[with additional patch] problem with copy() on sage.rings.integer_mod.IntegerMod_gmp

[dfdeshom]

John Cremona:

sage: a=[Mod(2,next_prime(2^n)) for n in range(28,35)]
sage: [type(x) for x in a]

[<type 'sage.rings.integer_mod.IntegerMod_int64'>,
 <type 'sage.rings.integer_mod.IntegerMod_int64'>,
 <type 'sage.rings.integer_mod.IntegerMod_int64'>,
 <type 'sage.rings.integer_mod.IntegerMod_gmp'>,
 <type 'sage.rings.integer_mod.IntegerMod_gmp'>,
 <type 'sage.rings.integer_mod.IntegerMod_gmp'>,
 <type 'sage.rings.integer_mod.IntegerMod_gmp'>]

sage: [copy(x) for x in a]
[2, 2, 2, None, None, None, None]

sage: [deepcopy(x) for x in a]
[2, 2, 2, 2, 2, 2, 2]

Component: basic arithmetic

Issue created by migration from https://trac.sagemath.org/ticket/2571

[JohnCremona]

Attachment: 8950.patch.gz

[JohnCremona]

comment:1

To fix this I added a return statement to the __copy__() function of IntegerMod_gmp. Patch attached (based on 2.10.4.rc0)

[dfdeshom]

comment:2

While you're editing that function, would you mind adding a doctest or 2 to it? That would help with the overall 3.0 goal of >50% function doctest coverage

[JohnCremona]

apply after 8950.patch

[JohnCremona]

comment:3

Attachment: 8951.patch.gz

As suggested, a few doctests have been added, in the second patch.

[sagetrac-jbmohler]

comment:5

These patches fix the bug and add and improve documentation bits. I've tested everything I could see that could go wrong (including the copy and sqrt/square_root change) and all looks good to me.

This is a positive review to both patches!

[sagetrac-mabshoff]

comment:6

Merged both patches in Sage 2.11.alpha0