@hazeljs/data 0.2.0-alpha.1

package/dist/pipelines/pipeline.builder.js

@@ -0,0 +1,212 @@
+"use strict";
+var __decorate = (this && this.__decorate) || function (decorators, target, key, desc) {
+    var c = arguments.length, r = c < 3 ? target : desc === null ? desc = Object.getOwnPropertyDescriptor(target, key) : desc, d;
+    if (typeof Reflect === "object" && typeof Reflect.decorate === "function") r = Reflect.decorate(decorators, target, key, desc);
+    else for (var i = decorators.length - 1; i >= 0; i--) if (d = decorators[i]) r = (c < 3 ? d(r) : c > 3 ? d(target, key, r) : d(target, key)) || r;
+    return c > 3 && r && Object.defineProperty(target, key, r), r;
+};
+var __metadata = (this && this.__metadata) || function (k, v) {
+    if (typeof Reflect === "object" && typeof Reflect.metadata === "function") return Reflect.metadata(k, v);
+};
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+var PipelineBuilder_1;
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.PipelineBuilder = void 0;
+const core_1 = require("@hazeljs/core");
+const core_2 = __importDefault(require("@hazeljs/core"));
+async function runWithRetry(fn, retry, _stepName) {
+    const { attempts, delay = 500, backoff = 'fixed' } = retry;
+    let lastError = new Error('Unknown');
+    for (let attempt = 1; attempt <= attempts; attempt++) {
+        try {
+            return await fn();
+        }
+        catch (err) {
+            lastError = err instanceof Error ? err : new Error(String(err));
+            if (attempt < attempts) {
+                const wait = backoff === 'exponential' ? delay * Math.pow(2, attempt - 1) : delay;
+                await new Promise((r) => setTimeout(r, wait));
+            }
+        }
+    }
+    throw lastError;
+}
+async function runWithTimeout(fn, ms, stepName) {
+    return new Promise((resolve, reject) => {
+        const id = setTimeout(() => reject(new Error(`Step "${stepName}" timed out after ${ms}ms`)), ms);
+        fn().then((v) => {
+            clearTimeout(id);
+            resolve(v);
+        }, (e) => {
+            clearTimeout(id);
+            reject(e);
+        });
+    });
+}
+/**
+ * PipelineBuilder — immutable, fluent DSL for building ETL pipelines programmatically.
+ *
+ * Each builder method returns a **new** instance — the original is never mutated.
+ *
+ * @example
+ * const pipeline = new PipelineBuilder()
+ *   .setName('orders')
+ *   .addTransform('normalize', (d) => ({ ...d, email: d.email.toLowerCase() }))
+ *   .when((d) => d.active, (b) => b.addTransform('enrich', enrich))
+ *   .addValidate('validate', validateFn);
+ *
+ * const result = await pipeline.execute(rawData);
+ */
+let PipelineBuilder = PipelineBuilder_1 = class PipelineBuilder {
+    constructor(name = 'unnamed-pipeline', steps = []) {
+        this._name = name;
+        this._steps = steps;
+    }
+    // ─── Identity ──────────────────────────────────────────────────────────────
+    setName(name) {
+        return new PipelineBuilder_1(name, [...this._steps]);
+    }
+    // ─── Steps ─────────────────────────────────────────────────────────────────
+    addTransform(name, transform, options = {}) {
+        return new PipelineBuilder_1(this._name, [...this._steps, { name, transform, ...options }]);
+    }
+    addValidate(name, validate, options = {}) {
+        return new PipelineBuilder_1(this._name, [...this._steps, { name, validate, ...options }]);
+    }
+    /**
+     * Run multiple transforms concurrently. Results are merged (Object.assign) into
+     * the current data if they are objects, otherwise replaced with an array of results.
+     */
+    parallel(name, transforms) {
+        return new PipelineBuilder_1(this._name, [...this._steps, { name, parallel: transforms }]);
+    }
+    /**
+     * Conditional branch: if `condition(data)` is true, run `thenBuilder` steps,
+     * otherwise run `elseBuilder` steps (default: identity).
+     */
+    branch(name, condition, thenBuilder, elseBuilder) {
+        const left = thenBuilder(new PipelineBuilder_1());
+        const right = elseBuilder ? elseBuilder(new PipelineBuilder_1()) : new PipelineBuilder_1();
+        return new PipelineBuilder_1(this._name, [
+            ...this._steps,
+            { name, branch: { condition, left, right } },
+        ]);
+    }
+    /**
+     * Attach a per-step error handler. If the previous step throws, `handler` is
+     * called with `(data, error)` and its return value becomes the new data.
+     */
+    catch(handler) {
+        if (this._steps.length === 0)
+            return this;
+        const steps = [...this._steps];
+        const last = { ...steps[steps.length - 1], catch: handler };
+        steps[steps.length - 1] = last;
+        return new PipelineBuilder_1(this._name, steps);
+    }
+    // ─── Execution ─────────────────────────────────────────────────────────────
+    async execute(input) {
+        let data = input;
+        for (let i = 0; i < this._steps.length; i++) {
+            const step = this._steps[i];
+            core_2.default.debug(`Pipeline "${this._name}": step ${i + 1} — ${step.name}`);
+            // Conditional skip
+            if (step.when && !step.when(data)) {
+                core_2.default.debug(`Step "${step.name}" skipped`);
+                continue;
+            }
+            const runStep = async () => {
+                if (step.branch) {
+                    const { condition, left, right } = step.branch;
+                    return condition(data) ? left.execute(data) : right.execute(data);
+                }
+                if (step.parallel && step.parallel.length > 0) {
+                    const results = await Promise.all(step.parallel.map((fn) => {
+                        const r = fn(data);
+                        return r instanceof Promise ? r : Promise.resolve(r);
+                    }));
+                    if (results.every((r) => r !== null && typeof r === 'object' && !Array.isArray(r))) {
+                        return Object.assign({}, data, ...results);
+                    }
+                    return results;
+                }
+                if (step.transform) {
+                    const r = step.transform(data);
+                    return r instanceof Promise ? await r : r;
+                }
+                if (step.validate) {
+                    return step.validate(data);
+                }
+                return data;
+            };
+            try {
+                let promise;
+                if (step.retry) {
+                    promise = runWithRetry(runStep, step.retry, step.name);
+                }
+                else {
+                    promise = runStep();
+                }
+                if (step.timeoutMs) {
+                    promise = runWithTimeout(() => promise, step.timeoutMs, step.name);
+                }
+                data = await promise;
+            }
+            catch (err) {
+                const error = err instanceof Error ? err : new Error(String(err));
+                if (step.dlq) {
+                    await Promise.resolve(step.dlq.handler(data, error, step.name));
+                }
+                else if (step.catch) {
+                    data = await Promise.resolve(step.catch(data, error));
+                }
+                else {
+                    throw error;
+                }
+            }
+        }
+        return data;
+    }
+    // ─── Serialization ─────────────────────────────────────────────────────────
+    /** Serialize the pipeline definition to a plain object (steps with functions are omitted). */
+    toSchema() {
+        return {
+            name: this._name,
+            steps: this._steps.map((s) => ({
+                name: s.name,
+                ...(s.when ? { conditional: true } : {}),
+                ...(s.parallel ? { parallel: true, parallelCount: s.parallel.length } : {}),
+                ...(s.branch ? { branch: true } : {}),
+                ...(s.retry ? { retry: s.retry } : {}),
+                ...(s.timeoutMs ? { timeoutMs: s.timeoutMs } : {}),
+            })),
+        };
+    }
+    build() {
+        return this.toSchema();
+    }
+    /** Create a fresh pipeline from a definition (transforms must be re-registered). */
+    static create(name) {
+        return new PipelineBuilder_1(name);
+    }
+    /**
+     * @deprecated Use `new PipelineBuilder()` directly. Kept for backward compat.
+     * Note: this instance is now immutable — reset() returns a new empty builder.
+     */
+    reset() {
+        return new PipelineBuilder_1(this._name);
+    }
+    get name() {
+        return this._name;
+    }
+    get steps() {
+        return this._steps;
+    }
+};
+exports.PipelineBuilder = PipelineBuilder;
+exports.PipelineBuilder = PipelineBuilder = PipelineBuilder_1 = __decorate([
+    (0, core_1.Service)(),
+    __metadata("design:paramtypes", [Object, Array])
+], PipelineBuilder);

package/dist/pipelines/pipeline.builder.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=pipeline.builder.test.d.ts.map

package/dist/pipelines/pipeline.builder.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"pipeline.builder.test.d.ts","sourceRoot":"","sources":["../../src/pipelines/pipeline.builder.test.ts"],"names":[],"mappings":""}

package/dist/pipelines/pipeline.builder.test.js

@@ -0,0 +1,185 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+const pipeline_builder_1 = require("./pipeline.builder");
+describe('PipelineBuilder', () => {
+    let builder;
+    beforeEach(() => {
+        builder = new pipeline_builder_1.PipelineBuilder();
+    });
+    it('adds transform and executes', async () => {
+        const pipeline = builder
+            .addTransform('step1', (d) => Object.assign({}, d, { a: 1 }))
+            .addTransform('step2', (d) => Object.assign({}, d, { b: 2 }));
+        const result = await pipeline.execute({ x: 0 });
+        expect(result).toEqual({ x: 0, a: 1, b: 2 });
+    });
+    it('handles async transforms', async () => {
+        const pipeline = builder.addTransform('async', async (d) => Object.assign({}, d, { done: true }));
+        const result = await pipeline.execute({});
+        expect(result.done).toBe(true);
+    });
+    it('setName sets pipeline name', () => {
+        const built = builder.setName('my-pipeline').build();
+        expect(built.name).toBe('my-pipeline');
+    });
+    it('build returns config', () => {
+        const config = builder
+            .addTransform('s1', (d) => d)
+            .addValidate('s2', (d) => d)
+            .build();
+        expect(config.steps).toHaveLength(2);
+        expect(config.name).toBe('unnamed-pipeline');
+    });
+    it('reset clears steps', async () => {
+        const withStep = builder.addTransform('s1', (d) => d);
+        const cleared = withStep.reset();
+        const config = cleared.build();
+        expect(config.steps).toHaveLength(0);
+    });
+    it('is immutable — original is unmodified after chaining', () => {
+        builder.addTransform('step1', (d) => d);
+        expect(builder.steps).toHaveLength(0); // original unchanged
+    });
+    it('branch runs left path when condition is true', async () => {
+        const pipeline = builder.branch('classify', (d) => d.type === 'a', (b) => b.addTransform('enrichA', (d) => Object.assign({}, d, { enriched: 'A' })), (b) => b.addTransform('enrichB', (d) => Object.assign({}, d, { enriched: 'B' })));
+        const resultA = await pipeline.execute({ type: 'a' });
+        expect(resultA.enriched).toBe('A');
+        const resultB = await pipeline.execute({ type: 'b' });
+        expect(resultB.enriched).toBe('B');
+    });
+    it('parallel runs transforms concurrently and merges results', async () => {
+        const pipeline = builder.parallel('enrich', [
+            (d) => Object.assign({}, d, { a: 1 }),
+            (d) => Object.assign({}, d, { b: 2 }),
+        ]);
+        const result = await pipeline.execute({});
+        expect(result.a).toBe(1);
+        expect(result.b).toBe(2);
+    });
+    it('catch handles step errors', async () => {
+        const pipeline = builder
+            .addTransform('fail', () => {
+            throw new Error('step failed');
+        })
+            .catch((_data, _err) => ({ recovered: true }));
+        const result = await pipeline.execute({});
+        expect(result.recovered).toBe(true);
+    });
+    it('toSchema returns serializable definition', () => {
+        const schema = builder
+            .addTransform('step1', (d) => d)
+            .addValidate('step2', (d) => d)
+            .toSchema();
+        expect(schema.name).toBe('unnamed-pipeline');
+        expect(schema.steps).toHaveLength(2);
+        expect(schema.steps[0].name).toBe('step1');
+    });
+    it('retry retries failed steps', async () => {
+        let attempts = 0;
+        const pipeline = builder.addTransform('flaky', () => {
+            attempts++;
+            if (attempts < 3)
+                throw new Error('transient');
+            return { ok: true };
+        }, { retry: { attempts: 3, delay: 0 } });
+        const result = await pipeline.execute({});
+        expect(result.ok).toBe(true);
+        expect(attempts).toBe(3);
+    });
+    it('addTransform with when option skips when false', async () => {
+        const pipeline = builder.addTransform('cond', (d) => Object.assign({}, d, { ran: true }), {
+            when: (d) => d.run === true,
+        });
+        const result = await pipeline.execute({ run: false });
+        expect(result.ran).toBeUndefined();
+        const result2 = await pipeline.execute({ run: true });
+        expect(result2.ran).toBe(true);
+    });
+    it('addValidate with when option', async () => {
+        const pipeline = builder.addValidate('v', (d) => d, { when: (d) => d.ok });
+        const result = await pipeline.execute({ ok: true });
+        expect(result).toEqual({ ok: true });
+    });
+    it('PipelineBuilder.create returns new instance', () => {
+        const b = pipeline_builder_1.PipelineBuilder.create('test');
+        expect(b.name).toBe('test');
+        expect(b.steps).toHaveLength(0);
+    });
+    it('conditional step is skipped when predicate returns false', async () => {
+        let ran = false;
+        const pipeline = builder.addTransform('conditional', (d) => {
+            ran = true;
+            return d;
+        }, { when: () => false });
+        await pipeline.execute({});
+        expect(ran).toBe(false);
+    });
+    it('parallel returns array when results are not all objects', async () => {
+        const pipeline = builder.parallel('mixed', [() => 1, () => 2]);
+        const result = await pipeline.execute({});
+        expect(result).toEqual([1, 2]);
+    });
+    it('parallel returns array when some results are null', async () => {
+        const pipeline = builder.parallel('withNull', [
+            (d) => Object.assign({}, d, { a: 1 }),
+            () => null,
+        ]);
+        const result = await pipeline.execute({});
+        expect(Array.isArray(result)).toBe(true);
+        expect(result).toHaveLength(2);
+    });
+    it('branch without elseBuilder uses identity for right path', async () => {
+        const pipeline = builder.branch('cond', (d) => d.flag, (b) => b.addTransform('left', (d) => Object.assign({}, d, { side: 'left' })));
+        const left = await pipeline.execute({ flag: true });
+        expect(left.side).toBe('left');
+        const right = await pipeline.execute({ flag: false });
+        expect(right).toEqual({ flag: false });
+    });
+    it('catch on empty steps returns this', () => {
+        const result = builder.catch(() => ({}));
+        expect(result).toBe(builder);
+        expect(result.steps).toHaveLength(0);
+    });
+    it('timeoutMs rejects when step exceeds timeout', async () => {
+        const pipeline = builder.addTransform('slow', () => new Promise((r) => setTimeout(() => r({}), 200)), { timeoutMs: 10 });
+        await expect(pipeline.execute({})).rejects.toThrow('timed out');
+    });
+    it('dlq handler is called on step failure', async () => {
+        const dlqItems = [];
+        const pipeline = builder.addTransform('fail', () => {
+            throw new Error('fail');
+        }, {
+            dlq: {
+                handler: (data, err, step) => {
+                    dlqItems.push({ data, err: err.message, step });
+                },
+            },
+        });
+        const result = await pipeline.execute({});
+        expect(result).toEqual({});
+        expect(dlqItems).toHaveLength(1);
+        expect(dlqItems[0]).toMatchObject({ step: 'fail', err: 'fail' });
+    });
+    it('validate step runs when no transform', async () => {
+        const pipeline = builder.addValidate('v', (d) => d);
+        const result = await pipeline.execute({ x: 1 });
+        expect(result).toEqual({ x: 1 });
+    });
+    it('step with no transform or validate returns data', async () => {
+        const pipeline = builder.addTransform('id', (d) => d);
+        const result = await pipeline.execute({ a: 1 });
+        expect(result).toEqual({ a: 1 });
+    });
+    it('retry with exponential backoff', async () => {
+        let attempts = 0;
+        const pipeline = builder.addTransform('flaky', () => {
+            attempts++;
+            if (attempts < 2)
+                throw new Error('retry');
+            return { ok: true };
+        }, { retry: { attempts: 2, delay: 1, backoff: 'exponential' } });
+        const result = await pipeline.execute({});
+        expect(result.ok).toBe(true);
+        expect(attempts).toBe(2);
+    });
+});

package/dist/pipelines/stream.service.d.ts

@@ -0,0 +1,12 @@
+import { ETLService } from './etl.service';
+/**
+ * Stream Service - Streaming pipeline execution
+ * Processes data through pipeline steps (for in-process streaming, not Flink deployment)
+ */
+export declare class StreamService {
+    private readonly etlService;
+    constructor(etlService: ETLService);
+    processStream<T = unknown>(pipelineInstance: object, source: AsyncIterable<unknown>): AsyncGenerator<T>;
+    processBatch<T = unknown>(pipelineInstance: object, items: unknown[]): Promise<T[]>;
+}
+//# sourceMappingURL=stream.service.d.ts.map

package/dist/pipelines/stream.service.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"stream.service.d.ts","sourceRoot":"","sources":["../../src/pipelines/stream.service.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,UAAU,EAAE,MAAM,eAAe,CAAC;AAI3C;;;GAGG;AACH,qBACa,aAAa;IACZ,OAAO,CAAC,QAAQ,CAAC,UAAU;gBAAV,UAAU,EAAE,UAAU;IAE5C,aAAa,CAAC,CAAC,GAAG,OAAO,EAC9B,gBAAgB,EAAE,MAAM,EACxB,MAAM,EAAE,aAAa,CAAC,OAAO,CAAC,GAC7B,cAAc,CAAC,CAAC,CAAC;IAmBd,YAAY,CAAC,CAAC,GAAG,OAAO,EAAE,gBAAgB,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,GAAG,OAAO,CAAC,CAAC,EAAE,CAAC;CAQ1F"}

package/dist/pipelines/stream.service.js

@@ -0,0 +1,58 @@
+"use strict";
+var __decorate = (this && this.__decorate) || function (decorators, target, key, desc) {
+    var c = arguments.length, r = c < 3 ? target : desc === null ? desc = Object.getOwnPropertyDescriptor(target, key) : desc, d;
+    if (typeof Reflect === "object" && typeof Reflect.decorate === "function") r = Reflect.decorate(decorators, target, key, desc);
+    else for (var i = decorators.length - 1; i >= 0; i--) if (d = decorators[i]) r = (c < 3 ? d(r) : c > 3 ? d(target, key, r) : d(target, key)) || r;
+    return c > 3 && r && Object.defineProperty(target, key, r), r;
+};
+var __metadata = (this && this.__metadata) || function (k, v) {
+    if (typeof Reflect === "object" && typeof Reflect.metadata === "function") return Reflect.metadata(k, v);
+};
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.StreamService = void 0;
+const core_1 = require("@hazeljs/core");
+const etl_service_1 = require("./etl.service");
+const decorators_1 = require("../decorators");
+const core_2 = __importDefault(require("@hazeljs/core"));
+/**
+ * Stream Service - Streaming pipeline execution
+ * Processes data through pipeline steps (for in-process streaming, not Flink deployment)
+ */
+let StreamService = class StreamService {
+    constructor(etlService) {
+        this.etlService = etlService;
+    }
+    async *processStream(pipelineInstance, source) {
+        const metadata = (0, decorators_1.getStreamMetadata)(pipelineInstance.constructor);
+        if (!metadata) {
+            throw new Error('Pipeline is not decorated with @Stream');
+        }
+        core_2.default.debug(`Processing stream ${metadata.name}`);
+        for await (const item of source) {
+            try {
+                const result = await this.etlService.execute(pipelineInstance, item);
+                yield result;
+            }
+            catch (error) {
+                core_2.default.error(`Stream ${metadata.name} error processing item:`, error);
+                throw error;
+            }
+        }
+    }
+    async processBatch(pipelineInstance, items) {
+        const results = [];
+        for (const item of items) {
+            const result = await this.etlService.execute(pipelineInstance, item);
+            results.push(result);
+        }
+        return results;
+    }
+};
+exports.StreamService = StreamService;
+exports.StreamService = StreamService = __decorate([
+    (0, core_1.Service)(),
+    __metadata("design:paramtypes", [etl_service_1.ETLService])
+], StreamService);

package/dist/pipelines/stream.service.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=stream.service.test.d.ts.map

package/dist/pipelines/stream.service.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"stream.service.test.d.ts","sourceRoot":"","sources":["../../src/pipelines/stream.service.test.ts"],"names":[],"mappings":""}

package/dist/pipelines/stream.service.test.js

@@ -0,0 +1,103 @@
+"use strict";
+var __decorate = (this && this.__decorate) || function (decorators, target, key, desc) {
+    var c = arguments.length, r = c < 3 ? target : desc === null ? desc = Object.getOwnPropertyDescriptor(target, key) : desc, d;
+    if (typeof Reflect === "object" && typeof Reflect.decorate === "function") r = Reflect.decorate(decorators, target, key, desc);
+    else for (var i = decorators.length - 1; i >= 0; i--) if (d = decorators[i]) r = (c < 3 ? d(r) : c > 3 ? d(target, key, r) : d(target, key)) || r;
+    return c > 3 && r && Object.defineProperty(target, key, r), r;
+};
+var __metadata = (this && this.__metadata) || function (k, v) {
+    if (typeof Reflect === "object" && typeof Reflect.metadata === "function") return Reflect.metadata(k, v);
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+const etl_service_1 = require("./etl.service");
+const stream_service_1 = require("./stream.service");
+const schema_validator_1 = require("../validators/schema.validator");
+const decorators_1 = require("../decorators");
+let BatchTestPipeline = class BatchTestPipeline {
+    inc(data) {
+        return { v: (data.v || 0) + 1 };
+    }
+};
+__decorate([
+    (0, decorators_1.Transform)({ step: 1, name: 'inc' }),
+    __metadata("design:type", Function),
+    __metadata("design:paramtypes", [Object]),
+    __metadata("design:returntype", void 0)
+], BatchTestPipeline.prototype, "inc", null);
+BatchTestPipeline = __decorate([
+    (0, decorators_1.Pipeline)('batch-test'),
+    (0, decorators_1.Stream)({ name: 'test', source: 'kafka://in', sink: 'kafka://out' })
+], BatchTestPipeline);
+describe('StreamService', () => {
+    let streamService;
+    beforeEach(() => {
+        const etlService = new etl_service_1.ETLService(new schema_validator_1.SchemaValidator());
+        streamService = new stream_service_1.StreamService(etlService);
+    });
+    it('processes batch of items', async () => {
+        const pipeline = new BatchTestPipeline();
+        const results = await streamService.processBatch(pipeline, [{ v: 0 }, { v: 1 }]);
+        expect(results).toEqual([{ v: 1 }, { v: 2 }]);
+    });
+    it('processStream yields results', async () => {
+        const pipeline = new BatchTestPipeline();
+        async function* source() {
+            yield { v: 0 };
+            yield { v: 5 };
+        }
+        const results = [];
+        for await (const r of streamService.processStream(pipeline, source())) {
+            results.push(r);
+        }
+        expect(results).toEqual([{ v: 1 }, { v: 6 }]);
+    });
+    it('processStream throws when pipeline not @Stream decorated', async () => {
+        let NoStreamPipeline = class NoStreamPipeline {
+            x(d) {
+                return d;
+            }
+        };
+        __decorate([
+            (0, decorators_1.Transform)({ step: 1, name: 'x' }),
+            __metadata("design:type", Function),
+            __metadata("design:paramtypes", [Object]),
+            __metadata("design:returntype", void 0)
+        ], NoStreamPipeline.prototype, "x", null);
+        NoStreamPipeline = __decorate([
+            (0, decorators_1.Pipeline)('no-stream')
+        ], NoStreamPipeline);
+        async function* source() {
+            yield {};
+        }
+        await expect((async () => {
+            for await (const _ of streamService.processStream(new NoStreamPipeline(), source())) {
+                break;
+            }
+        })()).rejects.toThrow('not decorated with @Stream');
+    });
+    it('processStream throws when item processing fails', async () => {
+        let FailStreamPipeline = class FailStreamPipeline {
+            fail() {
+                throw new Error('Item failed');
+            }
+        };
+        __decorate([
+            (0, decorators_1.Transform)({ step: 1, name: 'fail' }),
+            __metadata("design:type", Function),
+            __metadata("design:paramtypes", []),
+            __metadata("design:returntype", void 0)
+        ], FailStreamPipeline.prototype, "fail", null);
+        FailStreamPipeline = __decorate([
+            (0, decorators_1.Pipeline)('fail-stream'),
+            (0, decorators_1.Stream)({ name: 'fail', source: 'kafka://in', sink: 'kafka://out' })
+        ], FailStreamPipeline);
+        async function* source() {
+            yield {};
+        }
+        await expect((async () => {
+            for await (const _ of streamService.processStream(new FailStreamPipeline(), source())) {
+                break;
+            }
+        })()).rejects.toThrow('Item failed');
+    });
+});

package/dist/quality/quality.service.d.ts

@@ -0,0 +1,87 @@
+export interface QualityCheckResult {
+    name: string;
+    passed: boolean;
+    score?: number;
+    message?: string;
+    details?: Record<string, unknown>;
+}
+export interface DataQualityReport {
+    timestamp: Date;
+    dataset: string;
+    totalRows: number;
+    checks: QualityCheckResult[];
+    passed: boolean;
+    /** Composite quality score 0–100 (average of individual check scores). */
+    score: number;
+}
+export interface FieldProfile {
+    count: number;
+    nullCount: number;
+    nullPct: number;
+    uniqueCount: number;
+    cardinality: number;
+    min?: number | string;
+    max?: number | string;
+    mean?: number;
+    stddev?: number;
+    topValues: Array<{
+        value: unknown;
+        count: number;
+    }>;
+}
+export interface DataProfile {
+    dataset: string;
+    totalRows: number;
+    fields: Record<string, FieldProfile>;
+    generatedAt: Date;
+}
+export interface AnomalyResult {
+    field: string;
+    rowIndex: number;
+    value: unknown;
+    zScore: number;
+    message: string;
+}
+export type CheckFn = (data: unknown) => QualityCheckResult | Promise<QualityCheckResult>;
+type SyncCheckFn = (data: unknown) => QualityCheckResult;
+/**
+ * Quality Service — data quality checks, profiling, and anomaly detection.
+ *
+ * Built-in check factories:
+ * - completeness(fields[])
+ * - notNull(fields[])
+ * - uniqueness(fields[])
+ * - range(field, { min, max })
+ * - pattern(field, regex, message?)
+ * - referentialIntegrity(field, allowedValues[])
+ *
+ * Profiling:
+ * - profile(dataset, records[]) → DataProfile
+ *
+ * Anomaly detection:
+ * - detectAnomalies(records[], fields[], threshold?) → AnomalyResult[]
+ */
+export declare class QualityService {
+    private checks;
+    registerCheck(name: string, check: CheckFn): void;
+    runChecks(dataset: string, data: unknown): Promise<DataQualityReport>;
+    completeness(requiredFields: string[]): SyncCheckFn;
+    notNull(fields: string[]): SyncCheckFn;
+    uniqueness(fields: string[]): SyncCheckFn;
+    range(field: string, options: {
+        min?: number;
+        max?: number;
+    }): SyncCheckFn;
+    pattern(field: string, regex: RegExp, message?: string): SyncCheckFn;
+    referentialIntegrity(field: string, allowedValues: unknown[]): SyncCheckFn;
+    profile(dataset: string, records: Record<string, unknown>[]): DataProfile;
+    /**
+     * Detect statistical anomalies using Z-score.
+     * @param records Dataset rows
+     * @param fields  Numeric fields to analyze
+     * @param threshold Z-score threshold (default: 3.0)
+     */
+    detectAnomalies(records: Record<string, unknown>[], fields: string[], threshold?: number): AnomalyResult[];
+}
+export {};
+//# sourceMappingURL=quality.service.d.ts.map