Analyzer for SameAs when used on Value Types.

[manfred-brands]

CodeFix to replace SameAs with IsEqualTo
Fixes #276

[mikkelbu]

Good work @manfred-brands. I've only found a handful of nitpick and nothing major.

[manfred-brands]

Thanks @mikkelbu for the review. Review changes applied and squashed. Ready for merge.

[jnm2]

What do you think about handling Assert.AreSame? Should the only diagnostic there be recommending that you switch to using the constraint model, and only then you see the problem that you're using value types?

[manfred-brands]

@jnm2 I did think of that and assumed people would convert first. But what if they disable the Classic To Constraint rules then they would get no warning.
The same holds for SameAsIncompatibleTypesAnalyzerTests which also only works on the Constraint Model.
But ConstActualValueUsageAnalyzer works on both.

[manfred-brands]

@mikkelbu I changed the Identifier from NUnit2040 into NUnit1029. The 2k range is used for ClassicAssert conversions.

[manfred-brands]

Updated the branch to support the Classic Assert.AreSame and Assert.AreNotSame methods including matching code fix to convert to AreEqual and AreNotEqual respectively.

[Dreamescaper]

@mikkelbu I changed the Identifier from NUnit2040 into NUnit1029. The 2k range is used for ClassicAssert conversions.

As far as I remember 1k range is used for Structure analyzers, and 2k - for Assertion analyzers.
I think that 2k is more applicable here.

[manfred-brands]

I see now that the SameAsIncompatibleTypesAnalyzer uses 2020. Reverted the commit.

[mikkelbu]

@Dreamescaper Is right (I plan to make this more clear in the TOC by having two separate tables.

[mikkelbu]

@manfred-brands It looks good to me (and good that you also solved the classic case), so I'm happy to merge it as it is, but if you want to merge commits or similar then please tell me.

[manfred-brands]

@mikkelbu Branch updates. Ready for merge again.

[mikkelbu]

Thanks for the work @manfred-brands, I'll merge it now.