euredact 0.0.1 → 0.2.0

package/dist/esm/sdk.js

@@ -0,0 +1,76 @@
+import { ResultCache } from "./cache.js";
+import { normalize, mapOffsetToOriginal } from "./normalizer.js";
+import { RuleEngine } from "./rules/engine.js";
+import { EntityType } from "./types.js";
+const DATE_TYPES = new Set([EntityType.DOB, EntityType.DATE_OF_DEATH]);
+class ReferentialMapper {
+    counters = new Map();
+    mapping = new Map();
+    getLabel(text, entityType) {
+        if (!this.mapping.has(text)) {
+            const count = (this.counters.get(entityType) ?? 0) + 1;
+            this.counters.set(entityType, count);
+            this.mapping.set(text, `${entityType}_${count}`);
+        }
+        return this.mapping.get(text);
+    }
+}
+export class EuRedact {
+    engine = new RuleEngine();
+    cache = new ResultCache();
+    referentialMapper = new ReferentialMapper();
+    addCustomPattern(name, pattern) {
+        this.engine.addCustomPattern(name, pattern);
+        this.cache.clear();
+    }
+    redact(text, options = {}) {
+        const { countries = null, mode = "rules", referentialIntegrity = false, detectDates = false, cache = true, } = options;
+        const [normalizedText, offsetMapping] = normalize(text);
+        const countriesTuple = countries
+            ? countries.map(c => c.toUpperCase()).sort()
+            : ["ALL"];
+        const cacheMode = `${mode}|dates=${detectDates}`;
+        let cacheKey;
+        if (cache) {
+            cacheKey = this.cache.key(normalizedText, countriesTuple, cacheMode);
+            const cached = this.cache.get(cacheKey);
+            if (cached !== null)
+                return cached;
+        }
+        let detections = this.engine.detect(normalizedText, countries);
+        if (offsetMapping !== null) {
+            detections = detections.map(d => ({
+                ...d,
+                start: mapOffsetToOriginal(d.start, offsetMapping),
+                end: mapOffsetToOriginal(d.end, offsetMapping),
+            }));
+        }
+        if (!detectDates) {
+            detections = detections.filter(d => !DATE_TYPES.has(d.entityType));
+        }
+        detections.sort((a, b) => a.start - b.start || b.end - a.end);
+        let redacted = text;
+        for (let i = detections.length - 1; i >= 0; i--) {
+            const det = detections[i];
+            const replacement = referentialIntegrity
+                ? this.referentialMapper.getLabel(det.text, det.entityType)
+                : `[${det.entityType}]`;
+            redacted = redacted.slice(0, det.start) + replacement + redacted.slice(det.end);
+        }
+        const result = {
+            redactedText: redacted,
+            detections,
+            source: "rules",
+            degraded: false,
+        };
+        if (cache && cacheKey) {
+            this.cache.put(cacheKey, result);
+        }
+        return result;
+    }
+    redactBatch(texts, options = {}) {
+        const countries = options.countries ?? null;
+        this.engine.loadCountries(countries ? countries.map(c => c.toUpperCase()) : null);
+        return texts.map(text => this.redact(text, options));
+    }
+}

package/dist/esm/types.d.ts

@@ -0,0 +1,65 @@
+export declare enum EntityType {
+    NAME = "NAME",
+    ADDRESS = "ADDRESS",
+    IBAN = "IBAN",
+    BIC = "BIC",
+    CREDIT_CARD = "CREDIT_CARD",
+    PHONE = "PHONE",
+    EMAIL = "EMAIL",
+    DOB = "DOB",
+    DATE_OF_DEATH = "DATE_OF_DEATH",
+    NATIONAL_ID = "NATIONAL_ID",
+    SSN = "SSN",
+    TAX_ID = "TAX_ID",
+    PASSPORT = "PASSPORT",
+    DRIVERS_LICENSE = "DRIVERS_LICENSE",
+    RESIDENCE_PERMIT = "RESIDENCE_PERMIT",
+    LICENSE_PLATE = "LICENSE_PLATE",
+    VIN = "VIN",
+    VAT = "VAT",
+    POSTAL_CODE = "POSTAL_CODE",
+    IP_ADDRESS = "IP_ADDRESS",
+    IPV6_ADDRESS = "IPV6_ADDRESS",
+    MAC_ADDRESS = "MAC_ADDRESS",
+    HEALTH_INSURANCE = "HEALTH_INSURANCE",
+    HEALTHCARE_PROVIDER = "HEALTHCARE_PROVIDER",
+    CHAMBER_OF_COMMERCE = "CHAMBER_OF_COMMERCE",
+    IMEI = "IMEI",
+    GPS_COORDINATES = "GPS_COORDINATES",
+    UUID = "UUID",
+    SOCIAL_HANDLE = "SOCIAL_HANDLE",
+    SECRET = "SECRET",
+    OTHER = "OTHER"
+}
+export declare enum DetectionSource {
+    RULES = "rules",
+    CLOUD = "cloud"
+}
+export interface Detection {
+    entityType: EntityType | string;
+    start: number;
+    end: number;
+    text: string;
+    source: DetectionSource;
+    country: string | null;
+    confidence: string;
+}
+export interface RedactResult {
+    redactedText: string;
+    detections: Detection[];
+    source: string;
+    degraded: boolean;
+}
+export interface PatternDef {
+    entityType: EntityType | string;
+    pattern: string;
+    validator: string | null;
+    description: string;
+    contextKeywords: string[];
+    requiresContext: boolean;
+}
+export interface CountryConfig {
+    code: string;
+    name: string;
+    patterns: PatternDef[];
+}

package/dist/esm/types.js

@@ -0,0 +1,39 @@
+export var EntityType;
+(function (EntityType) {
+    EntityType["NAME"] = "NAME";
+    EntityType["ADDRESS"] = "ADDRESS";
+    EntityType["IBAN"] = "IBAN";
+    EntityType["BIC"] = "BIC";
+    EntityType["CREDIT_CARD"] = "CREDIT_CARD";
+    EntityType["PHONE"] = "PHONE";
+    EntityType["EMAIL"] = "EMAIL";
+    EntityType["DOB"] = "DOB";
+    EntityType["DATE_OF_DEATH"] = "DATE_OF_DEATH";
+    EntityType["NATIONAL_ID"] = "NATIONAL_ID";
+    EntityType["SSN"] = "SSN";
+    EntityType["TAX_ID"] = "TAX_ID";
+    EntityType["PASSPORT"] = "PASSPORT";
+    EntityType["DRIVERS_LICENSE"] = "DRIVERS_LICENSE";
+    EntityType["RESIDENCE_PERMIT"] = "RESIDENCE_PERMIT";
+    EntityType["LICENSE_PLATE"] = "LICENSE_PLATE";
+    EntityType["VIN"] = "VIN";
+    EntityType["VAT"] = "VAT";
+    EntityType["POSTAL_CODE"] = "POSTAL_CODE";
+    EntityType["IP_ADDRESS"] = "IP_ADDRESS";
+    EntityType["IPV6_ADDRESS"] = "IPV6_ADDRESS";
+    EntityType["MAC_ADDRESS"] = "MAC_ADDRESS";
+    EntityType["HEALTH_INSURANCE"] = "HEALTH_INSURANCE";
+    EntityType["HEALTHCARE_PROVIDER"] = "HEALTHCARE_PROVIDER";
+    EntityType["CHAMBER_OF_COMMERCE"] = "CHAMBER_OF_COMMERCE";
+    EntityType["IMEI"] = "IMEI";
+    EntityType["GPS_COORDINATES"] = "GPS_COORDINATES";
+    EntityType["UUID"] = "UUID";
+    EntityType["SOCIAL_HANDLE"] = "SOCIAL_HANDLE";
+    EntityType["SECRET"] = "SECRET";
+    EntityType["OTHER"] = "OTHER";
+})(EntityType || (EntityType = {}));
+export var DetectionSource;
+(function (DetectionSource) {
+    DetectionSource["RULES"] = "rules";
+    DetectionSource["CLOUD"] = "cloud";
+})(DetectionSource || (DetectionSource = {}));

package/package.json

@@ -1,17 +1,63 @@
 {
   "name": "euredact",
-  "version": "0.0.1",
-  "description": "European PII redaction SDK — rule engine + cloud tier",
-  "main": "index.js",
-  "keywords": ["pii", "redaction", "gdpr", "privacy", "europe", "nlp"],
-  "author": "Joren Janssens <joren.janssens@proton.me>",
+  "version": "0.2.0",
+  "description": "European PII redaction SDK -- fast, zero-dependency, rule-based detection and redaction of personal data across 31 European countries",
   "license": "Apache-2.0",
-  "homepage": "https://euredact.eu",
+  "type": "module",
+  "main": "./dist/cjs/index.js",
+  "module": "./dist/esm/index.js",
+  "types": "./dist/esm/index.d.ts",
+  "exports": {
+    ".": {
+      "import": {
+        "types": "./dist/esm/index.d.ts",
+        "default": "./dist/esm/index.js"
+      },
+      "require": {
+        "types": "./dist/esm/index.d.ts",
+        "default": "./dist/cjs/index.js"
+      }
+    }
+  },
+  "files": [
+    "dist",
+    "README.md",
+    "LICENSE"
+  ],
+  "engines": {
+    "node": ">=16.0.0"
+  },
+  "keywords": [
+    "pii",
+    "redaction",
+    "gdpr",
+    "privacy",
+    "europe",
+    "eu",
+    "iban",
+    "bsn",
+    "national-id",
+    "phone",
+    "email",
+    "anonymization",
+    "data-protection",
+    "personal-data",
+    "regex"
+  ],
   "repository": {
     "type": "git",
-    "url": "https://github.com/euredact/euredact.git"
+    "url": "https://github.com/euredact/euredact"
+  },
+  "scripts": {
+    "build:esm": "tsc",
+    "build:cjs": "tsc -p tsconfig.cjs.json",
+    "postbuild:cjs": "echo '{\"type\":\"commonjs\"}' > dist/cjs/package.json",
+    "build": "rm -rf dist && npm run build:esm && npm run build:cjs",
+    "test": "node --import tsx src/__tests__/run.ts",
+    "prepublishOnly": "npm run build"
   },
-  "bugs": {
-    "url": "https://github.com/euredact/euredact/issues"
+  "devDependencies": {
+    "typescript": "^5.4.0",
+    "tsx": "^4.7.0"
   }
 }

package/index.js

@@ -1,3 +0,0 @@
-// EUREdact — European PII redaction SDK
-// TypeScript/Node SDK coming soon. See https://euredact.eu
-throw new Error("euredact: Node.js SDK not yet available. Use the Python SDK: pip install euredact");