feat(binding-builder): support using custom binding

[h-a-n-a]

Summary

This PR supports using custom binding with environment variable RSPACK_BINDING.
Path-to-binding is a path to the root of the binding package.

RSPACK_BINDING=path-to-binding

This is achieved by our already used ast-grep package to scan and replace the original content of module @rspack/binding and @rspack/binding/package.json.

In order to either use or test the new binding package (which contains support for custom plugin), this PR also adds support for exposing experiments.createNativePlugin which is an alias to the original plugin create and would throw error if there's any naming collision with the internal builtin plugins. createNativePlugin is used to create wrappers around native plugins:

const MyPlugin = createNativePlugin("MyPlugin", (options) => {
 return options;
});

new MyPlugin({
 foo: "bar"
}).apply(compiler);

Tests are now containing two parts, the original binding and the builder testing binding. The latter part is to test if rspck-binding-builder successfully extended all the exposed symbols and functionalities provided by crates/node_binding. In order to cover that part, tests are written in a mixed style with the original rspack tests. With test.filter.js, these tests are only enabled when running test:base-builder script in rspack-test-tools. I talked to @stormslowly about these changes to the workflows. Due to the complexity, I would separate the workflow change to a separate PR.

Related links

Checklist

• Tests updated (or not required). Required, but to heavy to implement in the same PR.
• Documentation updated (or not required).

[netlify]

✅ Deploy Preview for rspack canceled.

Name	Link
🔨 Latest commit	3477666
🔍 Latest deploy log	https://app.netlify.com/projects/rspack/deploys/6863b44a84076b000877fcb1

[codspeed-hq]

CodSpeed Performance Report

Merging #10857 will not alter performance

_{Comparing binding-builder (3477666) with main (fc9cb1e)}

🎉 Hooray! codspeed-node just leveled up to 4.0.1!

A heads-up, this is a breaking change and it might affect your current performance baseline a bit. But here's the exciting part - it's packed with new, cool features and promises improved result stability 🥳!
Curious about what's new? Visit our releases page to delve into all the awesome details about this new version.

Summary

✅ 16 untouched benchmarks

[hardfist]

@ityuany can you help explain your current workflow of integration and see whether this solution meets your needs

[ityuany]

@ityuany can you help explain your current workflow of integration and see whether this solution meets your needs

https://github.com/sheinsight/spack

We are currently forking rspack to attempt customized feature development, and then releasing our own binding. In our internal framework, we forward requests to @rspack/binding to @shined/spack-binding by hijacking require.resolve.