bpo-36675: Doc: Reveal doctest directives#23620
Uh oh!
There was an error while loading. Please reload this page.
Conversation
JulienPalard commented Dec 2, 2020
Backport won't be possible, we have older sphinx on older branches. |
JulienPalard commented Dec 3, 2020
@andresdelfino I see you a lot on the doc those time, though you would be interested in proofreading this :) |
andresdelfino commented Dec 3, 2020
@JulienPalard Sure! |
andresdelfino commented Dec 15, 2020
@JulienPalard could you check the comment I made? I believe something is missing. |
JulienPalard commented Dec 16, 2020
@andresdelfino Gladly, but, which comment? |
| Another bad idea is to print things that embed an object address, like | ||
| .. doctest:: | ||
There was a problem hiding this comment.
Perhaps
| :no-trim-doctest-flags: |
is missing here?
There was a problem hiding this comment.
The rendered docs doesn't show the doctest directive in this case.
There was a problem hiding this comment.
I think it's better to keep it hidden in this case. The case is:
>>> id(1.0) # certain to fail some of the time 7948648 >>> class C: pass >>> C() # the default repr() for instances embeds an address <C object at 0x00AC18F0> the comment in the case is true: it's expected to fail. If we reveal the hidden doctest: +SKIP it's no longer expected to fail, the comment becomes erroneous.
The point of this example is to demo usefull cases for +ELLIPSIS, not +SKIP.
andresdelfino commented Dec 16, 2020
@JulienPalard sorry, I thought it was submitted, but I was seeing a draft. |
asottile commented Jan 5, 2021
This breaks the minimum version of sphinx listed in Docs/conf.py: Line 48 in f7f0ed5
|
JulienPalard commented Jan 6, 2021
@asottile that's right, let's bump this, thanks for noticing! |
asottile commented Jan 6, 2021
@JulienPalard there's been additional efforts in this release to maintain compatibility with 1.x so there might be competing opinions here |
Ok so let's discuss this futher in an issue. I'm opening it → https://bugs.python.org/issue42843 |
asottile commented Jan 6, 2021
sounds good -- I only noticed this because it broke the packaging for deadsnakes (on 20.04, 18.04 and 16.04) so I reverted this patch there: deadsnakes/python3.10@f69292e |
JulienPalard commented Jan 6, 2021
Sorry for this :( |
asottile commented Jan 6, 2021
no big deal, I'm considering dropping the documentation build since it's ~95% of my maintenance and I'm not sure anybody actually uses it. |
This reverts commit c8a10d2.
5 participants
Doctest directive were hidden in HTML and PDF output, this PR show them back.
https://bugs.python.org/issue36675