Changes FiniteField_givaro to Python

[roed314]

Residue fields and others would like to be able to multiply inherit from finite field parents. This is the first of the two switches needed to enable that (the other being sage.rings.finite_rings.element_ntl_gf2e.FiniteField_ntl_gf2e).

This depends on #8218.

Component: algebra

Author: David Roe

Reviewer: David Loeffler, John Cremona

Merged: sage-4.4.alpha0

Issue created by migration from https://trac.sagemath.org/ticket/8332

[roed314]

Attachment: trac_8332_givaro_python.patch.gz

[roed314]

comment:2

Part of a series:

8218 -> 8332 -> 7880 -> 7883 -> 8333 -> 8334 -> 8335

I tried to make each of these mostly self contained, with doctests passing after every ticket, but I didn't entirely succeed. If you're reviewing one of these tickets, applying later tickets will hopefully fix doctest failures that you're seeing.

[loefflerd]

comment:3

This applies fine to 4.3.4.rc0 (on top of 8218), and all doctests pass on 64-bit Linux except for a tiny failure in sage/structure/parent.pyx. This is just because some random code has used GF(9,'a') as an example of a Cython object, so it's trivial to fix. I am still waiting for Sage to build on T2, and once that happens I will test this there as well, but if that passes I think this is fine to go in.

[loefflerd]

comment:4

(BTW, the aforementioned failure doesn't seem to be dealt with by any of the other tickets in the series -- none of them change that line of code.)

[loefflerd]

apply over previous patch

[loefflerd]

comment:5

Attachment: trac_8332_fix.patch.gz

Here is a tiny patch to fix that failure. With this patch installed, all tests (including long) pass on selmer.warwick.ac.uk (Ubuntu x86_64), except an unrelated existing problem in sage/schemes/elliptic_curves/heegner.py; and a selection of tests in sage/rings/finite_rings pass on t2.math.washington.edu (Solaris) as well.

[JohnCremona]

Description changed:

--- 
+++ 
@@ -1,3 +1,3 @@
 Residue fields and others would like to be able to multiply inherit from finite field parents.  This is the first of the two switches needed to enable that (the other being sage.rings.finite_rings.element_ntl_gf2e.FiniteField_ntl_gf2e).
 
-This depends on 8218.
+This depends on #8218.