Common base class for FiniteRankFreeModule and TensorFreeModule

[mkoeppe]

Currently TensorFreeModule (and similar classes) inherits many methods from FiniteRankFreeModule that actually only work with the base module.
Trying to use them leads to incomprehensible error messages. For example:

sage: M = FiniteRankFreeModule(ZZ, 3, name='M')
sage: M.tensor_module(2, 1).tensor_module(3, 3)
Free module of type-(3,3) tensors on the Free module of type-(2,1) tensors on the Rank-3 free module M over the Integer Ring
sage: _.an_element()
ValueError: the None has not been defined on the Free module of type-(2,1) tensors on the Rank-3 free module M over the Integer Ring

As discussed in #30229 comment:6, we introduce a common base class that only defines the methods that should be inherited by TensorFreeModule and other such classes.

Depends on #31276

CC: @egourgoulhon @tscrim

Component: linear algebra

Author: Matthias Koeppe

Branch/Commit: 451ebe5

Reviewer: Eric Gourgoulhon

Issue created by migration from https://trac.sagemath.org/ticket/34424

[mkoeppe]

comment:1

I'd just call it FiniteRankFreeModule_abstract in analogy to the existing Basis_abstract unless there are better suggestions

[mkoeppe]

Branch: u/mkoeppe/common_base_class_for_finiterankfreemodule_and_tensorfreemodule

[mkoeppe]

comment:3

Here's a basic version that works.

Some more of the methods can probably be moved to FiniteRankFreeModule_abstract -- but currently all doctests pass as is, so in this case we should add some examples that use subclasses.

---

New commits:

e76ee2f

FiniteRankFreeModule_abstract: New common base class for FiniteRankFreeModule, TensorFreeModule, ExtPower*FreeModule

[mkoeppe]

Commit: e76ee2f

[mkoeppe]

Author: Matthias Koeppe

[sagetrac-git]

Changed commit from e76ee2f to ac6d465

[sagetrac-git]

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

ac6d465

FiniteRankFreeModule_abstract: No need for _sindex

[sagetrac-git]

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

7e3ae02

FiniteRankFreeModule_abstract: No need for _output_formatter

[sagetrac-git]

Changed commit from ac6d465 to 7e3ae02

[egourgoulhon]

comment:7

Thanks for implementing this!

Replying to @mkoeppe:

I'd just call it FiniteRankFreeModule_abstract in analogy to the existing Basis_abstract unless there are better suggestions

Sounds good to me.

[egourgoulhon]

comment:8

Just one suggestion: in the docstring of FiniteRankFreeModule_abstract:

- Free module of finite rank over a commutative ring.
+ Abstract base class for free modules of finite rank over a commutative ring.

Otherwise, LGTM.

[sagetrac-git]

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

134c44f

src/sage/tensor/modules/finite_rank_free_module.py: Better docstring for FiniteRankFreeModule_abstract

[sagetrac-git]

Changed commit from 7e3ae02 to 134c44f

[egourgoulhon]

Reviewer: Eric Gourgoulhon

[egourgoulhon]

comment:10

Thanks!

[mkoeppe]

comment:11

Thank you!

[mkoeppe]

Dependencies: #31276

[mkoeppe]

comment:12

When merged with #31276, this gives a doctest failure

[sagetrac-git]

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

1ea94e1	FiniteRankFreeModule.tensor_type: New
0fe1f0d	FiniteRankFreeModule.tensor_product: New
3929d2e	FiniteRankFreeModule.tensor_power: New
bc1dc24	Merge #31276
451ebe5	src/sage/tensor/modules/finite_rank_free_module.py (FiniteRankFreeModule_abstract): Move tensor_power, tensor_product here from FiniteRankFreeModule

[sagetrac-git]

Changed commit from 134c44f to 451ebe5

[mkoeppe]

comment:14

Bots are green, back to positive review (hoping that's OK)

[vbraun]

Changed branch from u/mkoeppe/common_base_class_for_finiterankfreemodule_and_tensorfreemodule to 451ebe5