chore(url): release url@1.0.0

[iuioiua]

To be merged July 7.

Closes #4924

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.82%. Comparing base (b6906a9) to head (c245151).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4992   +/-   ##
=======================================
  Coverage   95.82%   95.82%           
=======================================
  Files         458      458           
  Lines       37853    37853           
  Branches     5563     5563           
=======================================
  Hits        36272    36272           
  Misses       1541     1541           
  Partials       40       40           

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

[kt3k]

LGTM

[lucacasonato]

I don't think the basename behaviour makes sense - we should not stabilize it.

Namely the fact that it is really

import { basename } from "@std/path/posix";

if (url.path !== "/") {
  return basename(url.path);
} else {
  return url.hostname;
}

Additionally, the connotation that url.* methods operates on a URL is not really correct. It operates on URLs that have posix formatted paths. This is weird, because not all URLs have posix formatted paths. Take for example data URLs:

> new URL("data:application/javascript,1+1;")
URL {
  href: "data:application/javascript,1+1;",
  origin: "null",
  protocol: "data:",
  username: "",
  password: "",
  host: "",
  hostname: "",
  port: "",
  pathname: "application/javascript,1+1;",
  hash: "",
  search: ""
}

This is not a posix formatted path.

Because of this, I think that really we shouldn't be doing this - we should just document that people can use join, basename, and friends from @std/path/posix.

[iuioiua]

Frankly, the purpose of @std/url has always been a little unclear to me because the APIs are such thin wrappers around @std/path APIs, as you point out. I'll create a PR which aims to improve @std/path documentation by including equivalent examples for @std/url. Then we can see what to do with @std/url 🙂

[iuioiua]

Perhaps, one of the only things @std/url offers over just using @std/path/posix APIs is that it strips URL.hash and URL.search properties before processing the values. I'm not sure whether that provides sufficient justification for keeping the package...

[iuioiua]

Closing as we're instead removing @std/url. See #5530 for details.