fix: use .mjs extension for injected client modules

[sodatea]

Description

Otherwise, they will not be recognized by Node.js as ES modules.

Usually, it doesn't matter, as they're intended to be run in browsers.

I only encountered this problem in vite-jest, which tries to run vite-transpiled code in Node.js environments.
After this PR, the following workaround would no longer be necessary:
https://github.com/sodatea/vite-jest/blob/1e05b4c62df714b1537c76052d5fec761a46c619/packages/vite-jest/bin/vite-jest.js#L11-L12

Additional context

---

What is the purpose of this pull request?

• Bug fix
• New Feature
• Documentation update
• Other

Before submitting the PR, please make sure you do the following

• Read the Contributing Guidelines.
• Read the Pull Request Guidelines and follow the Commit Convention.
• Check that there isn't already a PR that solves the problem the same way to avoid creating a duplicate.
• Provide a description in this PR that addresses what the PR is solving, or reference the issue that it solves (e.g. fixes #123).
• Ideally, include relevant tests that fail without this PR but pass with it.

[patak-dev]

Should we also use package.json type module here if we merge #4063 ?

[sodatea]

Not necessary.

Per Node.js documentation:

Files ending with .mjs are always loaded as ES modules regardless of the nearest parent package.json.

[patak-dev]

I mean about being consistent. Both approaches look fine 👍🏼

[sodatea]

I've just fixed the tests in #4064
If consistency is a concern, maybe we should merge that one rather than #4063.

I don't have an opinion on package.json or .mjs for the cache directory.
But in the case of these client modules, as they're inside the vite package, I prefer not to add another package.json.