Testing secting is user guide renders incorrectly

[eni-awowale]

What happened?

From zulip chat testing section in user guide renders incorrectly.

[eni-awowale]

I am seeing the same thing from a local docs build.

[VeckoTheGecko]

Ummm, do you mean the random strings everywhere? That's intentional - xrst is our hypothesis testing strategies (i.e, how to randomly create xarray datasets - which include random data, attrs and var naming)

[keewis]

I think the problem is more that there's characters that mess with the character count, causing weird line breaks (and maybe there are non-normalized newline characters, or similar). The existence of the characters is not much of a problem on its own, it's just that we don't normalize them before printing.

[eni-awowale]

Yeah @VeckoTheGecko I saw in the example section of the code in strategies.py it included the random attrs and data, but the line breaks made it look kind of wonky to me.

@keewis do we usually normalize outputs like this before printing?

[taus]

it was indeed the random strings that confused me in combination with the formatting. I was comparing to an earlier version where the same section renders like this. Here the formatting is more clear on what is happening

It would seems it is rendering correctly, even though the formatting is a bit wonky. I apologize for the confusion

[keewis]

with that I think we could also just seed the hypothesis rng such that we always get the same short attributes and don't run into the formatting issues.

do we usually normalize outputs like this before printing?

we don't yet, but we could (in addition to seeding)

[VeckoTheGecko]

If we can have a property that can detect broken formatting, I think that would make a great hypothesis test for the repr

[VeckoTheGecko]

I apologize for the confusion

That's all good @taus ! If you were confused by it, I bet other users were too :) . Its good to have these doc/repr improvements

[nightcityblade]

Hi, I'd like to work on this. I'll submit a PR shortly.