SA1126: PrefixCallsCorrectly

[sharwell]

http://www.stylecop.com/docs/SA1126.html

[sharwell]

It seems like this diagnostic is not needed now that we are using Roslyn for analysis (its SemanticModel tells us everything we need to know).

[sharwell]

I am planning to close this issue as wontfix because it is so vaguely worded. If we are in agreement on this, any specific functionality originally part of it can be proposed and implemented as part of new well-defined rules.

[vweijsters]

Sounds like a good plan.

As far as I could determine, the following scenarios should still be addressed:

• Usage of Equals(objA, objB) and ReferenceEquals(objA, objB)
• A overridden method where the analyzer cannot decide if the developer intended to call the this implementation or the base implementation

[sharwell]

@vweijsters How might those scenarios be addressed?

💭 For the second one, all of our analyzers currently assume that the intended target is the method that was actually called. That seems to be a reasonable assumption for a style checker.

[AArnott]

Is SA1126 the rule responsible for enforcing folks prefix this. everywhere? If that's the case, that's a valuable rule that I am eager to resume enforcing once StyleCop.Analyzers supports.

It's fantastic for when you write thread-safe code because it makes it abundantly clear whether you're reading or modifying instance state, or calling an instance method that might.

Please, let's support this rule.

[sharwell]

@AArnott That would be SA1101, and it's been available with a code fix since Alpha 3. In this project we actually have the severity for violations set to Error.

[AArnott]

I see. In that case, I agree with not implementing this diagnostic.

[naamunds]

@sharwell - is there currently a rule for enforcing that a reference to a class-level member is prefixed with the class's name? This would be analogous to SA1101, but for static methods and variables, const variables, etc.

For example, if I define class Person with const string JuniorSuffix = "Jr.", I would want to enforce that references to JuniorSuffix within the Person class are written as Person.JuniorSuffix rather than just JuniorSuffix.

[sharwell]

@naamunds I am not aware of any such rule.