http2: improve http2 code a bit

[jasnell]

Multiple general improvements to http2 internals for readability and efficiency

Checklist

• make -j4 test (UNIX), or vcbuild test (Windows) passes
• tests and/or benchmarks are included
• documentation is changed or added
• commit message follows commit guidelines

[nodejs-github-bot]

@jasnell build started: https://ci.nodejs.org/blue/organizations/jenkins/node-test-pull-request-lite-pipeline/detail/node-test-pull-request-lite-pipeline/1429/pipeline

[jasnell]

CI: https://ci.nodejs.org/job/node-test-pull-request/18276/

[jasnell]

ping @nodejs/http2

[mcollina]

LGTM

Are there any performance benefits?

[Trott]

Landed in 7825045

[GrosSacASac]

From https://github.com/nodejs/node/blob/7825045ee695e9e5c048133255a3b614e04c98d3/lib/internal/http2/util.js
Is it a good idea to use Object.entries loop in this case ?

it would replace

 const keys = Object.keys(map);

  let i;
  let key;
  let value;

  for (i = 0; i < keys.length; i++) {
    key = keys[i];
    value = map[key];
  }

with something like

  Object.entries(map).forEach(([key, value]) => {
  
  });

[GrosSacASac]

node/lib/internal/http2/util.js

Line 486 in 7825045

ret += `${key}\0${value}\0`;

Is it a good idea to push into an array, and join all the strings at the end, instead of doing string concatenation each time in the loop ?

[jasnell]

Using entries and forEach is quite a bit more expensive performance wise, as is join, because it forces additional unnecessary iterations over the values.

[GrosSacASac]

What is the meaning of irritations in this context ?

[jasnell]

Sorry, that was a phone autocorrect error... Lol

I meant additional unnecessary iterations 😂

[GrosSacASac]

The most natural way to write a program should result in top-of-the-line runtime performance. In places where performance is king, the optimal code should be clearly expressible. Should we open issue in the engine ?

[jasnell]

The v8 team is continually making improvements to the performance in this area, and I suspect they will continue to do so. We'll get there eventually.

[codebytere]

@jasnell do you want to potentially backport this?

[jasnell]

Yeah, we should

…

On Wed, Nov 28, 2018, 18:59 Shelley Vohr ***@***.*** wrote: @jasnell <https://github.com/jasnell> do you want to potentially backport this? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#23984 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAa2ebmROWc4uqrmlFd-msYrR2w0rHT5ks5uz02sgaJpZM4YDDzo> .

[codebytere]

@jasnell do you have the bandwidth for this atm?

[jasnell]

Not so much. Will take a few weeks for me to get to it