feat(context): infer TypedResponse for TextRespond

[NamesMT]

Author should do the followings, if applicable

• Add tests Adjusted tests
• Run tests - Ran all tests except for deno and wrangler because I don't have them
• yarn denoify to generate files for Deno

[yusukebe]

Hi @NamesMT

This is awesome. I'd like to merge it into the "next" branch for the next minor release will be shipped soon.

I've fixed that res.text() and res.json() on the client side did not return the correct types with 9d690a6. Now, these are correct:

I think this is good. How about you?

[NamesMT]

Thanks @yusukebe, that's great.

But please wait a bit before merging, I found an untested regression, will document and add tests for it maybe tonight, urgent have to go now.

[yusukebe]

@NamesMT Okay!

[NamesMT]

Hi @yusukebe, I've done the updates to fix the regression and improve the type inferring for the Response.

The regression is: c.json() is able to return string JSONPrimitive, in the latest update to res.json() they were broken and always typed as never.

The final scope of work to get everything perfectly working was much bigger than I thought and I don't know if I did it the best way, so I will create a new draft PR for it to discuss if it's good or not.

It's an un-splitted commit but heres the readable updates that I've made:
// Edit for information: Simplify is copied over from type-fest

• Switch from InterfaceToType to Simplify for better performance + easier working with types, InterfaceToType was yelling Type instantiation is excessively deep and possibly infinite.
• TypedResponse: add ResponseFormat generic ('json' | 'text')
• ClientResponse: accept ResponseFormat generic to properly return the correct type
• Endpoint (and related interfaces): now includes outputFormat: ResponseFormat

Bonus:

• ClientResponse now properly extends globalThis.Response for better experience when working with functions that expects a fetch's Response

+ add new + update related tests

New PR: #2581

[yusukebe]

Hi @NamesMT

That's super great. Anyway, I want to go with #2581 though we may have improvements. Let's talk on #2581

[NamesMT]

Close to continue in #2581