Add Analyzer and CodeFix for duplicate method aliases (ORLEANS0011)

[ledjon-behluli]

cc @ReubenBond
close #8661

Analyzer:

Fixer:

Tests:

Microsoft Reviewers: Open in CodeFlow

[ledjon-behluli]

Note, this PR is based on #8643 (which has not yet been merged), thats why it contains the GenerateAliasAttributesAnalyzer code

[ReubenBond]

1:

2:

3:

Should this be Echo2 instead of Echo11?

[ledjon-behluli]

@ReubenBond If "Fix for Document" than yes, it changes to 'Echo1' and 'Echo2'

But if you change them 1-by-1 its actually quite the problem, let me explain why:
A "duplicate" is defined only if its the same alias (as in the same text), if we analyzed all other "EchoX" in order to determine the 'X' its going to get very ambiguous what really is a "Echo".

Thats why if the user applies the code fix for the document, and there are multiple "Echo"s it will go 1, 2, 3. But if the above is true, than we look the only "true" duplicate which would be "Echo" and assign the next number which would be "Echo1"

We need to always think that the user's may always name the alias themselves i.e: Echo_, Echo__, Echo_1 ...
I hope I make sense here

[ReubenBond]

In your example, I would expect the result to be "Echo1". If there was a method with the alias "Echo1" already, then I would expect the result to be "Echo2", and so on until there is no conflict. The method name does not matter (the developer could have aliased a method named GetFoo as "Echo" and the result should be the same), but the resulting alias value does.

[ledjon-behluli]

"The method name does not matter" - I see what you mean, i'll address it 👍

[ledjon-behluli]

@ReubenBond Maybe it makes sense to hold off on approving the other PRs which are based on this (since some common code is shared), until this is resolved and i'll port the change to the other 2 PRs. You could still review and leave comments ofc.

[ledjon-behluli]

@ReubenBond It should work as expected now! I've ported the fix to the other PRs too, so feel free to continue when you can

[ledjon-behluli]

@ReubenBond are we good with this, so we can move onto the other?

[ReubenBond]

I am tweaking the suffix incrementing code a little, but otherwise LGTM

[ReubenBond]

@ledjon-behluli if you can rebase, I'm ready to merge.

[ReubenBond]

@ledjon-behluli I performed a squash merge + force push of your latest, since it included changes from main which made final review harder.

[ledjon-behluli]

@ReubenBond whatever it takes, stuff got mangled up

[ReubenBond]

Includes #8664 & #8672