issue using urljoin from urllib.parse

[sebastian-bearing]

Bug report

Bug description:

on joining a url where the base url not finish on "/" have the following impact

from urllib.parse import urljoin

print(urljoin('https://localhost/1/','2'))
# => https://localhost/1/2
print(urljoin('https://localhost/1/','/2'))
# => https://localhost/2
print(urljoin('https://localhost/1','/2'))
# =>https://localhost/2
print(urljoin('https://localhost/1','2'))
# => https://localhost/2

CPython versions tested on:

3.12

Operating systems tested on:

macOS

[zangjiucheng]

Hello,

I tested with this input and reproduced the output you mentioned, but this seems as expected with RFC 3986 and Python docs behaviour?

To be more detail, can you please mention what you expect the output to be for this?

[sebastian-bearing]

its expected the base urljoin(base, url, allow_fragments=True): part of the base url is remove when do not finish on /?

[StanFromIreland]

part of the base url is remove when do not finish on /?

From RFC 3986 to which the function conforms:

return a string consisting of the reference's path component
appended to all but the last segment of the base URI's path (i.e.,
excluding any characters after the right-most "/" in the base URI
path, or excluding the entire base URI path if it does not contain
any "/" characters).

[zware]

While it's easy to expect different behavior from urljoin if one is accustomed to os.path.join (in fact, I recently ran into the same issue and had to re-educate myself), as mentioned by @zangjiucheng urljoin is behaving as documented and expected following RFC 3986.

[sebastian-bearing]

i the same line argue to following "//" is expected this result?

print(urljoin('https://localhost/1/','/2'))

https://localhost/2

[ericvsmith]

Please ask usage questions on https://discuss.python.org/c/help/7, not the issue tracker.