difference of behavior in polyhedra with different backends

[videlec]

The following error used to occur:

sage: from sage.geometry.polyhedron.parent import Polyhedra
sage: P_ppl = Polyhedra(ZZ, 3, 'ppl')
sage: P_cdd = Polyhedra(ZZ, 3, 'cdd')
sage: P_ppl([[],[],[]], None)
The empty polyhedron in ZZ^3
sage: P_cdd([[],[],[]], None)
Traceback (most recent call last):
...
TypeError: can't multiply sequence by non-int of type 'NoneType'

Apparently it got fixed along the way and we add doctests verifying it.

CC: @vbraun @jplab @mo271

Component: geometry

Keywords: empty polyhedron, cdd

Author: Jonathan Kliem

Branch/Commit: 96dad18

Reviewer: Laith Rastanawi

Issue created by migration from https://trac.sagemath.org/ticket/19803

[jplab]

comment:1

This might be related to #17339.

[jplab]

comment:2

This is because the cdd backend does not have an _init_empty_polyhedron method. That should be an easy fix.

The problem in the description is that it tries to get the common length of elements in the v-representation while they are all None. This should be grabbed beforehand and initialize an empty polyhedron like we do in the other cases.

[kliem]

Changed keywords from none to empty polyhedron, cdd

[kliem]

Commit: 96dad18

[kliem]

Branch: public/19803

[kliem]

comment:4

Looks like this got fixed along the way.

---

New commits:

96dad18

add doctests that cdd can handle empty polyhedron now

[kliem]

Author: Jonathan Kliem

[LaisRast]

Reviewer: Laith Rastanawi

[vbraun]

Changed branch from public/19803 to 96dad18