Sphinx showing tuple doc strings in using qcodes 0.32.0

[ereehuis]

Hi! I'm chasing an elusive bug in building our quantify-core documentation

=> In using qcodes 0.32.0, I get doc strings from tuple built-in instead of our own doc strings

python -m pydoc tuple

Help on class tuple in module builtins:

class tuple(object)
 |  tuple(iterable=(), /)
 |
 |  Built-in immutable sequence.
 |
 |  If no argument is given, the constructor returns an empty tuple.
 |  If iterable is specified the tuple is initialized from iterable's items.
 |
 |  If the argument is a tuple, the return value is the same object.

=> When changing to qcodes 0.31.0, all is fine

=> Lastly, when commenting out "qcodes.sphinx_extensions.parse_parameter_attr" in our sphinx extensions https://gitlab.com/quantify-os/quantify-core/blob/c33b26d589672a6e2598b074d39d70afc678ba17/docs/conf.py#L72-L72
methods that showed the tuple doc string before are now empty, as a second confirmation of the link to qcodes / qcodes.sphinx_extensions.parse_parameter_attr

=> Have you ever seen this tuple message show up in docs like this, any idea how to circumvent or what change/PR might have led to this? 🙏

Scanned through PRs merged into version 0.32.0, found this one that looked interesting as it deals with doc strings and Tuple and instruments, but is a very loose lead: https://github.com/QCoDeS/Qcodes/pull/3864/files

[jenshnielsen]

Thats strange. I don't see any obvious reason for why this is happening.
We are not seeing any issue like this in the qcodes documentation.

Could you try to bisect between 0.31 and 0.32 to see which commit triggered this?
https://www.tutorialdocs.com/article/git-bisect-tutorial.html

[ereehuis]

Will do, thanks for the suggestion 👍

[ereehuis]

Actually, issue with Ellipsis in docstrings was due to us using Sphinx v3. I'm moving us to Sphinx v4 which resolves the issue (see https://gitlab.com/quantify-os/quantify-core/-/merge_requests/317)

Thanks again!

[jenshnielsen]

Thanks for the info. We don't put a min supported sphinx version in our docs but perhaps we should