stream: add forEach method

Add a `forEach` method to readable streams to enable concurrent
iteration and align with the iterator-helpers proposal.

Co-Authored-By: Robert Nagy <ronagy@icloud.com>
PR-URL: #41445

Author: benjamingr

doc/api/stream.md

@@ -1830,6 +1830,65 @@ for await (const result of dnsResults) {
 }
 ```
 
+### `readable.forEach(fn[, options])`
+
+<!-- YAML
+added: REPLACEME
+-->
+
+> Stability: 1 - Experimental
+
+* `fn` {Function|AsyncFunction} a function to call on each item of the stream.
+  * `data` {any} a chunk of data from the stream.
+  * `options` {Object}
+    * `signal` {AbortSignal} aborted if the stream is destroyed allowing to
+      abort the `fn` call early.
+* `options` {Object}
+  * `concurrency` {number} the maximal concurrent invocation of `fn` to call
+    on the stream at once. **Default:** `1`.
+  * `signal` {AbortSignal} allows destroying the stream if the signal is
+    aborted.
+* Returns: {Promise} a promise for when the stream has finished.
+
+This method allows iterating a stream. For each item in the stream the
+`fn` function will be called. If the `fn` function returns a promise - that
+promise will be `await`ed.
+
+This method is different from `for... await` loops in that it supports setting
+the maximal concurrent invocation of `fn` through the `concurrency` option. It
+is also possible to `break` from a `for... await` destroying the stream but
+`forEach` is only breakable by passing it a `signal` and aborting the related
+`AbortController`.
+
+This method is different from listening to the [`'data'`][] event in that it
+uses the [`readable`][] event in the underlying machinary and can limit the
+number of concurrent `fn` calls.
+
+```mjs
+import { Readable } from 'stream';
+import { Resolver } from 'dns/promises';
+
+// With a synchronous predicate.
+for await (const item of Readable.from([1, 2, 3, 4]).filter((x) => x > 2)) {
+  console.log(item); // 3, 4
+}
+// With an asynchronous predicate, making at most 2 queries at a time.
+const resolver = new Resolver();
+const dnsResults = await Readable.from([
+  'nodejs.org',
+  'openjsf.org',
+  'www.linuxfoundation.org',
+]).map(async (domain) => {
+  const { address } = await resolver.resolve4(domain, { ttl: true });
+  return address;
+}, { concurrency: 2 });
+await dnsResults.forEach((result) => {
+  // Logs result, similar to `for await (const result of dnsResults)`
+  console.log(result);
+});
+console.log('done'); // Stream has finished
+```
+
 ### Duplex and transform streams
 
 #### Class: `stream.Duplex`

lib/internal/streams/operators.js

@@ -147,10 +147,23 @@ async function * map(fn, options) {
   }
 }
 
+async function forEach(fn, options) {
+  if (typeof fn !== 'function') {
+    throw new ERR_INVALID_ARG_TYPE(
+      'fn', ['Function', 'AsyncFunction'], this);
+  }
+  async function forEachFn(value, options) {
+    await fn(value, options);
+    return kEmpty;
+  }
+  // eslint-disable-next-line no-unused-vars
+  for await (const unused of this.map(forEachFn, options));
+}
+
 async function * filter(fn, options) {
   if (typeof fn !== 'function') {
-    throw (new ERR_INVALID_ARG_TYPE(
-      'fn', ['Function', 'AsyncFunction'], this));
+    throw new ERR_INVALID_ARG_TYPE(
+      'fn', ['Function', 'AsyncFunction'], this);
   }
   async function filterFn(value, options) {
     if (await fn(value, options)) {
@@ -160,7 +173,9 @@ async function * filter(fn, options) {
   }
   yield* this.map(filterFn, options);
 }
+
 module.exports = {
+  filter,
+  forEach,
   map,
-  filter
 };

test/parallel/test-stream-forEach.js

@@ -0,0 +1,86 @@
+'use strict';
+
+const common = require('../common');
+const {
+  Readable,
+} = require('stream');
+const assert = require('assert');
+const { setTimeout } = require('timers/promises');
+
+{
+  // forEach works on synchronous streams with a synchronous predicate
+  const stream = Readable.from([1, 2, 3]);
+  const result = [1, 2, 3];
+  (async () => {
+    await stream.forEach((value) => assert.strictEqual(value, result.shift()));
+  })().then(common.mustCall());
+}
+
+{
+  // forEach works an asynchronous streams
+  const stream = Readable.from([1, 2, 3]).filter(async (x) => {
+    await Promise.resolve();
+    return true;
+  });
+  const result = [1, 2, 3];
+  (async () => {
+    await stream.forEach((value) => assert.strictEqual(value, result.shift()));
+  })().then(common.mustCall());
+}
+
+{
+  // forEach works on asynchronous streams with a asynchronous forEach fn
+  const stream = Readable.from([1, 2, 3]).filter(async (x) => {
+    await Promise.resolve();
+    return true;
+  });
+  const result = [1, 2, 3];
+  (async () => {
+    await stream.forEach(async (value) => {
+      await Promise.resolve();
+      assert.strictEqual(value, result.shift());
+    });
+  })().then(common.mustCall());
+}
+
+{
+  // Concurrency + AbortSignal
+  const ac = new AbortController();
+  let calls = 0;
+  const forEachPromise =
+    Readable.from([1, 2, 3, 4]).forEach(async (_, { signal }) => {
+      calls++;
+      await setTimeout(100, { signal });
+    }, { signal: ac.signal, concurrency: 2 });
+  // pump
+  assert.rejects(async () => {
+    await forEachPromise;
+  }, {
+    name: 'AbortError',
+  }).then(common.mustCall());
+
+  setImmediate(() => {
+    ac.abort();
+    assert.strictEqual(calls, 2);
+  });
+}
+
+{
+  // Error cases
+  assert.rejects(async () => {
+    Readable.from([1]).forEach(1);
+  }, /ERR_INVALID_ARG_TYPE/).then(common.mustCall());
+  assert.rejects(async () => {
+    Readable.from([1]).forEach((x) => x, {
+      concurrency: 'Foo'
+    });
+  }, /ERR_OUT_OF_RANGE/).then(common.mustCall());
+  assert.rejects(async () => {
+    Readable.from([1]).forEach((x) => x, 1);
+  }, /ERR_INVALID_ARG_TYPE/).then(common.mustCall());
+}
+{
+  // Test result is a Promise
+  const stream = Readable.from([1, 2, 3, 4, 5]).forEach((_) => true);
+  assert.strictEqual(typeof stream.then, 'function');
+}