Files
oxc-project__oxc/packages/codegen/tsdown_plugins/const_functions.ts
overlookmotel a4478e960a feat(codegen): add oxc-codegen package (#25488)
Add `oxc-codegen` package - a printer written in TS.

### Features

- Prints JS, JSX, TS, TSX.
- Matches `oxc_codegen` exactly on all Test262, TypeScript, Acorn-JSX fixtures.
- Very fast.

### Deficiencies

- Pretty-printing only (no compact/minified output).
- No support for printing comments.
- Source map support is semi-there, but largely untested, and poor API (design doc has some ideas on how to improve it).

### This PR adds

- The printer code.
- Conformance testing infra.
- Further tests for edge cases.
- README.
- Design doc - how it differs from `oxc_codegen` and why, what produces the perf.

### This PR doesn't add

- Not integrated into CI or release pipeline.

### Notes

This started as a slop port of `oxc_codegen`. I've iterated on that base, cleaned it up a lot, and at least skimmed through all of it. I think it's in fairly decent shape.

The design doc is I think worth a read. It explains the rationale for the choices, in particular the "1 codebase, 4 builds" approach.

I've put it in a new top-level directory called `packages`. It's not an app, so didn't seem to me to fit alongside `apps/oxlint` and `apps/oxfmt`. It's pure JS, so not like e.g. `napi/parser`. And it felt too heavyweight to put into `npm` dir. But maybe there's a better place for it.

Co-authored-by: Cameron <cameron.clark@hey.com>
2026-08-12 13:41:31 +00:00

69 lines
2.3 KiB
TypeScript

import { parseSync } from "oxc-parser";
import type { Plugin } from "rolldown";
/**
* Plugin to rewrite the printer's top-level function declarations as `const` bindings.
*
* ```ts
* // Original code
* function printIf(node: ESTree.IfStatement, state: State): void {
* // ...
* }
*
* // After transform
* const printIf = (node: ESTree.IfStatement, state: State): void => {
* // ...
* };
* ```
*
* A function declaration creates a binding which could be assigned to later, so a call has to read
* the current value out of the module's scope. A `const` binding cannot, so V8 can treat it
* as a constant and call it directly. Measured on all 7 benchmark fixtures, it is worth ~3%.
*
* It really is the `const`, and not the arrow function or the smaller output it allows - the same rewrite
* with `let` measures ~4% slower than this one, and slower than leaving the declarations alone.
* Do not relax it to `let`.
*
* The source keeps its declarations - they hoist, they read better, and they are what the rest of
* the codebase looks like. This is why the rewrite is a build step rather than how printer is written.
*/
const plugin: Plugin = {
name: "const-functions",
transform: {
// Only process TS files in `src-js/print` directory
filter: { id: /\/src-js\/print\/.+\.ts$/ },
handler(code, path, meta) {
const magicString = meta.magicString!;
const { program, errors } = parseSync(path, code);
if (errors.length !== 0) throw new Error(`Failed to parse ${path}: ${errors[0].message}`);
for (const statement of program.body) {
let fn;
if (statement.type === "FunctionDeclaration") {
fn = statement;
} else if (
statement.type === "ExportNamedDeclaration" &&
statement.declaration?.type === "FunctionDeclaration"
) {
fn = statement.declaration;
} else {
continue;
}
if (fn.async || fn.generator) continue;
// `function name` -> `const name =`, `{` of the body -> `=> {`, and `}` -> `};`
magicString.overwrite(fn.start, fn.id!.end, `const ${fn.id!.name} =`);
magicString.appendLeft(fn.body!.start, "=> ");
magicString.appendRight(fn.end, ";");
}
return { code: magicString };
},
},
};
export default plugin;