Don't warn CA1802 when type's name matches enum's name

[CollinAlpert]

Currently the analyzer suggests to change a static readonly field declaration to a const if possible. This results in a compilation error if the field has the same name as the type.

Fixes #6681.

[codecov]

Codecov Report

Merging #6682 (4d4f961) into main (c6b6e10) will increase coverage by 0.00%.
The diff coverage is 100.00%.

❗ Current head 4d4f961 differs from pull request most recent head 85c163c. Consider uploading reports for the commit 85c163c to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6682   +/-   ##
=======================================
  Coverage   96.36%   96.36%           
=======================================
  Files        1395     1395           
  Lines      324940   325005   +65     
  Branches    10665    10666    +1     
=======================================
+ Hits       313135   313202   +67     
+ Misses       9241     9238    -3     
- Partials     2564     2565    +1     

[Youssef1313]

I don't know what accurate we want to handle this case, but the following is valid:

private const BindingFlags BindingFlags = (BindingFlags)5;

[CollinAlpert]

Yes, true. What do you think about extending the condition to the following:

lastField.Type.TypeKind == TypeKind.Enum && lastField.Type.Name == lastField.Name && fieldInitializerValue is IBinaryOperation or IFieldReferenceOperation

[mavasani]

private const BindingFlags BindingFlags = (BindingFlags)(BindingFlags.Public | BindingFlags.NonPublic)

This still produces the error, and the right side is an IConversionOperation. I am fine with the simple check you have in the PR. If you want to make it stricter, the correct check would be to verify that the IOperation tree rooted at fieldInitializer does not contain any IFieldReferenceOperation whose field symbol's containing type is same as lastField.Type. Either way works for me - please do add both the test cases mentioned in this comment thread to the new unit test added in the PR.

[CollinAlpert]

Done. I opted for the more extensive check.