Latex pictures for Dynkin diagram and misc improvements to Cartan types

[wdj]

Displaying nice latex pictures for all Dynkin diagrams of affine and classical types.

Also:

• Improves the description of which conventions are used for Dynkin diagrams and Cartan matrices (Bourbaki/Kac?).

• Misc improvements to Cartan types, including a global option for displaying Cartan types using Kac's conventions and much documentation cleanup.

[mabshoff]

Mike,

this has been sitting around for a while. What is the status here?

Cheers,

Michael

[nthiery]

It's indeed a bit of an abuse to have Dynkin diagram derive from Digraphs (some edges are not oriented). But they are not Graphs either (some edges are). They don't really deserve a special class graph just for themselves, do they? So, I guess we can live with this abuse.

That being said, plot should definitely be overriden to get appropriate pictures. Volunteers?

See also #5502 for ascii art drawing.

[tscrim]

I've uploaded a patch that gives custom latex printing for Dynkin diagrams for crystallographic types.

[tscrim]

New version which uses the global options framework for notation choices.

[tscrim]

Cleaned up some docstrings.

[nthiery]

Hi Travis,

I pushed a reviewer patch on the queue which makes the logic more concise as we had discussed this morning. Please check my changes and fold them. Due to some changes I undid in my patch, you probably want to have a look at the folded patch, and strip away trivial space changes that could be left due to uncomplete undoes.

I'll then have a final check on the updated patch.

Cheers,

Nicolas

[tscrim]

Hey Nicolas,

Thank you for the review. I had to make some minor tweaks to affine types B,C, and D. However this patch will change depending on what happens in #14248.

Thanks,
Travis

[tscrim]

Updated with a better note about the conventions used in sage.

[nthiery]

I have just been through the patch, and wrote a little reviewer patch which I just pushed to the Sage-Combinat queue. It sounds good to go assuming all tests pass.

Travis: if you are happy with the reviewer patch, please fold upload and set a positive review on my behalf.

[jdemeyer]

dochtml.log:[combinat ] /mazur/release/merger/sage-5.10.beta3/devel/sage/doc/en/reference/combinat/sage/combinat/root_system/cartan_type.rst:11: WARNING: error while formatting signature for sage.combinat.root_system.cartan_type.CartanType_crystalographic.ascii_art: invalid syntax (<unknown>, line 1)

[nthiery]

Sorry, we should have caught that. Worked around in the attached patch. See also #14553.

The updated patch was checked by Travis. I am running the tests now.

[nthiery]

All test passed on sage.math.u-psud.fr and documentation built smoothly:

For the full logs, see:

• ​http://sage.math.washington.edu/home/nthiery/2023-buildlog
• ​http://sage.math.washington.edu/home/nthiery/2023-docbuildlog
• ​http://sage.math.washington.edu/home/nthiery/2023-testlog

[nthiery]

Some doctests are ignored with the new doctest framework, and the framework complains about it with long tests.

See discussion on:

​https://groups.google.com/forum/?fromgroups=#!topic/sage-devel/4m1ydGdiGf8

[nthiery]

The new patches contains two little changes we agreed with Travis / on sage-devel:

• Some trailing whitespace in new lines
• Updating the number of currently ignored tests in doctest/sources

All long test passed.

Back to positive review!

[nthiery]

Minor doc tweak