates.find((img) => !!img.file_path && img.file_path === primaryPath);
|
||||
const primaryLang = primary ? posterLang3(primary.iso_639_1) : 'und';
|
||||
members[`posters/${primaryLang}/${primaryPosterCid}`] = 'true';
|
||||
let count = 1;
|
||||
|
||||
const ranked = candidates
|
||||
.filter((img) => typeof img.file_path === 'string' && img.file_path.trim() !== '' && img.file_path !== primaryPath)
|
||||
.sort(compareCandidates);
|
||||
for (const img of ranked) {
|
||||
if (count >= MAX_POSTER_MEMBERS) break;
|
||||
const cid = encodeUrlCid(`${ALT_POSTER_BASE_URL}/${img.file_path!.replace(/^\/+/, '')}`);
|
||||
if (!cid) continue;
|
||||
const key = `posters/${posterLang3(img.iso_639_1)}/${cid}`;
|
||||
if (!(key in members)) {
|
||||
members[key] = 'true';
|
||||
count++;
|
||||
}
|
||||
}
|
||||
return members;
|
||||
}
|
||||
@@ -0,0 +1,88 @@
|
||||
import { describe, expect, it } from 'vitest';
|
||||
import { encodeUrlCid } from '../src/cid.js';
|
||||
import { ALT_POSTER_BASE_URL, MAX_POSTER_MEMBERS, posterLang3, posterMembers, type TmdbImage } from '../src/posters.js';
|
||||
|
||||
const alt = (path: string) => encodeUrlCid(`${ALT_POSTER_BASE_URL}${path}`)!;
|
||||
const img = (file_path: string, iso_639_1: string | null, vote_average = 0, vote_count = 0): TmdbImage => ({
|
||||
file_path,
|
||||
iso_639_1,
|
||||
vote_average,
|
||||
vote_count,
|
||||
});
|
||||
|
||||
describe('encodeUrlCid', () => {
|
||||
// Same vectors as meta-feeder-tmdb's vendored `hash.rs::url_cid_matches_pinned_vectors`.
|
||||
it('matches the pinned Rust vectors', () => {
|
||||
expect(encodeUrlCid('https://image.tmdb.org/t/p/w500/dqZENchTd7lp5zit1Q7Bkjzcxpi.jpg')).toBe(
|
||||
'bagdcaab7nb2hi4dthixs62lnmftwkltunvsgeltpojts65bpoaxxonjqgaxwi4k2ivhgg2cumq3wy4bvpjuximkrg5bgw2t2mn4ha2jonjygo',
|
||||
);
|
||||
});
|
||||
|
||||
it('uses a two-byte length varint past 127 bytes', () => {
|
||||
const long = `https://image.tmdb.org/t/p/w500/${'a'.repeat(100)}.jpg`;
|
||||
expect(long.length).toBe(136);
|
||||
expect(encodeUrlCid(long)).toBe(
|
||||
'bagdcaaeiafuhi5dqom5c6l3jnvqwozjoorwwiyron5zgol3uf5yc65zvgayc6ylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcylbmfqwcltkobtq',
|
||||
);
|
||||
});
|
||||
|
||||
it('rejects anything but an absolute http(s) URL', () => {
|
||||
expect(encodeUrlCid('/relative.jpg')).toBeNull();
|
||||
expect(encodeUrlCid('ftp://example.com/x.jpg')).toBeNull();
|
||||
expect(encodeUrlCid('')).toBeNull();
|
||||
});
|
||||
});
|
||||
|
||||
describe('posterLang3', () => {
|
||||
it('maps to ISO 639-3, never 639-2/B', () => {
|
||||
expect(posterLang3('fr')).toBe('fra');
|
||||
expect(posterLang3('de')).toBe('deu');
|
||||
expect(posterLang3('EN')).toBe('eng');
|
||||
});
|
||||
|
||||
it('files a textless poster under zxx and an unknown code under und', () => {
|
||||
expect(posterLang3(null)).toBe('zxx');
|
||||
expect(posterLang3(undefined)).toBe('zxx');
|
||||
expect(posterLang3('xx')).toBe('und');
|
||||
});
|
||||
});
|
||||
|
||||
describe('posterMembers', () => {
|
||||
const PRIMARY_CID = 'bagacbabaecontentcid';
|
||||
|
||||
it('lists the primary under its own language, once, beside the alternatives', () => {
|
||||
const data = {
|
||||
poster_path: '/primary.jpg',
|
||||
images: {
|
||||
posters: [img('/alt-fr.jpg', 'fr', 5.5, 10), img('/primary.jpg', 'ja', 5, 3), img('/textless.jpg', null, 6, 1)],
|
||||
},
|
||||
};
|
||||
expect(posterMembers(data, PRIMARY_CID)).toEqual({
|
||||
[`posters/jpn/${PRIMARY_CID}`]: 'true',
|
||||
[`posters/zxx/${alt('/textless.jpg')}`]: 'true',
|
||||
[`posters/fra/${alt('/alt-fr.jpg')}`]: 'true',
|
||||
});
|
||||
});
|
||||
|
||||
it('ranks by vote_average then vote_count, and caps at ten including the primary', () => {
|
||||
const posters = Array.from({ length: 15 }, (_, i) => img(`/p${String(i).padStart(2, '0')}.jpg`, 'en', i, 1));
|
||||
const keys = Object.keys(posterMembers({ poster_path: '/primary.jpg', images: { posters } }, PRIMARY_CID));
|
||||
expect(keys).toHaveLength(MAX_POSTER_MEMBERS);
|
||||
expect(keys[0]).toBe(`posters/und/${PRIMARY_CID}`);
|
||||
expect(keys[1]).toBe(`posters/eng/${alt('/p14.jpg')}`);
|
||||
expect(keys).not.toContain(`posters/eng/${alt('/p05.jpg')}`);
|
||||
|
||||
const tie = Object.keys(
|
||||
posterMembers(
|
||||
{ poster_path: '/primary.jpg', images: { posters: [img('/few.jpg', null, 5, 1), img('/many.jpg', null, 5, 9)] } },
|
||||
PRIMARY_CID,
|
||||
),
|
||||
);
|
||||
expect(tie[1]).toBe(`posters/zxx/${alt('/many.jpg')}`);
|
||||
});
|
||||
|
||||
it('writes no set when the payload has no image list', () => {
|
||||
expect(posterMembers({ poster_path: '/primary.jpg' }, PRIMARY_CID)).toEqual({});
|
||||
expect(posterMembers({ poster_path: '/primary.jpg', images: { posters: [] } }, PRIMARY_CID)).toEqual({});
|
||||
});
|
||||
});
|
||||
@@ -0,0 +1,9 @@
|
||||
import { defineConfig } from 'vitest/config';
|
||||
|
||||
export default defineConfig({
|
||||
test: {
|
||||
globals: true,
|
||||
environment: 'node',
|
||||
include: ['test/**/*.test.ts'],
|
||||
},
|
||||
});
|
||||
Reference in New Issue
Block a user