@itslil/jquery

jQuery 3.7.1, reimplemented in LilScript and published as a dependency-free drop-in.

This is not the official jquery package. It is an independent runtime that implements the [email protected] public API.

Site: yeargun.github.io/jquerylil

npm install @itslil/jquery

import { jQuery as $ } from "@itslil/jquery"

$("button").on("click", function () {
  $(this).toggleClass("on")
})

Alias the import if you already write from "jquery":

{
  "dependencies": { "@itslil/jquery": "3.7.1" },
  "overrides": { "jquery": "npm:@itslil/[email protected]" }
}

Current snapshot

This is the port as it stands. Official jquery.min.js is still smaller. Throughput against [email protected] is faster on every suite we ran.

Size

Reusable package artifacts, measured with lilscript-codec gzip-9 / Brotli-11.

| Lane | Raw | gzip-9 | Brotli-11 | vs official min | | --- | ---: | ---: | ---: | ---: | | Official jquery.js | 285,314 | 83,619 | 69,545 | 2.53× | | Official jquery.min.js | 87,533 | 30,336 | 27,445 | 1.00× | | @itslil/jquery ESM | 92,765 | 34,544 | 30,973 | 1.13× |

Brotli is 1.13× official min and 0.45× official unminified jquery.js. The published ESM is the LilScript compiler output plus a license banner and a default export. It is not pretty-printed and not run through Terser.

Example app bundles

Same six apps, Vite production minify, official [email protected] vs @itslil/jquery. The number is the emitted JS chunk.

| App | jquery Brotli | @itslil/jquery Brotli | Ratio | | --- | ---: | ---: | ---: | | Todos | 28,847 | 30,496 | 1.06× | | Tabs | 28,634 | 30,273 | 1.06× | | Search | 28,737 | 30,424 | 1.06× | | Cart | 28,877 | 30,503 | 1.06× | | Accordion | 28,548 | 30,214 | 1.06× | | Gallery | 28,790 | 30,421 | 1.06× |

Vite minifies both lanes. Official still wins the app chunks.

Performance

Isolated Node v20.12.0 processes versus [email protected]. 8 samples, first 2 discarded, median of the rest. Ratio is @itslil/jquery / official (lower is faster). Checksums match on every suite. Mean retained memory 1.00×.

| Suite | [email protected] | @itslil/jquery | Ratio | | --- | ---: | ---: | ---: | | core | 224.78 ms | 212.69 ms | 0.95× | | events | 37.63 ms | 33.55 ms | 0.89× | | deferred | 52.46 ms | 17.77 ms | 0.34× | | collection | 4.22 ms | 4.04 ms | 0.96× | | event-state | 8.62 ms | 7.24 ms | 0.84× |

5 / 5 suites ≤ 1.05×. All five are faster. The deferred suite is a much cheaper implementation of the same resolve/done checksum; it is not a different workload.

Compatibility

• jQuery / $ named exports, default export, and window.jQuery / window.$ on script-tag builds
• ESM, CJS, and UMD artifacts
• fn.jquery === "3.7.1"
• Node / jsdom: put window and document on globalThis before import. Official CJS is a factory(window). This package is a singleton bound at load time.
• Some fn.length values differ because LilScript emits rest wrappers. Call the documented arguments; do not branch on .length.

Zero runtime dependencies.

Rebuild

Compiled JavaScript in dist/ is what npm installs. Rebuilding from src/**/*.lil needs a release LilScript compiler next to this repo, or LILSCRIPT_COMPILER.

npm run build            # wrap the current compiler artifact
npm run build -- --compile
npm test                 # jsdom vs [email protected]
npm run bench
npm run measure
npm run examples         # http://127.0.0.1:4178/examples/

License

MIT. See NOTICE.md.