euredact 0.3.9 → 0.4.0

package/dist/cjs/sdk.js

@@ -1,4 +1,37 @@
 "use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.EuRedact = void 0;
 const cache_js_1 = require("./cache.js");
@@ -48,6 +81,40 @@ class ReferentialMapper {
     }
 }
 const DEFAULT_MAX_INPUT_LENGTH = 10_485_760; // ~10 MB of text
+/**
+ * Send a document to the cloud tier.
+ *
+ * Options the service cannot honour throw rather than being ignored. Silently
+ * dropping one would mean returning a result that does not match what was
+ * asked for — which, for anything that changes which spans come back, is
+ * under-redaction wearing a plausible face.
+ */
+async function redactViaCloud(text, options) {
+    const { CloudClient } = await Promise.resolve().then(() => __importStar(require("./cloud/client.js")));
+    const countries = options.countries ?? null;
+    if (!countries || countries.length !== 1) {
+        throw new Error('cloud mode needs exactly one country, e.g. { countries: ["BE"] }. The ' +
+            "model is trained and evaluated per country, so a multi-country request " +
+            "has no defined behaviour.");
+    }
+    if (options.countryHint) {
+        throw new Error("countryHint is not supported in cloud mode");
+    }
+    if (options.context || options.chunkOffset) {
+        throw new Error("context/chunkOffset are not supported in cloud mode: the model has never " +
+            "seen a chunk boundary, so the service rejects oversized input rather " +
+            "than splitting it");
+    }
+    if (options.referentialIntegrity) {
+        throw new Error("referentialIntegrity is not supported in cloud mode");
+    }
+    // detectDates is deliberately NOT forwarded. The service always runs its
+    // rules engine with dates on, because that is what the model was trained
+    // against; the caller's value cannot change that. It is the one ignored
+    // option that is safe to ignore — it can only cause MORE to be detected,
+    // never less, so it cannot produce under-redaction.
+    return new CloudClient().redact(text, { country: countries[0] });
+}
 /**
  * Reject a bare string where a list of country codes is expected.
  *
@@ -95,6 +162,19 @@ class EuRedact {
     redact(text, options = {}) {
         checkCountryArg(options.countries, "countries");
         checkCountryArg(options.countryHint, "countryHint");
+        const requestedMode = options.mode ?? "rules";
+        if (requestedMode === "cloud") {
+            // Cannot be served synchronously: the cloud tier is a network call, and
+            // returning rules-only output instead would be the silent under-redaction
+            // this guard exists to prevent. Point at the async entry point rather
+            // than inventing a Promise-or-value union return type.
+            throw new Error('mode: "cloud" is asynchronous — use redactAsync(text, { mode: "cloud" }) ' +
+                "(or EuRedact#redactAsync). redact() is synchronous and can only serve " +
+                'mode: "rules".');
+        }
+        if (requestedMode !== "rules") {
+            throw new Error(`unknown mode ${JSON.stringify(requestedMode)}: expected "rules" or "cloud"`);
+        }
         if (text.length > this.maxInputLength) {
             throw new Error(`Input text length (${text.length.toLocaleString()} chars) exceeds the maximum ` +
                 `(${this.maxInputLength.toLocaleString()} chars). Split the input or ` +
@@ -178,6 +258,18 @@ class EuRedact {
         }
         return result;
     }
+    /**
+     * Redact, awaiting the cloud tier when `mode: "cloud"` is asked for.
+     *
+     * `mode: "rules"` resolves immediately with exactly what `redact()` returns,
+     * so callers that may or may not use the cloud tier can hold one code path.
+     */
+    async redactAsync(text, options = {}) {
+        const mode = options.mode ?? "rules";
+        if (mode !== "cloud")
+            return this.redact(text, options);
+        return redactViaCloud(text, options);
+    }
     redactBatch(texts, options = {}) {
         // Generation is country-blind, so every country's patterns are loaded
         // regardless of what the caller declared.

package/dist/cjs/types.js

@@ -1,9 +1,25 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.DetectionSource = exports.LEGACY_TYPE_ALIASES = exports.EntityType = void 0;
+exports.canonicalType = canonicalType;
 var EntityType;
 (function (EntityType) {
-    EntityType["NAME"] = "NAME";
+    /**
+     * Canonical name for a person's name. `NAME` is kept below as a legacy alias
+     * with the same value, exactly as `IBAN` aliases `BANK_ACCOUNT`, so
+     * `EntityType.NAME === EntityType.PERSON_NAME` and existing code keeps
+     * working — but the emitted value, and therefore the `[PERSON_NAME]`
+     * placeholder written into redacted text, is canonical.
+     *
+     * Renamed rather than translated at the cloud boundary: two names for one
+     * type is how a whole category goes missing when someone filters on the
+     * spelling they happened to know. Nothing could have depended on the old
+     * value — the type is cloud-only and the cloud tier was stubbed, so it was
+     * never emitted.
+     */
+    EntityType["PERSON_NAME"] = "PERSON_NAME";
+    /** @deprecated legacy alias of {@link EntityType.PERSON_NAME} */
+    EntityType["NAME"] = "PERSON_NAME";
     EntityType["ADDRESS"] = "ADDRESS";
     /**
      * Canonical name for a bank account identifier. `IBAN` is kept below as a
@@ -51,6 +67,20 @@ var EntityType;
     EntityType["UUID"] = "UUID";
     EntityType["SOCIAL_HANDLE"] = "SOCIAL_HANDLE";
     EntityType["SECRET"] = "SECRET";
+    // ── [CLOUD EXTENSION] ────────────────────────────────────────────────────
+    // Types only the cloud tier can detect. The rule engine never emits these —
+    // there is no shape to match on — which is exactly why they exist: the
+    // model's job is what the rules cannot catch. Named to match the detection
+    // canon the service is trained and evaluated against.
+    EntityType["ORGANISATION_NAME"] = "ORGANISATION_NAME";
+    EntityType["JOB_TITLE"] = "JOB_TITLE";
+    EntityType["MEDICAL_CONDITION"] = "MEDICAL_CONDITION";
+    EntityType["SENSITIVE_ATTRIBUTE"] = "SENSITIVE_ATTRIBUTE";
+    EntityType["BIOMETRIC_REF"] = "BIOMETRIC_REF";
+    EntityType["FINANCIAL_AMOUNT"] = "FINANCIAL_AMOUNT";
+    EntityType["QUASI_IDENTIFIER"] = "QUASI_IDENTIFIER";
+    EntityType["CREDENTIAL"] = "CREDENTIAL";
+    EntityType["URL"] = "URL";
     EntityType["OTHER"] = "OTHER";
 })(EntityType || (exports.EntityType = EntityType = {}));
 /**
@@ -60,7 +90,21 @@ var EntityType;
  */
 exports.LEGACY_TYPE_ALIASES = {
     IBAN: "BANK_ACCOUNT",
+    NAME: "PERSON_NAME",
+    STREET_ADDRESS: "ADDRESS",
+    NATIONALITY_ETHNICITY: "SENSITIVE_ATTRIBUTE",
 };
+/**
+ * Resolve a possibly-legacy type name to its canonical form.
+ *
+ * Mirrors `EntityType._missing_` in the Python SDK. An unrecognised name is
+ * returned unchanged rather than dropped: a client one release behind the
+ * service must not silently lose a whole category of PII.
+ */
+function canonicalType(name) {
+    const upper = String(name ?? "").trim().toUpperCase();
+    return exports.LEGACY_TYPE_ALIASES[upper] ?? upper;
+}
 var DetectionSource;
 (function (DetectionSource) {
     DetectionSource["RULES"] = "rules";

package/dist/esm/cloud/client.d.ts

@@ -0,0 +1,42 @@
+/**
+ * [CLOUD EXTENSION] The cloud client.
+ *
+ * Speaks to the euRedact inference service: a deterministic rules engine
+ * followed by a fine-tuned model asked only *what did the rules miss?* The
+ * service returns both the redacted document and located spans, so this
+ * module's job is transport and translation, not detection.
+ *
+ * Three things it hides from the caller:
+ *
+ * - **Whether the document finished inside the service's sync window.** A short
+ *   document comes back on the original request; a long one, or one arriving
+ *   while a GPU node is cold, returns a job handle and is polled. Callers
+ *   should never write that branch, so `redact()` resolves either way.
+ * - **Retries.** Every request carries an `Idempotency-Key`, so a retry after a
+ *   timeout cannot produce a second job or a second usage row.
+ * - **`Retry-After`.** When the service says how long to wait, that is obeyed
+ *   rather than second-guessed with a backoff curve.
+ *
+ * Zero dependencies, like the rest of the package: this uses the platform's
+ * own `fetch`.
+ */
+import { type CloudConfig } from "./config.js";
+import { type RedactResult } from "../types.js";
+export interface CloudRedactOptions {
+    country: string;
+    language?: string;
+    priority?: "interactive" | "batch";
+    rulesOnly?: boolean;
+    idempotencyKey?: string;
+    /** Injectable for tests. Defaults to the platform `fetch`. */
+    fetchImpl?: typeof fetch;
+}
+export declare class CloudClient {
+    readonly config: CloudConfig;
+    constructor(config?: CloudConfig | null);
+    private headers;
+    private raiseFor;
+    redact(text: string, options: CloudRedactOptions): Promise<RedactResult>;
+    private poll;
+    private withTimeout;
+}

package/dist/esm/cloud/client.js

@@ -0,0 +1,215 @@
+/**
+ * [CLOUD EXTENSION] The cloud client.
+ *
+ * Speaks to the euRedact inference service: a deterministic rules engine
+ * followed by a fine-tuned model asked only *what did the rules miss?* The
+ * service returns both the redacted document and located spans, so this
+ * module's job is transport and translation, not detection.
+ *
+ * Three things it hides from the caller:
+ *
+ * - **Whether the document finished inside the service's sync window.** A short
+ *   document comes back on the original request; a long one, or one arriving
+ *   while a GPU node is cold, returns a job handle and is polled. Callers
+ *   should never write that branch, so `redact()` resolves either way.
+ * - **Retries.** Every request carries an `Idempotency-Key`, so a retry after a
+ *   timeout cannot produce a second job or a second usage row.
+ * - **`Retry-After`.** When the service says how long to wait, that is obeyed
+ *   rather than second-guessed with a backoff curve.
+ *
+ * Zero dependencies, like the rest of the package: this uses the platform's
+ * own `fetch`.
+ */
+import { getConfig, } from "./config.js";
+import { CloudError, NotConfiguredError, QuotaExceededError, TooLargeError, } from "./errors.js";
+import { canonicalType, DetectionSource, EntityType, } from "../types.js";
+const RETRY_STATUS = new Set([429, 500, 502, 503, 504]);
+const MAX_BACKOFF_MS = 30_000;
+function requireFetch(supplied) {
+    const impl = supplied ?? globalThis.fetch;
+    if (typeof impl !== "function") {
+        throw new NotConfiguredError("the cloud tier needs a global fetch. Node 18+ provides one; on Node 16 " +
+            "supply one via the fetchImpl option or install a polyfill.");
+    }
+    return impl;
+}
+function sleep(ms) {
+    return new Promise(resolve => setTimeout(resolve, ms));
+}
+function retryAfterMs(headers) {
+    const raw = headers?.get("Retry-After");
+    if (!raw)
+        return null;
+    const seconds = Number(raw);
+    // The HTTP-date form is legal too; fall back to backoff rather than parse it.
+    return Number.isFinite(seconds) ? Math.max(0, seconds * 1000) : null;
+}
+/**
+ * Full jitter. A synchronised retry storm from many clients is how a
+ * recovering service is knocked back over.
+ */
+function backoffMs(attempt, retryAfter) {
+    if (retryAfter !== null)
+        return Math.min(retryAfter, MAX_BACKOFF_MS);
+    return Math.min(MAX_BACKOFF_MS, Math.random() * 2 ** attempt * 1000);
+}
+function uuid() {
+    const c = globalThis.crypto;
+    if (c && typeof c.randomUUID === "function")
+        return c.randomUUID();
+    return `idem-${Date.now().toString(16)}-${Math.random().toString(16).slice(2)}`;
+}
+/**
+ * Canon name -> EntityType, tolerating one this build has not heard of.
+ *
+ * An unknown type is passed through as a plain string rather than dropped:
+ * `Detection.entityType` is `EntityType | string` precisely so a client one
+ * release behind the service does not silently lose a whole category of PII.
+ */
+function toEntityType(raw) {
+    const canonical = canonicalType(raw);
+    const known = Object.values(EntityType).includes(canonical);
+    return known ? canonical : canonical;
+}
+function toResult(payload, text) {
+    const detections = (payload.entities ?? []).map(span => ({
+        entityType: toEntityType(span.type),
+        start: span.start,
+        end: span.end,
+        text: span.text,
+        source: span.source === "model" ? DetectionSource.CLOUD : DetectionSource.RULES,
+        country: null,
+        confidence: span.confidence ?? "high",
+    }));
+    detections.sort((a, b) => a.start - b.start || b.end - a.end);
+    return {
+        redactedText: payload.redacted_text ?? text,
+        detections,
+        source: "cloud",
+        degraded: false,
+        inferredCountries: [],
+        evidence: [],
+        detectionMode: "declared",
+    };
+}
+export class CloudClient {
+    config;
+    constructor(config) {
+        const resolved = config ?? getConfig();
+        if (resolved === null)
+            throw new NotConfiguredError();
+        this.config = resolved;
+    }
+    headers(idempotencyKey) {
+        return {
+            Authorization: `Bearer ${this.config.apiKey}`,
+            "Content-Type": "application/json",
+            // A retry after a timeout must not create a second job or a second usage
+            // row. The key is per logical document, not per attempt.
+            "Idempotency-Key": idempotencyKey,
+            ...this.config.headers,
+        };
+    }
+    raiseFor(status, payload) {
+        const message = payload.error ?? `HTTP ${status}`;
+        const detail = payload.detail ?? {};
+        if (status === 401)
+            throw new CloudError(`authentication failed: ${message}`, status);
+        if (status === 413)
+            throw new TooLargeError(message, status, detail);
+        if (status === 429)
+            throw new QuotaExceededError(message, status, detail);
+        throw new CloudError(message, status, detail);
+    }
+    async redact(text, options) {
+        const doFetch = requireFetch(options.fetchImpl);
+        const key = options.idempotencyKey ?? uuid();
+        const body = JSON.stringify({
+            text,
+            country: options.country.toUpperCase(),
+            language: options.language ?? "",
+            priority: options.priority ?? "interactive",
+            rules_only: options.rulesOnly ?? false,
+        });
+        let attempt = 0;
+        for (;;) {
+            let response;
+            let payload = {};
+            try {
+                response = await this.withTimeout(doFetch, `${this.config.baseUrl}/v1/redact`, {
+                    method: "POST",
+                    headers: this.headers(key),
+                    body,
+                });
+                payload = await readJson(response);
+            }
+            catch (e) {
+                if (attempt >= this.config.maxRetries) {
+                    throw new CloudError(`cloud request failed: ${e instanceof Error ? e.message : String(e)}`);
+                }
+            }
+            if (response) {
+                if (response.status === 202) {
+                    // Outlived the sync window. Location is where the answer will appear;
+                    // polling it is this client's job, not the caller's.
+                    const location = response.headers.get("Location") ??
+                        payload.location ??
+                        `/v1/jobs/${payload.job_id ?? ""}`;
+                    return toResult(await this.poll(doFetch, location, key), text);
+                }
+                if (response.status === 200)
+                    return toResult(payload, text);
+                if (!RETRY_STATUS.has(response.status) || attempt >= this.config.maxRetries) {
+                    this.raiseFor(response.status, payload);
+                }
+            }
+            await sleep(backoffMs(attempt, retryAfterMs(response?.headers)));
+            attempt++;
+        }
+    }
+    async poll(doFetch, location, key) {
+        const url = location.startsWith("http")
+            ? location
+            : `${this.config.baseUrl}${location}`;
+        const deadline = Date.now() + this.config.pollTimeoutMs;
+        let delay = 500;
+        for (;;) {
+            const response = await this.withTimeout(doFetch, url, {
+                method: "GET",
+                headers: this.headers(key),
+            });
+            const payload = await readJson(response);
+            if (response.status === 200 && payload.redacted_text !== undefined) {
+                return payload;
+            }
+            if (response.status !== 200 && !RETRY_STATUS.has(response.status)) {
+                this.raiseFor(response.status, payload);
+            }
+            if (Date.now() >= deadline) {
+                throw new CloudError(`document did not complete within ${Math.round(this.config.pollTimeoutMs / 1000)}s (${url})`);
+            }
+            const after = retryAfterMs(response.headers);
+            await sleep(after ?? delay);
+            delay = Math.min(delay * 1.5, 5000);
+        }
+    }
+    async withTimeout(doFetch, url, init) {
+        const controller = new AbortController();
+        const timer = setTimeout(() => controller.abort(), this.config.timeoutMs);
+        try {
+            return await doFetch(url, { ...init, signal: controller.signal });
+        }
+        finally {
+            clearTimeout(timer);
+        }
+    }
+}
+async function readJson(response) {
+    try {
+        const data = await response.json();
+        return data && typeof data === "object" ? data : {};
+    }
+    catch {
+        return {};
+    }
+}

package/dist/esm/cloud/config.d.ts

@@ -0,0 +1,43 @@
+/** [CLOUD EXTENSION] Cloud tier configuration. */
+export declare const DEFAULT_BASE_URL = "https://api.euredact.dev";
+/**
+ * How long to wait for a single document. The service holds a connection open
+ * for its sync window and then hands back a job handle, so this has to exceed
+ * that window or the client abandons requests the service is about to answer.
+ */
+export declare const DEFAULT_TIMEOUT_MS = 90000;
+/**
+ * Total wall-clock budget for a document that outlives the sync window and has
+ * to be polled. Cold-starting a GPU node and loading an ~18 GB checkpoint takes
+ * minutes, and an async job absorbs that invisibly — but not forever.
+ */
+export declare const DEFAULT_POLL_TIMEOUT_MS = 900000;
+export declare const DEFAULT_MAX_RETRIES = 3;
+export interface CloudConfig {
+    apiKey: string;
+    baseUrl: string;
+    timeoutMs: number;
+    pollTimeoutMs: number;
+    maxRetries: number;
+    /** Extra headers, for a proxy or a self-hosted deployment. */
+    headers: Record<string, string>;
+}
+export interface ConfigureOptions {
+    apiKey?: string;
+    baseUrl?: string;
+    timeoutMs?: number;
+    pollTimeoutMs?: number;
+    maxRetries?: number;
+    headers?: Record<string, string>;
+}
+/**
+ * Configure the cloud tier.
+ *
+ * Falls back to `EUREDACT_API_KEY` and `EUREDACT_BASE_URL` so a key never has
+ * to be written into source to get started.
+ */
+export declare function configure(options?: ConfigureOptions): CloudConfig;
+/** The active configuration, or null. Never throws. */
+export declare function getConfig(): CloudConfig | null;
+/** Forget the configuration. Mainly for tests. */
+export declare function reset(): void;

package/dist/esm/cloud/config.js

@@ -0,0 +1,51 @@
+/** [CLOUD EXTENSION] Cloud tier configuration. */
+export const DEFAULT_BASE_URL = "https://api.euredact.dev";
+/**
+ * How long to wait for a single document. The service holds a connection open
+ * for its sync window and then hands back a job handle, so this has to exceed
+ * that window or the client abandons requests the service is about to answer.
+ */
+export const DEFAULT_TIMEOUT_MS = 90_000;
+/**
+ * Total wall-clock budget for a document that outlives the sync window and has
+ * to be polled. Cold-starting a GPU node and loading an ~18 GB checkpoint takes
+ * minutes, and an async job absorbs that invisibly — but not forever.
+ */
+export const DEFAULT_POLL_TIMEOUT_MS = 900_000;
+export const DEFAULT_MAX_RETRIES = 3;
+let config = null;
+/** Read an env var where there is an environment to read. Browsers have none. */
+function env(name) {
+    const proc = globalThis.process;
+    return proc?.env?.[name];
+}
+/**
+ * Configure the cloud tier.
+ *
+ * Falls back to `EUREDACT_API_KEY` and `EUREDACT_BASE_URL` so a key never has
+ * to be written into source to get started.
+ */
+export function configure(options = {}) {
+    const apiKey = options.apiKey ?? env("EUREDACT_API_KEY") ?? "";
+    if (!apiKey) {
+        throw new Error("no API key: pass configure({ apiKey }) or set EUREDACT_API_KEY");
+    }
+    config = {
+        apiKey,
+        baseUrl: (options.baseUrl ?? env("EUREDACT_BASE_URL") ?? DEFAULT_BASE_URL)
+            .replace(/\/+$/, ""),
+        timeoutMs: options.timeoutMs ?? DEFAULT_TIMEOUT_MS,
+        pollTimeoutMs: options.pollTimeoutMs ?? DEFAULT_POLL_TIMEOUT_MS,
+        maxRetries: options.maxRetries ?? DEFAULT_MAX_RETRIES,
+        headers: { ...(options.headers ?? {}) },
+    };
+    return config;
+}
+/** The active configuration, or null. Never throws. */
+export function getConfig() {
+    return config;
+}
+/** Forget the configuration. Mainly for tests. */
+export function reset() {
+    config = null;
+}

package/dist/esm/cloud/errors.d.ts

@@ -0,0 +1,32 @@
+/** [CLOUD EXTENSION] Errors the cloud tier can raise. */
+export declare class CloudError extends Error {
+    readonly status?: number;
+    readonly detail: Record<string, unknown>;
+    constructor(message: string, status?: number, detail?: Record<string, unknown>);
+}
+/**
+ * Raised when the cloud tier is used without configuration.
+ *
+ * Deliberately an error rather than a quiet fallback to rules-only output.
+ * `mode: "cloud"` that silently returns rules-only results is the worst failure
+ * this library can have: the caller believes names, employers and diagnoses
+ * were checked, sees a plausible redacted document, and ships it with the PII
+ * still in it.
+ */
+export declare class NotConfiguredError extends CloudError {
+    constructor(message?: string);
+}
+/** 429 after retries were exhausted. */
+export declare class QuotaExceededError extends CloudError {
+    constructor(message: string, status?: number, detail?: Record<string, unknown>);
+}
+/**
+ * 413. The document is over the service's input cap.
+ *
+ * Permanent and not retryable. The model has never seen a chunk boundary, so
+ * the service refuses oversized input rather than splitting it and changing the
+ * accuracy story silently.
+ */
+export declare class TooLargeError extends CloudError {
+    constructor(message: string, status?: number, detail?: Record<string, unknown>);
+}

package/dist/esm/cloud/errors.js

@@ -0,0 +1,48 @@
+/** [CLOUD EXTENSION] Errors the cloud tier can raise. */
+export class CloudError extends Error {
+    status;
+    detail;
+    constructor(message, status, detail) {
+        super(message);
+        this.name = "CloudError";
+        this.status = status;
+        this.detail = detail ?? {};
+    }
+}
+/**
+ * Raised when the cloud tier is used without configuration.
+ *
+ * Deliberately an error rather than a quiet fallback to rules-only output.
+ * `mode: "cloud"` that silently returns rules-only results is the worst failure
+ * this library can have: the caller believes names, employers and diagnoses
+ * were checked, sees a plausible redacted document, and ships it with the PII
+ * still in it.
+ */
+export class NotConfiguredError extends CloudError {
+    constructor(message) {
+        super(message ??
+            "Cloud tier not configured. Call euredact.configure({ apiKey }) first, " +
+                "or set EUREDACT_API_KEY.");
+        this.name = "NotConfiguredError";
+    }
+}
+/** 429 after retries were exhausted. */
+export class QuotaExceededError extends CloudError {
+    constructor(message, status, detail) {
+        super(message, status, detail);
+        this.name = "QuotaExceededError";
+    }
+}
+/**
+ * 413. The document is over the service's input cap.
+ *
+ * Permanent and not retryable. The model has never seen a chunk boundary, so
+ * the service refuses oversized input rather than splitting it and changing the
+ * accuracy story silently.
+ */
+export class TooLargeError extends CloudError {
+    constructor(message, status, detail) {
+        super(message, status, detail);
+        this.name = "TooLargeError";
+    }
+}

package/dist/esm/cloud/index.d.ts

@@ -0,0 +1,19 @@
+/**
+ * [CLOUD EXTENSION] Cloud tier — rules engine plus a fine-tuned model.
+ *
+ * The rule engine catches what has a shape: IBANs, national IDs, phone
+ * numbers, anything with a checksum. It cannot catch what does not — a
+ * person's name, an employer, a diagnosis, a job title. Those are what the
+ * cloud tier adds, and the `EntityType` members marked `[CLOUD EXTENSION]` are
+ * exactly that gap.
+ *
+ * ```ts
+ * import { configure, redactAsync } from "euredact";
+ *
+ * configure({ apiKey: "erk_..." });
+ * const result = await redactAsync(text, { countries: ["BE"], mode: "cloud" });
+ * ```
+ */
+export { CloudClient, type CloudRedactOptions } from "./client.js";
+export { configure, getConfig, reset, DEFAULT_BASE_URL, type CloudConfig, type ConfigureOptions, } from "./config.js";
+export { CloudError, NotConfiguredError, QuotaExceededError, TooLargeError, } from "./errors.js";

package/dist/esm/cloud/index.js

@@ -0,0 +1,19 @@
+/**
+ * [CLOUD EXTENSION] Cloud tier — rules engine plus a fine-tuned model.
+ *
+ * The rule engine catches what has a shape: IBANs, national IDs, phone
+ * numbers, anything with a checksum. It cannot catch what does not — a
+ * person's name, an employer, a diagnosis, a job title. Those are what the
+ * cloud tier adds, and the `EntityType` members marked `[CLOUD EXTENSION]` are
+ * exactly that gap.
+ *
+ * ```ts
+ * import { configure, redactAsync } from "euredact";
+ *
+ * configure({ apiKey: "erk_..." });
+ * const result = await redactAsync(text, { countries: ["BE"], mode: "cloud" });
+ * ```
+ */
+export { CloudClient } from "./client.js";
+export { configure, getConfig, reset, DEFAULT_BASE_URL, } from "./config.js";
+export { CloudError, NotConfiguredError, QuotaExceededError, TooLargeError, } from "./errors.js";