Refers to regression introduced in 7b4a9945

A lambda all_items_equal() replaced a reduce() that was broken for potential
3+-way merges. A reduce(operator.eq, ...) accumulates bools and can't
generically check equality of all items in a sequence:

>>> bool(reduce(operator.eq, [('migrations', '0001_initial')] * 3))
False

The code now counts the number of common ancestors to calculate slice offsets
for the branches. Each branch shares the same number of common ancestors.

The common_ancestor for loop statement had incomplete branch coverage.

django.core.exceptions.ValidationError.messages() and
django.db.backends.schema.BaseDatabaseSchemaEditor._alter_field():
Replaced reduce(operator.add, ...) w/uncoupled, explicit sum()

Added relabeled_clone() method to sql.Query to fix the problem. It
manifested itself in rare cases where at least double nested subquery's
filter condition might target non-existing alias.

Thanks to Trac alias ris for reporting the problem.

This removes the concept of equality between operations to guarantee
compatilibity with Python 3.

Python 3 requires equality to result in identical object hashes. It's
impossible to implement a unique hash that preserves equality as
operations such as field creation depend on being able to accept
arbitrary dicts that cannot be hashed reliably.

Thanks Klaas van Schelven for the original patch in
13d613f8.

Refs #23748.

Refs #23748.

Removed inaccurate info about partial evaluation after refs #18702.
Added information on modifying sliced QuerySets; refs #22503.

Thanks Claude for writing the patch.

Also simplified the logic under the assumption that a false-ish object
won't have an exception_reporter_filter attribute.

load_hashers cached its result regardless of its password_hashers
argument which required fragile cache invalidation. Remove that
argument in favor of @override_settings and triggering cache
invalidation with a signal.

Thanks to yamila-moreno for the idea of a skip message.