Multiple bugs in Polynomial.reverse(degree)

[mezzarobba]

Polynomial.reverse(d) over ℚ is inconsistent with the generic implementation. The name of the optional argument is different, and its interpretation is slightly different:

sage: x = polygen(QQ)
sage: y = polygen(QQbar)
sage: (x+1).reverse(1)
1
sage: (y+1).reverse(1)
x + 1

In addition, the documentation of the generic reverse() (which arguably should specify what reverse() is supposed to do for sage polynomials) incorrectly states that “the reverse polynomial will have the specified degree”:

sage: (y^2).reverse(5)
x^3

Finally, the generic implementation is buggy when the optional argument is zero.

Component: commutative algebra

Author: Marc Mezzarobba

Branch/Commit: 554500b

Reviewer: Vincent Delecroix

Issue created by migration from https://trac.sagemath.org/ticket/21194

[mezzarobba]

Commit: 160f6b3

[mezzarobba]

Author: Marc Mezzarobba

[mezzarobba]

Branch: u/mmezzarobba/21194-reverse

[sagetrac-git]

Changed commit from 160f6b3 to 08cee24

[sagetrac-git]

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

08cee24

#21194 Fix multiple bugs in Polynomial.reverse()

[videlec]

Reviewer: Vincent Delecroix

[videlec]

comment:5

sage: x = polygen(QQ)
sage: (x+1).reverse(2**64-1)
0

Your casting should be <unsigned long> (degree + 1).

[sagetrac-git]

Changed commit from 08cee24 to 554500b

[sagetrac-git]

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

554500b

#21194 Fix handling of incorrect input

[mezzarobba]

comment:9

Thanks!

[vbraun]

Changed branch from u/mmezzarobba/21194-reverse to 554500b