character art fails for LieAlgebra elements

[gh-mwageringel]

sage: L.<x,y> = LieAlgebra(QQ)
sage: ascii_art(x)
...
AttributeError: 'FreeLieAlgebra.Lyndon_with_category' object has no attribute '_ascii_art_term'

This could be fixed by implementing _ascii_art_term or by implementing an appropriate fallback for cases where _ascii_art_term is not defined. The latter would be desirable in any case.

The same applies to the unicode art representation.

[mkoeppe]

Moving tickets to milestone sage-9.2 based on a review of last modification date, branch status, and severity.

[tscrim]

I am doing a full overhaul of the acii and unicode art for Lie algebras.

[tscrim]

Here is a branch that fixes up a bunch of issues with ascii/unicode art across the Lie algebra code and adds some improved displays.

---

New commits:

​3433479	Creating functions for converting numbers into unicode sub/superscripts.
​3fc2bdb	Adding support for negative numbers and superscript fractions.
​830d2f1	Fixing ascii/unicode art for free Lie algebras.
​8710b0b	Add ascii and unicode art for compact real forms.
​bcdd907	Streamline repr for wrapped Lie algebra elements.
​a454ac5	Better ascii/unicode art representative for Lie alg w/ structure coeffs.
​89182a3	Unicode art for Heisenberg Lie algebras.
​080907e	Unicode art for rank 2 Heisenberg Virasoro Lie algebras.
​e2be7c4	Adding unicode art for Virasoro Lie algebras.

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​0d244b5	Fixing unicode pdf build issues.
​d834c56	Merge branch 'public/typeset/unicode_sub_super-30119' into public/lie_algebras/fix_ascii_unicode_art-29205
​cd74400	Adding unicode art for affine Lie algebra elements.

[tscrim]

I merged in the latest fixes from #30119 and added one more unicode art with fixing the generic Lie algebra element one with multiline unicode art (that is a more general problem for another day).

[mkoeppe]

Not sure about this superscript comma.

[tscrim]

I am not sure either. That was the best thing I could find for that one. I don't know what else to use.

[mkoeppe]

Perhaps t⁽²⏐⁻⁴⁾ instead of t⁽²˴⁻⁴⁾, but neither is ideal.

[tscrim]

I think the current version looks closer to a comma and is more human readable without explanation. I agree neither is ideal.

[mkoeppe]

I agree, I just noticed that the appearance of what I suggested is very much font-dependent and not suitable.

[tscrim]

So is this a positive review then?

[chapoton]

is the patchbot plugin about (non-ascii characters without encoding top-line) obsolete ?

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​5dd390b

Adding utf-8 encoding header to files.

[tscrim]

Probably safer just to add them.

[mkoeppe]

Are you using u'₀' instead of '₀' to appease some style checker?

[tscrim]

Some leftover stuff from Python2 that is ignored by Python3.

[tscrim]

Hi Volker, it would be nice to get this into 9.2.