test: get rid of ring_to_mat

[fingolfin]

The 'similar' methods being tested are generic and defined &
tested in AA.

Getting rid of this makes it somewhat less annoying to run the
affect test files interactively.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 88.37%. Comparing base (1e48bb0) to head (5d20d83).
Report is 22 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1902      +/-   ##
==========================================
+ Coverage   87.25%   88.37%   +1.11%     
==========================================
  Files          97       97              
  Lines       35651    42376    +6725     
==========================================
+ Hits        31109    37451    +6342     
- Misses       4542     4925     +383     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[lgoettgens]

I agree that currently these tests are more or less useless. But these would make it very easy to test #1826 (once that is fixed) by just adding @test t isa M twice in each of the for-loops.

[fingolfin]

Sure, but I think this should be handled by adding such a test to the ring conformance tests in AA.

[fingolfin]

To move forward for this, my solution would be to integrate a similar test into the conformance test suite. Specifically into test_MatSpace_interface.

To handle conversion to different rings, we could use something similar to the adhoc_partner_rings to provide a list of "partner" rings against which to perform the similar tests. Note that M should be redundant and just the same as dense_matrix_type(R).