feat(GRO-2153): abstracted geocoding service (Nominatim + Google)

Phase 1.2 of Route Optimization. Adds a provider-agnostic geocoding service layer in the deployed src/ tree: - GeocodingProvider interface + GeocodeResult type - NominatimGeocodingProvider (default, free, self-hostable) with an internal rate limiter enforcing the 1 req/sec Nominatim usage policy - GoogleGeocodingProvider (optional fallback) keyed by the encrypted businessSettings.googleMapsApiKey (decrypted via decryptSecret) or GOOGLE_MAPS_API_KEY env fallback - resolveGeocodingProvider() selecting on businessSettings.routeOptimizationProvider, with safe fallback to Nominatim when google is configured but no usable key - geocodeBatch() throttled batch utility (honors provider rate limit, captures per-item errors, optional progress callback) - 20 unit tests covering both providers, selection, throttle spacing, and batch Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-06-08 09:01:36 +00:00
parent 6be78cae35
commit 2fa6e3d87b
2 changed files with 732 additions and 0 deletions
@@ -0,0 +1,419 @@
+import { decryptSecret } from "@groombook/db";
+
+/**
+ * Abstracted geocoding service layer (GRO-2153, Phase 1.2 of Route Optimization).
+ *
+ * Provides a provider-agnostic interface for turning a street address into
+ * latitude/longitude coordinates, with two concrete implementations:
+ *
+ *  - {@link NominatimGeocodingProvider} — OpenStreetMap Nominatim (default, free,
+ *    self-hostable). Enforces the public Nominatim usage policy of at most one
+ *    request per second.
+ *  - {@link GoogleGeocodingProvider} — Google Geocoding API (optional fallback,
+ *    requires an API key stored encrypted at rest in `businessSettings`).
+ *
+ * Provider selection is driven by `businessSettings.routeOptimizationProvider`
+ * via {@link resolveGeocodingProvider}. A {@link geocodeBatch} helper geocodes a
+ * list of addresses while respecting the active provider's rate limit.
+ */
+
+/** Identifier for a supported geocoding backend. Mirrors `route_optimization_provider`. */
+export type GeocodingProviderName = "nominatim" | "google";
+
+/** Successful geocoding result in WGS84 decimal degrees. */
+export interface GeocodeResult {
+  latitude: number;
+  longitude: number;
+  /** Provider-normalized display address, when available. */
+  formattedAddress: string | null;
+  /** Which provider produced this result. */
+  provider: GeocodingProviderName;
+}
+
+/** Abstract geocoding provider contract. */
+export interface GeocodingProvider {
+  /** Stable provider identifier. */
+  readonly name: GeocodingProviderName;
+  /**
+   * Minimum milliseconds to leave between consecutive requests. Used both by the
+   * provider's internal rate limiter and by {@link geocodeBatch} so callers do
+   * not need to know provider-specific limits.
+   */
+  readonly minRequestIntervalMs: number;
+  /**
+   * Geocode a single address.
+   * @returns the top match, or `null` when the address is blank or unresolvable.
+   * @throws on transport failures or provider-level errors (e.g. quota, bad key).
+   */
+  geocode(address: string): Promise<GeocodeResult | null>;
+}
+
+// --- Constants -------------------------------------------------------------
+
+/** Nominatim usage policy: at most 1 request per second. */
+const NOMINATIM_RATE_LIMIT_MS = 1000;
+const DEFAULT_NOMINATIM_BASE_URL = "https://nominatim.openstreetmap.org";
+/** Nominatim requires a descriptive User-Agent identifying the application. */
+const DEFAULT_NOMINATIM_USER_AGENT = "GroomBook/1.0 (+https://groombook.app route-optimization)";
+
+const GOOGLE_GEOCODE_BASE_URL = "https://maps.googleapis.com/maps/api/geocode/json";
+/**
+ * Google permits a high request rate; a small floor keeps batch traffic polite
+ * without throttling interactive single calls.
+ */
+const GOOGLE_RATE_LIMIT_MS = 20;
+
+// --- Injectable primitives (for testability) -------------------------------
+
+/** Minimal `fetch` shape this module depends on. Defaults to the global `fetch`. */
+export type FetchLike = (
+  input: string,
+  init?: { headers?: Record<string, string> }
+) => Promise<{
+  ok: boolean;
+  status: number;
+  statusText: string;
+  json(): Promise<unknown>;
+}>;
+
+type NowFn = () => number;
+type SleepFn = (ms: number) => Promise<void>;
+
+const defaultSleep: SleepFn = (ms) =>
+  new Promise((resolve) => setTimeout(resolve, ms));
+
+const defaultFetch: FetchLike = (input, init) =>
+  (globalThis.fetch as unknown as FetchLike)(input, init);
+
+/**
+ * Serializes async tasks and guarantees at least `intervalMs` between the start
+ * of consecutive tasks. A failing task never wedges the queue.
+ */
+class RateLimiter {
+  // Negative infinity so the very first task never waits.
+  private last = Number.NEGATIVE_INFINITY;
+  private chain: Promise<unknown> = Promise.resolve();
+
+  constructor(
+    private readonly intervalMs: number,
+    private readonly now: NowFn = Date.now,
+    private readonly sleep: SleepFn = defaultSleep
+  ) {}
+
+  run<T>(task: () => Promise<T>): Promise<T> {
+    const result = this.chain.then(async () => {
+      if (this.intervalMs > 0) {
+        const elapsed = this.now() - this.last;
+        const wait = this.intervalMs - elapsed;
+        if (wait > 0) await this.sleep(wait);
+      }
+      this.last = this.now();
+      return task();
+    });
+    // Keep the queue alive regardless of whether this task resolves or rejects.
+    this.chain = result.then(
+      () => undefined,
+      () => undefined
+    );
+    return result;
+  }
+}
+
+function normalizeAddress(address: string): string {
+  return address.trim();
+}
+
+// --- Nominatim provider ----------------------------------------------------
+
+interface NominatimSearchRow {
+  lat?: string;
+  lon?: string;
+  display_name?: string;
+}
+
+export interface NominatimProviderOptions {
+  /** Override the Nominatim instance base URL (e.g. a self-hosted mirror). */
+  baseUrl?: string;
+  /** User-Agent header identifying this application, per Nominatim policy. */
+  userAgent?: string;
+  /** Minimum spacing between requests; defaults to the 1 req/sec policy. */
+  minRequestIntervalMs?: number;
+  fetchImpl?: FetchLike;
+  now?: NowFn;
+  sleep?: SleepFn;
+}
+
+export class NominatimGeocodingProvider implements GeocodingProvider {
+  readonly name = "nominatim" as const;
+  readonly minRequestIntervalMs: number;
+
+  private readonly baseUrl: string;
+  private readonly userAgent: string;
+  private readonly fetchImpl: FetchLike;
+  private readonly limiter: RateLimiter;
+
+  constructor(options: NominatimProviderOptions = {}) {
+    this.baseUrl = (options.baseUrl ?? DEFAULT_NOMINATIM_BASE_URL).replace(/\/+$/, "");
+    this.userAgent = options.userAgent ?? DEFAULT_NOMINATIM_USER_AGENT;
+    this.minRequestIntervalMs = options.minRequestIntervalMs ?? NOMINATIM_RATE_LIMIT_MS;
+    this.fetchImpl = options.fetchImpl ?? defaultFetch;
+    this.limiter = new RateLimiter(this.minRequestIntervalMs, options.now, options.sleep);
+  }
+
+  async geocode(address: string): Promise<GeocodeResult | null> {
+    const query = normalizeAddress(address);
+    if (!query) return null;
+
+    return this.limiter.run(async () => {
+      const url = new URL(`${this.baseUrl}/search`);
+      url.searchParams.set("q", query);
+      url.searchParams.set("format", "jsonv2");
+      url.searchParams.set("limit", "1");
+
+      const res = await this.fetchImpl(url.toString(), {
+        headers: { "User-Agent": this.userAgent },
+      });
+      if (!res.ok) {
+        throw new Error(
+          `Nominatim geocoding failed: ${res.status} ${res.statusText}`
+        );
+      }
+
+      const body = (await res.json()) as NominatimSearchRow[];
+      if (!Array.isArray(body) || body.length === 0) return null;
+
+      const top = body[0]!;
+      const latitude = Number(top.lat);
+      const longitude = Number(top.lon);
+      if (!Number.isFinite(latitude) || !Number.isFinite(longitude)) return null;
+
+      return {
+        latitude,
+        longitude,
+        formattedAddress: top.display_name ?? null,
+        provider: this.name,
+      };
+    });
+  }
+}
+
+// --- Google provider -------------------------------------------------------
+
+interface GoogleGeocodeResponse {
+  status: string;
+  error_message?: string;
+  results?: Array<{
+    formatted_address?: string;
+    geometry?: { location?: { lat?: number; lng?: number } };
+  }>;
+}
+
+export interface GoogleProviderOptions {
+  baseUrl?: string;
+  minRequestIntervalMs?: number;
+  fetchImpl?: FetchLike;
+  now?: NowFn;
+  sleep?: SleepFn;
+}
+
+export class GoogleGeocodingProvider implements GeocodingProvider {
+  readonly name = "google" as const;
+  readonly minRequestIntervalMs: number;
+
+  private readonly apiKey: string;
+  private readonly baseUrl: string;
+  private readonly fetchImpl: FetchLike;
+  private readonly limiter: RateLimiter;
+
+  constructor(apiKey: string, options: GoogleProviderOptions = {}) {
+    if (!apiKey) {
+      throw new Error("GoogleGeocodingProvider requires a non-empty API key");
+    }
+    this.apiKey = apiKey;
+    this.baseUrl = options.baseUrl ?? GOOGLE_GEOCODE_BASE_URL;
+    this.minRequestIntervalMs = options.minRequestIntervalMs ?? GOOGLE_RATE_LIMIT_MS;
+    this.fetchImpl = options.fetchImpl ?? defaultFetch;
+    this.limiter = new RateLimiter(this.minRequestIntervalMs, options.now, options.sleep);
+  }
+
+  async geocode(address: string): Promise<GeocodeResult | null> {
+    const query = normalizeAddress(address);
+    if (!query) return null;
+
+    return this.limiter.run(async () => {
+      const url = new URL(this.baseUrl);
+      url.searchParams.set("address", query);
+      url.searchParams.set("key", this.apiKey);
+
+      const res = await this.fetchImpl(url.toString());
+      if (!res.ok) {
+        throw new Error(
+          `Google geocoding failed: ${res.status} ${res.statusText}`
+        );
+      }
+
+      const body = (await res.json()) as GoogleGeocodeResponse;
+      if (body.status === "ZERO_RESULTS") return null;
+      if (body.status !== "OK") {
+        const detail = body.error_message ? `: ${body.error_message}` : "";
+        throw new Error(`Google geocoding error: ${body.status}${detail}`);
+      }
+
+      const top = body.results?.[0];
+      const location = top?.geometry?.location;
+      const latitude = location?.lat;
+      const longitude = location?.lng;
+      if (
+        typeof latitude !== "number" ||
+        typeof longitude !== "number" ||
+        !Number.isFinite(latitude) ||
+        !Number.isFinite(longitude)
+      ) {
+        return null;
+      }
+
+      return {
+        latitude,
+        longitude,
+        formattedAddress: top?.formatted_address ?? null,
+        provider: this.name,
+      };
+    });
+  }
+}
+
+// --- Provider selection ----------------------------------------------------
+
+/** Subset of `businessSettings` relevant to geocoding provider selection. */
+export interface GeocodingSettings {
+  routeOptimizationProvider?: string | null;
+  /** Google API key, encrypted at rest (AES-256-GCM via `encryptSecret`). */
+  googleMapsApiKey?: string | null;
+}
+
+export interface ResolveProviderOptions {
+  /** Decryption function for the stored Google key. Defaults to `decryptSecret`. */
+  decrypt?: (ciphertext: string) => string;
+  /** Options forwarded to the constructed provider (base URL, fetch, timing). */
+  nominatim?: NominatimProviderOptions;
+  google?: GoogleProviderOptions;
+  /** Sink for non-fatal selection warnings. Defaults to `console.warn`. */
+  warn?: (message: string) => void;
+}
+
+/**
+ * Resolves the Google API key from settings (decrypting the at-rest value) or,
+ * as a development convenience, from the `GOOGLE_MAPS_API_KEY` env var.
+ * Returns `null` when no usable key is available.
+ */
+function resolveGoogleApiKey(
+  settings: GeocodingSettings,
+  decrypt: (ciphertext: string) => string,
+  warn: (message: string) => void
+): string | null {
+  const stored = settings.googleMapsApiKey?.trim();
+  if (stored) {
+    try {
+      const decrypted = decrypt(stored).trim();
+      if (decrypted) return decrypted;
+    } catch (err) {
+      warn(
+        `Failed to decrypt googleMapsApiKey; falling back to Nominatim: ${
+          err instanceof Error ? err.message : String(err)
+        }`
+      );
+      return null;
+    }
+  }
+  const fromEnv = process.env.GOOGLE_MAPS_API_KEY?.trim();
+  return fromEnv ? fromEnv : null;
+}
+
+/**
+ * Selects a geocoding provider based on `businessSettings.routeOptimizationProvider`.
+ *
+ * - `"google"` returns a {@link GoogleGeocodingProvider} when a usable key exists,
+ *   otherwise warns and falls back to Nominatim.
+ * - Any other value (including `null`/`undefined`) returns a
+ *   {@link NominatimGeocodingProvider}.
+ */
+export function resolveGeocodingProvider(
+  settings: GeocodingSettings | null | undefined,
+  options: ResolveProviderOptions = {}
+): GeocodingProvider {
+  const warn = options.warn ?? ((message: string) => console.warn(message));
+  const decrypt = options.decrypt ?? decryptSecret;
+  const requested = settings?.routeOptimizationProvider ?? "nominatim";
+
+  if (requested === "google") {
+    const apiKey = resolveGoogleApiKey(settings ?? {}, decrypt, warn);
+    if (apiKey) {
+      return new GoogleGeocodingProvider(apiKey, options.google);
+    }
+    warn(
+      "routeOptimizationProvider is 'google' but no usable API key was found; falling back to Nominatim"
+    );
+  }
+
+  return new NominatimGeocodingProvider(options.nominatim);
+}
+
+// --- Batch geocoding -------------------------------------------------------
+
+/** Input item for {@link geocodeBatch}: a caller-defined key and its address. */
+export interface BatchGeocodeItem<K> {
+  key: K;
+  address: string;
+}
+
+/** Per-item outcome from {@link geocodeBatch}. */
+export interface BatchGeocodeOutcome<K> {
+  key: K;
+  address: string;
+  /** Resolved coordinates, or `null` when unresolvable. */
+  result: GeocodeResult | null;
+  /** Present when the geocode call threw; the batch continues past errors. */
+  error?: string;
+}
+
+export interface GeocodeBatchOptions<K> {
+  /** Invoked after each item completes; useful for progress reporting. */
+  onProgress?: (
+    completed: number,
+    total: number,
+    outcome: BatchGeocodeOutcome<K>
+  ) => void;
+}
+
+/**
+ * Geocodes a list of addresses sequentially through the given provider. The
+ * provider's internal rate limiter enforces throttling (e.g. Nominatim's
+ * 1 req/sec), so addresses are processed one at a time and individual failures
+ * are captured per item rather than aborting the whole batch.
+ */
+export async function geocodeBatch<K>(
+  items: ReadonlyArray<BatchGeocodeItem<K>>,
+  provider: GeocodingProvider,
+  options: GeocodeBatchOptions<K> = {}
+): Promise<Array<BatchGeocodeOutcome<K>>> {
+  const outcomes: Array<BatchGeocodeOutcome<K>> = [];
+
+  for (const item of items) {
+    let outcome: BatchGeocodeOutcome<K>;
+    try {
+      const result = await provider.geocode(item.address);
+      outcome = { key: item.key, address: item.address, result };
+    } catch (err) {
+      outcome = {
+        key: item.key,
+        address: item.address,
+        result: null,
+        error: err instanceof Error ? err.message : String(err),
+      };
+    }
+    outcomes.push(outcome);
+    options.onProgress?.(outcomes.length, items.length, outcome);
+  }
+
+  return outcomes;
+}