Use rich_to_bool_sgn instead of rich_to_bool when comparing outputs of mpz_cmp

[tscrim]

According to its documentation, mpz_cmp yields arbitrary positive/negative numbers, but rich_to_bool assumes the input is in [-1, 0, 1]. Thus it yields wrong results:

sage: B.<i,j,k> = QuaternionAlgebra(6)
sage: B(1) == B(-1)
False
sage: B(1) != B(-1)
False

CC: @mmasdeu @fchapoton

Component: algebra

Keywords: comparison

Author: Travis Scrimshaw

Branch/Commit: 501d5af

Reviewer: Frédéric Chapoton

Issue created by migration from https://trac.sagemath.org/ticket/29094

[tscrim]

Branch: public/algebra/fix_comparisons_mpz_cmp-29094

[sagetrac-git]

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

a06ce62

Using rich_to_bool_sgn instead of rich_to_bool with mpz_cmp.

[sagetrac-git]

Commit: a06ce62

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

501d5af

Using rich_to_bool_sgn instead of rich_to_bool with mpz_cmp.

[sagetrac-git]

Changed commit from a06ce62 to 501d5af

[fchapoton]

comment:4

looks good. But why are there 2 changes ?

[tscrim]

comment:5

That second change is because the documentation says it returns [-1, 0, 1], but test might not be in that range.

[fchapoton]

comment:6

ok, you can set to positive as soon as the lights are green

by the way, the related #28945 seems not to work with the latest beta. Maybe the same kind of issue ?

[fchapoton]

Reviewer: Frédéric Chapoton

[tscrim]

comment:8

Doesn't seem like it, but I will have to investigate further.

Thank you for doing the review.

[vbraun]

Changed branch from public/algebra/fix_comparisons_mpz_cmp-29094 to 501d5af