Refactor backend_normaliz

[mkoeppe]

Preparation for #25097.

TODO: Make pynormaliz a feature.

CC: @videlec @jplab

Component: geometry

Author: Matthias Koeppe

Branch/Commit: ee5ff41

Reviewer: Jean-Philippe Labbé

Issue created by migration from https://trac.sagemath.org/ticket/27716

[mkoeppe]

Branch: public/normaliz-backend-refactoring

[sagetrac-git]

Commit: df4f20e

[sagetrac-git]

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

df4f20e

Refactor normaliz backend

[mkoeppe]

comment:3

This branch has a part of the changes on #25097 without dependencies.

[videlec]

comment:4

Do you intend to make #27682 a dependency?

[mkoeppe]

comment:5

No, not for this ticket

[sagetrac-git]

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

0dfec03

_format_function_call: Fix doctest

[sagetrac-git]

Changed commit from df4f20e to 0dfec03

[videlec]

comment:7

I think in #27682 there is a need to touch the normaliz backend due to some changed in the (not yet released) PyNormaliz. So there might be a small conflict.

[jplab]

comment:8

Replying to @videlec:

I think in #27682 there is a need to touch the normaliz backend due to some changed in the (not yet released) PyNormaliz. So there might be a small conflict.

Yes, Matthias merged cleanly so far. It's coming...

[jplab]

comment:9

Here are a few things:

• to get 100% coverage, the function _nmz_result should be tested. Perhaps also with a function name that does not exist?
• sage.matrix.constructor.matrix imported but not used
• pep8 around here, and blankline at end of file...........

921             if  box_points<threshold:                                                                                                                                                                          
922                 from sage.geometry.integral_points import rectangular_box_points
923                 return rectangular_box_points(list(box_min), list(box_max), self)

Otherwise, it looks quite good, let's see what the bot says. vdelecroix mentioned in #25091 in comment #25091 comment:69 that

import PyNormaliz

is not desirable... and we should use feature. I would not do this here but in a follow up ticket.

[jplab]

comment:10

There seems to be only two small things:

Pyflakes:

+src/sage/geometry/polyhedron/backend_normaliz.py:33: 'sage.matrix.constructor.matrix' imported but unused
+
+src/sage/geometry/polyhedron/parent.py:21: 'sage.geometry.polyhedron.base.Polyhedron_base' imported but unused

Python3:

+inside file:  b/src/sage/geometry/polyhedron/backend_normaliz.py
+@@ -496,25 +607,36 @@ class Polyhedron_normaliz(Polyhedron_base):
++        for key, value in sorted(data.iteritems()):

[sagetrac-git]

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

b370268	pyflakes fixes
328a04a	_nmz_result: Add doctest

[sagetrac-git]

Changed commit from 0dfec03 to 328a04a

[jplab]

Reviewer: Jean-Philippe Labbé