Implement NormalizeLineEndings

[Bartleby2718]

This fixes #4975.

[Bartleby2718]

Only two of these are used, but I just created singletons for all 6 StringComparison values.

[Bartleby2718]

CurrentCultureIgnoreCase and Ordinal to be consistent with the existing code.

[Bartleby2718]

Updated because I found Test Explorer misleading without this.

[manfred-brands]

Thanks @Bartleby2718 for your contribution.
You already found your casing problem in parameters and fields.
It would have been easier if you had just fixed the instead of putting a comment on your own PR that you need to fix it.

The actual comparer looks good.
The main change I would like is to rename the word "Normalize" and "Normalizing"
One what is normal for one might not be normal for another.
But more to match the other modifiers that it "Ignore" an aspect.
It doesn't change any strings but ignores different style line endings.

[manfred-brands]

Thanks @Bartleby2718. mostly fine.
A few things remain.

[manfred-brands]

Looking good now.

[manfred-brands]

@Bartleby2718 There is a serious performance degradation:

Warning : NUnit.Framework.Tests.Constraints.UniqueItemsConstraintTests.PerformanceTests_FastPath(System.Collections.Generic.List`1[System.String],True)
  Warn.Unless(() =>
            {
                if (ignoreCase)
                    Assert.That(values, Is.Unique.IgnoreCase);
                else
                    Assert.That(values, Is.Unique);
            }, HelperConstraints.HasMaxTime(100))
Elapsed time of 3696ms exceeds maximum of 100ms
   at NUnit.Framework.Tests.Constraints.UniqueItemsConstraintTests.PerformanceTests_FastPath(IEnumerable values, Boolean ignoreCase) in /_/src/NUnitFramework/tests/Constraints/UniqueItemsConstraintTests.cs:line 186

I have traced it down to a subtle line I missed in my review: comparer.GetHashCode() which should have been comparer.GetHashCode(s)
Putting all items in the same 'bucket' in a HashSet slowed down the test dramatically.