Benchmark: add destructuring object

[fundon]

Checklist

• make -j4 test (UNIX), or vcbuild test nosign (Windows) passes
• tests and/or benchmarks are included

Affected core subsystem(s)

Description of change

Maybe throse codes should be changed:

FROM

https://github.com/nodejs/node/blob/master/lib/fs.js#L1-L40

TO

// Maintainers, keep in mind that ES1-style octal literals (`0666`) are not
// allowed in strict mode. Use ES6-style octal literals instead (`0o666`).

'use strict';

const { fs: constants } = process.binding('constants');
const util = require('util');
const pathModule = require('path');

const binding = process.binding('fs');
const fs = exports;
const { Buffer } = require('buffer');
const { Stream } = require('stream');
const EventEmitter = require('events');
const { FSReqWrap } = binding;
const { FSEvent } = process.binding('fs_event_wrap');
const { assertEncoding, SyncWriteStream } = require('internal/fs');

Object.defineProperty(exports, 'constants', {
  configurable: false,
  enumerable: true,
  value: constants
});

const { Readable, Writable } = Stream;

const kMinPoolSpace = 128;
const { kMaxLength } = require('buffer');

const {
  O_APPEND = 0,
  O_CREAT = 0,
  O_EXCL = 0,
  O_RDONLY = 0,
  O_RDWR = 0,
  O_SYNC = 0,
  O_TRUNC = 0,
  O_WRONLY = 0
} = constants;

Benchmark Results: ~~~23% improvement~~ higher is better

 λ ~/D/o/node git:(master) 1≡ > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 392.4462171885431
es/destructuring-object-bench.js millions=100 method="destructureObject": 300.60502352237313
 λ ~/D/o/node git:(master) 1≡ > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 378.884718277464
es/destructuring-object-bench.js millions=100 method="destructureObject": 306.7606096492869
 λ ~/D/o/node git:(master) 1≡ > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 385.192791572974
es/destructuring-object-bench.js millions=100 method="destructureObject": 296.1235505392568

[mscdex]

These assert calls are unnecessary and will just slow down the benchmark.

[fundon]

I just follow other benchmark cases. Need to delete them?

[fundon]

Remove all assert and results: ~38% improvement higher is better

 λ ~/D/o/node git:(master) 1≡ 1M > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 903.2505158892741
es/destructuring-object-bench.js millions=100 method="destructureObject": 564.4660663547472
 λ ~/D/o/node git:(master) 1≡ 1M > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 890.7910802024818
es/destructuring-object-bench.js millions=100 method="destructureObject": 547.1594066200685
 λ ~/D/o/node git:(master) 1≡ 1M > ./node benchmark/es/destructuring-object-bench.js
es/destructuring-object-bench.js millions=100 method="normal": 900.7505972742099
es/destructuring-object-bench.js millions=100 method="destructureObject": 523.6524462787967

[mscdex]

The benchmark results you've posted show that non-destructuring is faster (higher number, since results are ops/sec.). Even so, having this benchmark might be nice to have anyway.

[fundon]

😂 I think it's excuted times. Sorry

[mscdex]

If you remove the assert calls, the adding of this benchmark LGTM.

[jasnell]

LGTM

[imyller]

LGTM

[imyller]

CI: https://ci.nodejs.org/job/node-test-pull-request/4248/

[imyller]

@fundon Before we can land this, could you please address following linter errors:

destructuring-object-bench.js
  15:9   error  'x' is defined but never used  no-unused-vars
  16:9   error  'y' is defined but never used  no-unused-vars
  17:9   error  'r' is defined but never used  no-unused-vars
  27:11  error  'x' is defined but never used  no-unused-vars
  27:14  error  'y' is defined but never used  no-unused-vars
  27:17  error  'r' is defined but never used  no-unused-vars

[mscdex]

I think we can add code comments to ignore those particular linter errors?

[imyller]

I think we can add code comments to ignore those particular linter errors?

I agree.

@fundon You can enclose lines 15-17 and line 27 with:

/* eslint-disable no-unused-vars */
 ...
/* eslint-enable no-unused-vars */

[fundon]

Ok, I update it.

[imyller]

CI: https://ci.nodejs.org/job/node-test-pull-request/4258/

[fundon]

Some tests failed. What should i do?

[imyller]

@fundon CI failures are unrelated. They all seem to be known issues.

I'll start prepping this PR for landing in few hours 👍

[imyller]

Landing:

• Three LGTMs
• No objections
• Requested changes have been made
• CI test passed (only known and unrelated CI failures)

[imyller]

landed in 7458fbb

Thank you, @fundon