Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 641624 - dev-python/matplotlib-1.5.3-r2[doc] - ValueError: The section Notes appears twice in the docstring of None in None.
Summary: dev-python/matplotlib-1.5.3-r2[doc] - ValueError: The section Notes appears t...
Status: RESOLVED OBSOLETE
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: Current packages (show other bugs)
Hardware: All Linux
: Normal normal (vote)
Assignee: Gentoo Science Related Packages
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2017-12-18 22:03 UTC by Martin von Gagern
Modified: 2018-01-08 00:01 UTC (History)
4 users (show)

See Also:
Package list:
Runtime testing required: ---


Attachments
sphinx-err-*.log (sphinx-err-eLfvZx.log,9.54 KB, text/plain)
2017-12-18 22:03 UTC, Martin von Gagern
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Martin von Gagern 2017-12-18 22:03:55 UTC
Created attachment 510830 [details]
sphinx-err-*.log

Exception occurred:
  File "/usr/lib64/python2.7/site-packages/numpydoc/docscrape.py", line 343, in _parse
    raise ValueError(msg)
ValueError: The section Notes appears twice in the docstring of None in None.
The full traceback has been saved in ‹attached sphinx-err-*.log›, if you want to report the issue to the developers.

This is using dev-python/sphinx-1.6.3-r2. Unfortunately the error message is none too helpful. Seems we lost some context somewhere. USE=-doc serves as a workaround.
Comment 1 Kalin KOZHUHAROV 2018-01-07 11:31:29 UTC
Currently this version is not in portage, guess that makes it INVALID/WONTFIX?
Comment 2 Jonas Stein gentoo-dev 2018-01-07 12:12:17 UTC
Thank you, Kalin.

Martin, where was your ebuild from? I did not see, that we do not have the mentioned version in our tree. Do you use an overlay?
Could you test this dev-python/matplotlib from the gentoo tree?
It could help also to add the logs here.
https://wiki.gentoo.org/wiki/Attach_the_logs_to_the_bug_ticket
Comment 3 Martin von Gagern 2018-01-07 23:11:19 UTC
This was the main Gentoo repository, with ARCH=~amd64. The ebuild got removed in https://github.com/gentoo/gentoo/commit/b35eee52212164c14ae6bfb6dff4a56cae401f1b by Michał Górny.

Updating to matplotlib-2.1.0-r1 required updating to sci-mathematics/sage-8.1-r2::sage-on-gentoo due to an explicit minor version dependency there. See https://github.com/cschwan/sage-on-gentoo/commit/12e1ca52181ad497cc5c05a2740e for the bump there.

So yes I'm using an overlay, namely sage-on-gentoo, but no, the ebuild in question did not come from that overlay, although me using the specific version I had been using is likely due to this.

Anyway, with matplotlib-2.1.0-r1[doc] I get this:

Exception occurred:
  File "conf.py", line 19, in <module>
    from sphinx_gallery.sorting import ExplicitOrder
ModuleNotFoundError: No module named 'sphinx_gallery'
The full traceback has been saved in …, if you want to report the issue to the developers.

which makes some sense since dev-python/sphinx-gallery isn't installed on my system. It would appear that the matplotlib ebuild does not reflect https://github.com/matplotlib/matplotlib/blob/v2.1.0/doc-requirements.txt in full. Do you want a separate ticket for this, or abuse this one here?

Installing sphinx-gallery itself fails for bug #626878 so I can't test whether manually installing that dependency would allow me to install current matplotlib with USE=doc. Will gladly revisit once that has been resolved.
Comment 4 Martin von Gagern 2018-01-08 00:01:32 UTC
(In reply to Martin von Gagern from comment #3)
> Do you want a separate ticket for this, or abuse this one here?

Filed bug 643840 for this in order to keep separate things separate.

> Installing sphinx-gallery itself fails for bug #626878 so I can't test
> whether manually installing that dependency would allow me to install
> current matplotlib with USE=doc.

Did work around by emerging dev-python/pytest-runner. Next I hit bug 581550 comment 10 but that appears to be a sporadic thing, since without changes the next run after that did proceed past that point.

So I can confirm that matplotlib-2.1.0-r1[doc] does not encounter this specific issue, although it has several other problems. Closing as obsolete since the ebuild in question is no longer in tree.