scan: validate arr URLs upfront, cache library once per scan

Two regressions from the radarr/sonarr fix:

1. ERR_INVALID_URL spam — when radarr_enabled='1' but radarr_url is empty
   or not http(s), every per-item fetch threw TypeError. We caught it but
   still ate the cost (and the log noise) on every movie. New isUsable()
   check on each service: enabled-in-config but URL doesn't parse →
   warn ONCE and skip arr lookups for the whole scan.

2. Per-item HTTP storm — for movies not in Radarr's library we used to
   hit /api/v3/movie (the WHOLE library) again per item, then two
   metadata-lookup calls. With 2000 items that's thousands of extra
   round-trips and the scan crawled. Now: pre-load the Radarr/Sonarr
   library once into Map<tmdbId,..>+Map<imdbId,..>+Map<tvdbId,..>,
   per-item lookups are O(1) memory hits, and only the genuinely-missing
   items make a single lookup-endpoint HTTP call.

The startup line now reports the library size:
  External language sources: radarr=enabled (https://..., 1287 movies in library), sonarr=...
so you can immediately see whether the cache loaded.

server/api/scan.ts

@@ -4,8 +4,16 @@ import { getAllConfig, getConfig, getDb, setConfig } from "../db/index";
 import { log, error as logError, warn } from "../lib/log";
 import { analyzeItem } from "../services/analyzer";
 import { extractOriginalLanguage, getAllItems, getDevItems, mapStream, normalizeLanguage } from "../services/jellyfin";
-import { getOriginalLanguage as radarrLang } from "../services/radarr";
-import { getOriginalLanguage as sonarrLang } from "../services/sonarr";
+import {
+	loadLibrary as loadRadarrLibrary,
+	getOriginalLanguage as radarrLang,
+	isUsable as radarrUsable,
+} from "../services/radarr";
+import {
+	loadLibrary as loadSonarrLibrary,
+	getOriginalLanguage as sonarrLang,
+	isUsable as sonarrUsable,
+} from "../services/sonarr";
 import type { MediaStream } from "../types";
 
 const app = new Hono();
@@ -152,13 +160,31 @@ async function runScan(limit: number | null = null): Promise<void> {
 	const jellyfinCfg = { url: cfg.jellyfin_url, apiKey: cfg.jellyfin_api_key, userId: cfg.jellyfin_user_id };
 	const subtitleLanguages = parseLanguageList(cfg.subtitle_languages ?? null, ["eng", "deu", "spa"]);
 	const audioLanguages = parseLanguageList(cfg.audio_languages ?? null, []);
-	const radarrEnabled = cfg.radarr_enabled === "1";
-	const sonarrEnabled = cfg.sonarr_enabled === "1";
+	const radarrCfg = { url: cfg.radarr_url, apiKey: cfg.radarr_api_key };
+	const sonarrCfg = { url: cfg.sonarr_url, apiKey: cfg.sonarr_api_key };
+	// 'enabled' in config means the user toggled it on. Only actually use it
+	// if the URL+key pass URL parsing — otherwise we'd hit ERR_INVALID_URL on
+	// every item. Refuse to call invalid endpoints rather than spamming logs.
+	const radarrEnabled = cfg.radarr_enabled === "1" && radarrUsable(radarrCfg);
+	const sonarrEnabled = cfg.sonarr_enabled === "1" && sonarrUsable(sonarrCfg);
+
+	if (cfg.radarr_enabled === "1" && !radarrEnabled) {
+		warn(`Radarr is enabled in config but URL/API key is invalid (url='${cfg.radarr_url}') — skipping Radarr lookups`);
+	}
+	if (cfg.sonarr_enabled === "1" && !sonarrEnabled) {
+		warn(`Sonarr is enabled in config but URL/API key is invalid (url='${cfg.sonarr_url}') — skipping Sonarr lookups`);
+	}
+
+	// Pre-load both libraries once so per-item lookups are O(1) cache hits
+	// instead of HTTP round-trips. The previous code called /api/v3/movie
+	// (the entire library!) once per item that didn't match by tmdbId.
+	const [radarrLibrary, sonarrLibrary] = await Promise.all([
+		radarrEnabled ? loadRadarrLibrary(radarrCfg) : Promise.resolve(null),
+		sonarrEnabled ? loadSonarrLibrary(sonarrCfg) : Promise.resolve(null),
+	]);
 
-	// Log the external-source situation once per scan so it's obvious from
-	// logs whether Radarr/Sonarr are actually being consulted.
 	log(
-		`External language sources: radarr=${radarrEnabled ? `enabled (${cfg.radarr_url || "NO URL"})` : "disabled"}, sonarr=${sonarrEnabled ? `enabled (${cfg.sonarr_url || "NO URL"})` : "disabled"}`,
+		`External language sources: radarr=${radarrEnabled ? `enabled (${cfg.radarr_url}, ${radarrLibrary?.byTmdbId.size ?? 0} movies in library)` : "disabled"}, sonarr=${sonarrEnabled ? `enabled (${cfg.sonarr_url}, ${sonarrLibrary?.byTvdbId.size ?? 0} series in library)` : "disabled"}`,
 	);
 	let radarrMisses = 0;
 	let radarrHits = 0;
@@ -250,14 +276,15 @@ async function runScan(limit: number | null = null): Promise<void> {
 			let needsReview = origLang ? 0 : 1;
 			let authoritative = false; // set when Radarr/Sonarr answers
 
-			if (jellyfinItem.Type === "Movie" && radarrEnabled) {
+			if (jellyfinItem.Type === "Movie" && radarrEnabled && radarrLibrary) {
 				if (!tmdbId && !imdbId) {
 					missingProviderIds++;
 					warn(`No tmdb/imdb id on '${jellyfinItem.Name}' — Radarr lookup skipped`);
 				} else {
 					const lang = await radarrLang(
-						{ url: cfg.radarr_url, apiKey: cfg.radarr_api_key },
+						radarrCfg,
 						{ tmdbId: tmdbId ?? undefined, imdbId: imdbId ?? undefined },
+						radarrLibrary,
 					);
 					if (lang) {
 						radarrHits++;
@@ -274,12 +301,12 @@ async function runScan(limit: number | null = null): Promise<void> {
 				}
 			}
 
-			if (jellyfinItem.Type === "Episode" && sonarrEnabled) {
+			if (jellyfinItem.Type === "Episode" && sonarrEnabled && sonarrLibrary) {
 				if (!tvdbId) {
 					missingProviderIds++;
 					warn(`No tvdb id on '${jellyfinItem.Name}' — Sonarr lookup skipped`);
 				} else {
-					const lang = await sonarrLang({ url: cfg.sonarr_url, apiKey: cfg.sonarr_api_key }, tvdbId);
+					const lang = await sonarrLang(sonarrCfg, tvdbId, sonarrLibrary);
 					if (lang) {
 						sonarrHits++;
 						if (origLang && normalizeLanguage(origLang) !== normalizeLanguage(lang)) needsReview = 1;

server/services/radarr.ts

@@ -10,7 +10,19 @@ function headers(apiKey: string): Record<string, string> {
 	return { "X-Api-Key": apiKey };
 }
 
+/** True only when url is an http(s) URL and apiKey is non-empty. */
+export function isUsable(cfg: RadarrConfig): boolean {
+	if (!cfg.apiKey) return false;
+	try {
+		const u = new URL(cfg.url);
+		return u.protocol === "http:" || u.protocol === "https:";
+	} catch {
+		return false;
+	}
+}
+
 export async function testConnection(cfg: RadarrConfig): Promise<{ ok: boolean; error?: string }> {
+	if (!isUsable(cfg)) return { ok: false, error: "Radarr URL/API key not configured" };
 	try {
 		const res = await fetch(`${cfg.url}/api/v3/system/status`, {
 			headers: headers(cfg.apiKey),
@@ -28,41 +40,68 @@ interface RadarrMovie {
 	originalLanguage?: { name: string; nameTranslated?: string };
 }
 
+/** Pre-loaded Radarr library indexed for O(1) per-item lookups during a scan. */
+export interface RadarrLibrary {
+	byTmdbId: Map<string, RadarrMovie>;
+	byImdbId: Map<string, RadarrMovie>;
+}
+
 async function fetchJson<T>(url: string, cfg: RadarrConfig, context: string): Promise<T | null> {
 	try {
 		const res = await fetch(url, { headers: headers(cfg.apiKey) });
 		if (!res.ok) {
-			warn(`Radarr ${context} → HTTP ${res.status} (${url.replace(cfg.apiKey, "***")})`);
+			warn(`Radarr ${context} → HTTP ${res.status}`);
 			return null;
 		}
 		return (await res.json()) as T;
 	} catch (e) {
-		logError(`Radarr ${context} failed:`, e);
+		logError(`Radarr ${context} failed:`, (e as Error).message);
 		return null;
 	}
 }
 
-/** Returns ISO 639-2 original language or null. */
+/**
+ * Fetch the full Radarr library once and index by tmdbId + imdbId. The
+ * caller passes this around for the duration of the scan so per-item
+ * lookups don't hit Radarr at all when the movie is in the library.
+ */
+export async function loadLibrary(cfg: RadarrConfig): Promise<RadarrLibrary> {
+	const empty: RadarrLibrary = { byTmdbId: new Map(), byImdbId: new Map() };
+	if (!isUsable(cfg)) return empty;
+	const list = await fetchJson<RadarrMovie[]>(`${cfg.url}/api/v3/movie`, cfg, "library/all");
+	if (!list) return empty;
+	const byTmdbId = new Map<string, RadarrMovie>();
+	const byImdbId = new Map<string, RadarrMovie>();
+	for (const m of list) {
+		if (m.tmdbId != null) byTmdbId.set(String(m.tmdbId), m);
+		if (m.imdbId) byImdbId.set(m.imdbId, m);
+	}
+	return { byTmdbId, byImdbId };
+}
+
+/**
+ * Returns ISO 639-2 original language or null. Uses the pre-loaded library
+ * first (no HTTP), falls back to a single metadata-lookup HTTP call when
+ * the movie isn't in Radarr's library (e.g. on disk + in Jellyfin only).
+ */
 export async function getOriginalLanguage(
 	cfg: RadarrConfig,
 	ids: { tmdbId?: string; imdbId?: string },
+	library: RadarrLibrary,
 ): Promise<string | null> {
-	// 1. Library lookup by tmdbId: only hits movies actively managed in Radarr.
+	// Library hit — no HTTP at all.
 	if (ids.tmdbId) {
-		const list = await fetchJson<RadarrMovie[]>(`${cfg.url}/api/v3/movie?tmdbId=${ids.tmdbId}`, cfg, "library/tmdb");
-		const movie = list?.[0];
-		if (movie?.originalLanguage) return nameToIso(movie.originalLanguage.name);
+		const m = library.byTmdbId.get(ids.tmdbId);
+		if (m?.originalLanguage) return nameToIso(m.originalLanguage.name);
 	}
-
-	// 2. Library lookup by imdbId (sidestep: pull all and filter).
 	if (ids.imdbId) {
-		const list = await fetchJson<(RadarrMovie & { imdbId?: string })[]>(`${cfg.url}/api/v3/movie`, cfg, "library/all");
-		const movie = list?.find((m) => m.imdbId === ids.imdbId);
-		if (movie?.originalLanguage) return nameToIso(movie.originalLanguage.name);
+		const m = library.byImdbId.get(ids.imdbId);
+		if (m?.originalLanguage) return nameToIso(m.originalLanguage.name);
 	}
 
-	// 3. Metadata lookup — asks Radarr to consult TMDB for movies NOT in the
-	//    library. This is the fix for the 8-Mile-in-Jellyfin-but-not-in-Radarr case.
+	// Missing from library — one metadata-lookup call. Prefer tmdbId since
+	// it goes straight to TMDB; fall back to imdb if that's all we have.
+	if (!isUsable(cfg)) return null;
 	if (ids.tmdbId) {
 		const result = await fetchJson<RadarrMovie>(
 			`${cfg.url}/api/v3/movie/lookup/tmdb?tmdbId=${ids.tmdbId}`,
@@ -70,9 +109,7 @@ export async function getOriginalLanguage(
 			"lookup/tmdb",
 		);
 		if (result?.originalLanguage) return nameToIso(result.originalLanguage.name);
-	}
-
-	if (ids.imdbId) {
+	} else if (ids.imdbId) {
 		const result = await fetchJson<RadarrMovie>(
 			`${cfg.url}/api/v3/movie/lookup/imdb?imdbId=${ids.imdbId}`,
 			cfg,
@@ -137,16 +174,10 @@ const NAME_TO_639_2: Record<string, string> = {
 	"portuguese (brazil)": "por",
 };
 
-/** Map a Radarr language name to ISO 639-2 or null. Logs unmapped names. */
 function nameToIso(name: string): string | null {
 	const key = name.toLowerCase().trim();
 	if (NAME_TO_639_2[key]) return NAME_TO_639_2[key];
-	// If Radarr ever returns a bare code (e.g., "en", "eng"), normalizeLanguage
-	// will pass it through. Anything else is unknown — return null instead of
-	// guessing so scan.ts can mark needs_review.
-	if (key.length === 2 || key.length === 3) {
-		return normalizeLanguage(key);
-	}
-	warn(`Radarr language name not recognised: '${name}'. Add it to NAME_TO_639_2 or review upstream.`);
+	if (key.length === 2 || key.length === 3) return normalizeLanguage(key);
+	warn(`Radarr language name not recognised: '${name}'.`);
 	return null;
 }

server/services/sonarr.ts

@@ -10,7 +10,19 @@ function headers(apiKey: string): Record<string, string> {
 	return { "X-Api-Key": apiKey };
 }
 
+/** True only when url is an http(s) URL and apiKey is non-empty. */
+export function isUsable(cfg: SonarrConfig): boolean {
+	if (!cfg.apiKey) return false;
+	try {
+		const u = new URL(cfg.url);
+		return u.protocol === "http:" || u.protocol === "https:";
+	} catch {
+		return false;
+	}
+}
+
 export async function testConnection(cfg: SonarrConfig): Promise<{ ok: boolean; error?: string }> {
+	if (!isUsable(cfg)) return { ok: false, error: "Sonarr URL/API key not configured" };
 	try {
 		const res = await fetch(`${cfg.url}/api/v3/system/status`, {
 			headers: headers(cfg.apiKey),
@@ -27,28 +39,47 @@ interface SonarrSeries {
 	originalLanguage?: { name: string };
 }
 
+/** Pre-loaded Sonarr library indexed for O(1) per-series lookups during a scan. */
+export interface SonarrLibrary {
+	byTvdbId: Map<string, SonarrSeries>;
+}
+
 async function fetchJson<T>(url: string, cfg: SonarrConfig, context: string): Promise<T | null> {
 	try {
 		const res = await fetch(url, { headers: headers(cfg.apiKey) });
 		if (!res.ok) {
-			warn(`Sonarr ${context} → HTTP ${res.status} (${url.replace(cfg.apiKey, "***")})`);
+			warn(`Sonarr ${context} → HTTP ${res.status}`);
 			return null;
 		}
 		return (await res.json()) as T;
 	} catch (e) {
-		logError(`Sonarr ${context} failed:`, e);
+		logError(`Sonarr ${context} failed:`, (e as Error).message);
 		return null;
 	}
 }
 
-/** Returns ISO 639-2 original language for a series or null. */
-export async function getOriginalLanguage(cfg: SonarrConfig, tvdbId: string): Promise<string | null> {
-	// 1. Library: only hits series actively managed in Sonarr.
-	const list = await fetchJson<SonarrSeries[]>(`${cfg.url}/api/v3/series?tvdbId=${tvdbId}`, cfg, "library/tvdb");
-	const series = list?.[0];
-	if (series?.originalLanguage) return nameToIso(series.originalLanguage.name);
+export async function loadLibrary(cfg: SonarrConfig): Promise<SonarrLibrary> {
+	const empty: SonarrLibrary = { byTvdbId: new Map() };
+	if (!isUsable(cfg)) return empty;
+	const list = await fetchJson<SonarrSeries[]>(`${cfg.url}/api/v3/series`, cfg, "library/all");
+	if (!list) return empty;
+	const byTvdbId = new Map<string, SonarrSeries>();
+	for (const s of list) {
+		if (s.tvdbId != null) byTvdbId.set(String(s.tvdbId), s);
+	}
+	return { byTvdbId };
+}
 
-	// 2. Metadata lookup via Sonarr's TVDB bridge for series not in the library.
+/** Returns ISO 639-2 original language for a series or null. */
+export async function getOriginalLanguage(
+	cfg: SonarrConfig,
+	tvdbId: string,
+	library: SonarrLibrary,
+): Promise<string | null> {
+	const m = library.byTvdbId.get(tvdbId);
+	if (m?.originalLanguage) return nameToIso(m.originalLanguage.name);
+
+	if (!isUsable(cfg)) return null;
 	const lookup = await fetchJson<SonarrSeries[]>(
 		`${cfg.url}/api/v3/series/lookup?term=tvdb%3A${tvdbId}`,
 		cfg,