Lit platform benchmarks

This repo contains a set of micro-benchmarks of basic platform APIs.

The benchmarks are useful to measure the performance fundamental operations across different DOM and JavaScript techniques and different browsers.

Benchmarks

dirty-check-attributes

This benchmark measures the performance impact of dirty-checking values before calling `setAttribute()``.

It measures three techniques:

• always call `setAttribute()`` (no dirty-check)
• dirty check against the DOM (with getAttribute())
• dirty check against a stored JS value

npm run bench:dirty-check-attributes

dirty-check-text-content

This benchmark measures the performance impact of dirty-checking values before setting Text.textContent.

It measures three techniques:

• always call textContent (no dirty-check)
• dirty check against the DOM
• dirty check against a stored JS value

Future benchmarks

• Setting .textContent vs .data
• append() vs insertBefore()
• template.innerHTML vs DOMParser
• template clone vs imperative DOM construction
• Reusing TreeWalkers
• Finding nodes with TreeWalker + comments vs query selectors, etc