URLLib Join Behaviour

[OfficiallySomeGuy]

Bug report

Bug description:

This bug proposes that

urllib.parse.urljoin('https://example.com/thing', 'v1')

should resolve to https://example.com/thing/v1 rather than the current behaviour which resolves to https://example.com/v1 without warning. I believe the relevant code is here. This would also resolve some of the concerns in #96015.

CPython versions tested on:

3.11

Operating systems tested on:

macOS

Linked PRs

• gh-131814: Remove logic stripping URL components not suffixed with / #131815

[ericvsmith]

This would be a breaking change. You'd need to discuss it at https://discuss.python.org/c/ideas/6 first.

[brianschubert]

The current behavior feels correct to me. This matches how relative URLs are generally handled in web browsers. For example, on this webpage (https://github.com/python/cpython/issues/131814), if I create a link with the relative URL foo, my browser resolves it to the absolute URL https://github.com/python/cpython/issues/foo, not https://github.com/python/cpython/issues/131814/foo.

Example: <a href=foo>link</a> becomes link

[ericvsmith]

I'm going to close this, pending a discussion on d.p.o. If there's a consensus there on a change, this can be re-opened.