Benchmark: Fastest way to list children: childNodes vs children vs firstChild/nextSibling vs firstElementChild/nextElementSibling v2 fixed

Fastest way to list children: childNodes vs children vs firstChild/nextSibling vs firstElementChild/nextElementSibling v2 fixed (version: 0)

Result: firstElementChild is the fastest yielding 3-4X performance. With lastElementChild close (margin of error) 2nd. Other observations: doing for loop i-- with children shows ~150% higher performance vs i++. And unloading it into a variable prior yields another ~200% boost

Comparing performance of: childNodes foreach vs children (direct) vs children (into variable) vs children-- (direct) vs children-- (into variable) vs ChildNodes for (direct) vs ChildNodes for (into variable) vs ChildNodes for-- (direct) vs ChildNodes for-- (into variable) vs firstElementChild nextElementChild vs lastElementChild previousElementChild vs firstChild nextSibling vs lastChild previousSibling

Created: 3 years ago by: Guest

Jump to the latest result

HTML Preparation code:

<html>

<body>
    <div id='test'>
        <div name='a'></div>
        <div name='b'></div>
        <div name='c'></div>
    </div>
</body>

</html>

Script Preparation code:

var parent = document.getElementById('test');

Tests:

childNodes foreach
parent.childNodes.forEach(function (node) { let n = node; });
children (direct)
for (let i = 0; i < parent.children.length; i++) { let n = parent.children[i]; }
children (into variable)
let children = parent.children; for (let i = 0; i < children.length; i++) { let n = children[i]; }
children-- (direct)
for (let i = parent.children.length-1; i >= 0; i--) { let n = parent.children[i]; }
children-- (into variable)
let children = parent.children; for (let i = children.length-1; i >= 0; i--) { let n = children[i]; }
ChildNodes for (direct)
for (let i = 0; i < parent.childNodes.length; i++) { let n = parent.childNodes[i]; }
ChildNodes for (into variable)
let children = parent.childNodes; for (let i = 0; i < children.length; i++) { let n = children[i]; }
ChildNodes for-- (direct)
for (let i = parent.childNodes.length-1; i >= 0; i--) { let n = parent.childNodes[i]; }
ChildNodes for-- (into variable)
let children = parent.childNodes; for (let i = children.length-1; i >= 0; i--) { let n = children[i]; }
firstElementChild nextElementChild
let elem = parent.firstElementChild; do { let n = elem; } while (elem = elem.nextElementSibling)
lastElementChild previousElementChild
let elem = parent.lastElementChild; do { let n = elem; } while (elem = elem.previousElementSibling)
firstChild nextSibling
let elem = parent.firstChild; do { let n = elem; } while (elem = elem.nextSibling)
lastChild previousSibling
let elem = parent.lastChild; do { let n = elem; } while (elem = elem.previousSibling)

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
childNodes foreach
children (direct)
children (into variable)
children-- (direct)
children-- (into variable)
ChildNodes for (direct)
ChildNodes for (into variable)
ChildNodes for-- (direct)
ChildNodes for-- (into variable)
firstElementChild nextElementChild
lastElementChild previousElementChild
firstChild nextSibling
lastChild previousSibling

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated one month ago):

LLMs can make mistakes. Check important info.

Latest run results:

Run details: (Test run date: one year ago)

User agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/117.0.0.0 Safari/537.36 Edg/117.0.2045.31

Browser/OS: Chrome 117 on Windows

View result in a separate tab

Test name	Executions per second
childNodes foreach	629319.8 Ops/sec
children (direct)	849352.2 Ops/sec
children (into variable)	2385627.5 Ops/sec
children-- (direct)	1382130.1 Ops/sec
children-- (into variable)	3051974.8 Ops/sec
ChildNodes for (direct)	446341.5 Ops/sec
ChildNodes for (into variable)	1631156.8 Ops/sec
ChildNodes for-- (direct)	758549.2 Ops/sec
ChildNodes for-- (into variable)	2075338.9 Ops/sec
firstElementChild nextElementChild	5528009.0 Ops/sec
lastElementChild previousElementChild	5510178.5 Ops/sec
firstChild nextSibling	4411911.0 Ops/sec
lastChild previousSibling	4426229.0 Ops/sec

Related benchmarks:

Fastest way to list children: childNodes vs children vs firstChild/nextSibling vs firstElementChild/nextElementSibling v2 fixed (version: 0)

Result: firstElementChild is the fastest yielding 3-4X performance. With lastElementChild close (margin of error) 2nd. Other observations: doing for loop i-- with children shows ~150% higher performance vs i++. And unloading it into a variable prior yields another ~200% boost

Created: 3 years ago by: Guest

Jump to the latest result

childNodes foreach

children (direct)

children (into variable)

children-- (direct)

children-- (into variable)

ChildNodes for (direct)

ChildNodes for (into variable)

ChildNodes for-- (direct)

ChildNodes for-- (into variable)

firstElementChild nextElementChild

lastElementChild previousElementChild

firstChild nextSibling

lastChild previousSibling

Suite status: <idle, ready to run>

Experimental features:

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated one month ago):