Not equal comparison for mixed forms

[Michael Jung]

At this stage, the not equal comparison fails:

sage: M = Manifold(4, 'M')
sage: X.<x,y,z,t> = M.chart()
sage: Form0 = M.mixed_form(comp=([0,0,0,0,0]))
sage: Form1 = M.mixed_form(comp=([1,0,0,0,0]))
sage: Form0 != Form1
False

(reported at ​https://ask.sagemath.org/question/52409/why-doesnt-or-work-for-mixed-forms/)

I already spotted the error: originally, richcmp is checked separately on each component and returns False if the check fails for at least one element. But for the inequality, it is enough when the check succeeds for at least one element because then both mixed forms are already unequal.

This bug gets fixed in this ticket. Furthermore, an equality check between mixed forms and differential forms will be possible.

[git]

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

​0203a6d

Trac #30108: doctest cleaned

[git]

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

​37ebad9

Trac #30108: doctest to check whether bug is fixed

[git]

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

​3312d46

Trac #30108: comment new doctest for documentation

[git]

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

​3eba4e1

Trac #30108: != bug fixed

[Michael Jung]

I have overwritten my former approach. This should be much cleaner.

[git]

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

​aa605ab

Trac #30108: eq check between scalar fields and mixed forms

[git]

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

​b7d9f47

Trac #30108: wrong import

[Michael Jung]

For the sake of consistency, and compatibility, I think that the method __eq__ should be replaced by _richcmp_ in scalarfield.py and tensorfield.py. The method _richcmp_ tries to coerce both elements into elements of a common parent first which makes most of the current checks obsolete and increases compatibility.

Eric, what do you think?

I have already opened a new ticket for this: ​https://trac.sagemath.org/ticket/30116

[git]

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

​0b93b22

Trac #30108: unused import richcmp removed

[Eric Gourgoulhon]

Replying to gh-mjungmath:

For the sake of consistency, and compatibility, I think that the method __eq__ should be replaced by _richcmp_ in scalarfield.py and tensorfield.py. The method _richcmp_ tries to coerce both elements into elements of a common parent first which makes most of the current checks obsolete and increases compatibility.

Eric, what do you think?

This sounds like a good idea, except maybe when efficiency matters, like for ManifoldPoint, since points are parameters of UniqueRepresentation objects, namely tangent spaces.

[git]

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

​4a39629

Trac #30108: use super()

[Michael Jung]

Can someone please review this ticket? There's not much change going on, and someone asked, rather indirectly, for a fix.

[Eric Gourgoulhon]

Replying to gh-mjungmath:

Can someone please review this ticket? There's not much change going on, and someone asked, rather indirectly, for a fix.

LGTM. Just waiting for the patchbot green light.

[Michael Jung]

Replying to egourgoulhon:

Replying to gh-mjungmath:

Can someone please review this ticket? There's not much change going on, and someone asked, rather indirectly, for a fix.

LGTM. Just waiting for the patchbot green light.

Patchbot appears to be green now. Thank you.

[Eric Gourgoulhon]

Good to go!

[Michael Jung]

Perfect. Thank you! :)

[git]

Branch pushed to git repo; I updated commit sha1 and set ticket back to needs_review. New commits:

​6f0f537

Trac #30108: Merge branch 'develop' into t/30108/inequality_test_mixed