Fix the homset category initialization for ModularAbelianVariety's homspaces

[nthiery]

Before the patch, the following was wrong (probably introduced by #9138):

    sage: End(J0(37)).homset_category()
    Join of Category of hom sets in Category of sets and Category of rings

After the patch:

    sage: End(J0(37)).homset_category()
    Category of modular abelian varieties over Rational Field

In both cases, we have, as desired:

    sage: End(J0(37)).category()
    Join of Category of hom sets in Category of sets and Category of rings

By the way, this removes a direct call to _Hom_, using Hom instead, preparing for #11935.

Note: #11935 depends on this ticket.

CC: @sagetrac-sage-combinat

Component: modular forms

Keywords: abelian varieties

Author: Nicolas M. Thiéry

Reviewer: Simon King

Merged: sage-5.1.beta0

Issue created by migration from https://trac.sagemath.org/ticket/12875

[nthiery]

comment:2

For the record: all tests passed on 5.0.beta13, with a couple unrelated sage-combinat patches just above (except for one doctest failure in sagedoc caused by those patches).

[simon-king-jena]

comment:3

Some small criticism: The commit message does not mention the ticket number. Apart from that, the patch looks fine, and I am now running doctests.

[nthiery]

Attachment: trac_12875-category-fix_abvar_homspace-nt.patch.gz

[nthiery]

comment:4

Replying to @simon-king-jena:

Some small criticism: The commit message does not mention the ticket number.

Fixed in the updated patch. Thanks for catching this, and for the quick review!

[simon-king-jena]

Reviewer: Simon King

[simon-king-jena]

comment:5

Thank you for updating the commit message! All tests pass, with sage-5.1.notebook, the patch applied after the patches from #12808 (I was to lazy to remove them). The patch looks fine, thus, I give it a positive review.