chore(http): release http@1.0.0

[iuioiua]

To be merged July 25.

The latest API docs: https://jsr.io/@std/http@1.0.0-rc.6/doc

Closes #5007

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 96.48%. Comparing base (7fd14ba) to head (12fcba2).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5217   +/-   ##
=======================================
  Coverage   96.48%   96.48%           
=======================================
  Files         465      465           
  Lines       37766    37766           
  Branches     5580     5580           
=======================================
  Hits        36439    36439           
  Misses       1285     1285           
  Partials       42       42           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[bartlomieju]

Do we expect to stabilize parseSignedCookie and signCookie before merging this one?

[iuioiua]

No, we're keeping them unstable until they've lived around long enough and until #5257 is resolved.

[iuioiua]

@kt3k, please provide final review.

[kt3k]

eTag return type can be Promise<string> for string and Uint8Array inputs. Not sure worth the breaking change though..

[iuioiua]

Can you create PR so we can get a better idea of what you mean?

[iuioiua]

Good to merge now, @kt3k?

[kt3k]

LGTM

[kt3k]

but I'll also suggest removing ETagResource symbol, as it's not referenced from public API anymore.

[iuioiua]

SGTM

[iuioiua]

I suggest renaming STATUS_CODE to STATUS_CODES (with an "s" suffix) and STATUS_TEXT to STATUS_TEXTS (also with an "s" suffix).

[kt3k]

I suggest renaming STATUS_CODE to STATUS_CODES (with an "s" suffix) and STATUS_TEXT to STATUS_TEXTS (also with an "s" suffix).

I'm not in favor of that change. Why do we need it?

It feels like a bit too late to start discussing that kind of idea.

[iuioiua]

Well, the name doesn't make sense. The object is not a "status code". Rather, it's an object whose entries are status codes. Same for "status text". The only reason we'd keep it is to avoid disruption, but we'd be settling for something that's incorrect. Plus, it's a very trivial change.

[iuioiua]

For context, node:http has STATUS_CODES and METHODS, which are also objects. Changing to the wrong name was made in #3781 by me.

[kt3k]

These can be seen as equivalent of TypeScript enums, and enums are usually named in singular form. https://www.typescriptlang.org/docs/handbook/enums.html

[iuioiua]

That seems like an obscure reason for the name. Again, it's an object whose entries are status codes. It itself, is not actually a status code.