urllib.parse.Misimplementation of urljoin

[SBlechmann]

Describe the bug
The urllib.parse.urljoin, e.g. when using orion base url and the version endpoint in the get_version() function, works a little different than originally expected.

urljoin(url1, url2)

E.g. if orion is running behind "example.org/orion" and I run the get_version, the function will end up querying "example.org/version".
This is because urljoin does not simply concatenate the URLs but cuts the first url depending on how many "/" are part of second url.

To Reproduce
see above, give the orion client a url with a path /orion

Expected behavior
Get version; instead getting nothing cause url is wrong.

[djs0109]

Hi, can you try this code snippet? Make sure you put a trailing slash / at the end of the baseurl.

from urllib.parse import urljoin
base_url = "example.org/orion/"
url = urljoin(base_url, "version")
assert url == "example.org/orion/version"

If it is still an issue, please specify your python environment more specifically.