RealSet: Use self.__class__.__base__ in methods when constructing a new RealSet

[mkoeppe]

... instead of hardcoding RealSet. This is preparation for #34278.

(split out from #32170 - sagemath/sagetrac-mirror@develop...public/32170)

Depends on #32181

CC: @yuan-zhou @tscrim

Component: basic arithmetic

Author: Binshuai Wang, Matthias Koeppe, Yuan Zhou

Branch/Commit: u/yzh/realset__use_self___class___in_methods_when_constructing_a_new_realset @ eef3886

Issue created by migration from https://trac.sagemath.org/ticket/34277

[mkoeppe]

Branch: u/mkoeppe/realset__use_self___class___in_methods_when_constructing_a_new_realset

[mkoeppe]

Author: Binshuai Wang, ...

[mkoeppe]

comment:2

Cherry-picked from #32170

---

New commits:

5f60c49

make a whole change for RealSet class with using 'self.__class__' and '@classmethod'

[mkoeppe]

Commit: 5f60c49

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. Last 10 new commits:

44926c8	better pyflakes, better handle of RealSet(x != oo)
c9d4746	update RealSet.union and RealSet.intersection to handle several real sets
ebd3b77	Better pygments style
73e5aa3	Make white logo transparent to match with furo
523b49a	Merge branch 'u/klee/34252' of git://trac.sagemath.org/sage into t/32181/realset__faster_operations_by_scan_line__merging__techniques
a864fdc	fix typo
3f6a0ac	revive RealSet.normalize
79f2014	revise the docstring of RealSet.is_connected
1ef7dd1	remove unnecessary bool()
a856e5d	Merge #32181

[sagetrac-git]

Changed commit from 5f60c49 to a856e5d

[mkoeppe]

Dependencies: #32181

[mkoeppe]

Changed author from Binshuai Wang, ... to Binshuai Wang, Matthias Koeppe, ...

[mkoeppe]

comment:4

Merged #32181 and resolved merge conflicts.

Haven't checked whether everything that needs to be changed to __class__ has been changed.

[yuan-zhou]

comment:6

Bug example:

sage: r = RealSet(2,10)
sage: RealSet(x > 2).intersection(RealSet(x < 10)) is RealSet(r[0], normalized=True)

Expect True. Got False.

[yuan-zhou]

comment:7

In the commit 5f60c49 from #32170,
why is it sometimes isinstance(other, RealSet) but isinstance(right, self.__class__) in other places? Why is it RealSet.point(...) and RealSet._prep(...) sometimes but self.__class__.real_line() in other places?

[yuan-zhou]

comment:8

In def union(self, *real_set_collection) and intersection(self, *real_set_collection),
why is it

intervals = tuple(self._scan_to_intervals(scan, lambda i: i > 0))

instead of self.__class__._scan_to_intervals?

(It was RealSet._scan_to_intervals)

[mkoeppe]

comment:9

Replying to @yuan-zhou:

In def union(self, *real_set_collection) and intersection(self, *real_set_collection),
why is it

intervals = tuple(self._scan_to_intervals(scan, lambda i: i > 0))

instead of self.__class__._scan_to_intervals?

(It was RealSet._scan_to_intervals)

_scan_to_intervals is a staticmethod, so self._scan_to_intervals(...) is the same thing as self.__class__._scan_to_intervals.

[mkoeppe]

comment:10

Replying to @yuan-zhou:

In the commit 5f60c49 from #32170,
why is it sometimes isinstance(other, RealSet) but isinstance(right, self.__class__) in other places? Why is it RealSet.point(...) and RealSet._prep(...) sometimes but self.__class__.real_line() in other places?

As I said, I haven't checked whether all necessary changes have been made

[yuan-zhou]

Changed branch from u/mkoeppe/realset__use_self___class___in_methods_when_constructing_a_new_realset to u/yzh/realset__use_self___class___in_methods_when_constructing_a_new_realset

[yuan-zhou]

Changed commit from a856e5d to c97b9f7

[yuan-zhou]

comment:12

Unfortunately, the previous bug example remains.

sage: r = RealSet(2,10)
sage: r is r.__class__(r)
True
sage: r is r.__class__(r[0])
False
sage: r is r.__class__(r[0], normalized=True)
False

---

New commits:

c97b9f7

change RealSet(...) constructor to self.__class__(...)

[mkoeppe]

Changed author from Binshuai Wang, Matthias Koeppe, ... to Binshuai Wang, Matthias Koeppe, Yuan Zhou

[sagetrac-git]

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

42f4dd6

Use self.__class__.__base__ in methods when constructing a new RealSet

[sagetrac-git]

Changed commit from c97b9f7 to 42f4dd6

[mkoeppe]

comment:17

More changes in the same direction can be cherry-picked from #34278

[sagetrac-git]

Changed commit from 42f4dd6 to eef3886

[sagetrac-git]

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

c3893aa	src/sage/sets/real_set.py: Use self.METHOD, cls.METHOD instead of RealSet.METHOD; use super()
dab16c2	src/sage/sets/real_set.py: Make normalize a @classmethod
0574efd	src/sage/sets/real_set.py (RealSet_base.__init__): Check normalized
eef3886	src/sage/sets/real_set.py: Use self.METHOD, cls.METHOD instead of RealSet.METHOD; use super() (fixup)

[yuan-zhou]

comment:19

cherry-picked the relevant commits from #34278.