Elliot Hillary

It is worth noting that unlike Java projects, Delphi projects are often designed to target a wide range of versions. There might be a value loss if we threw out...

Hi @Indigo744, Thanks for re-raising! Definitely open to changing some of this rule's behaviour. My two cents: I think there's two mindsets that this rule can be approached from. On...

> I'm not saying that you should blindly follow what others are doing, but it seems that this exception is commonly accepted everywhere in the Sonar ecosystem (and it is...

Hi @Indigo744, Agreed - let's make this rule match the other analyzers. A PR would be very welcome! The main files you'd need to take a look at are: *...

Some more observations: * Explicitly specifying the `register` calling convention results in the same type as specifying no calling convention at all * `near`, `far`, and `export` are all syntactically...

> * require a containment check to be enclosed only if the parent expression is a unary `not` I would probably go with this approach. Delphi's operator precedence is not...

An alternate approach to this rule could be to require all bitwise `not` invocations to be contained in parentheses. They're confusing and take precedence over logical `not`, which is almost...

The cursed code makes a good point 😢 In the PR description, > As a gotcha, the rule must not raise an issue if the "dead" statement is a goto...

Thanks for the rule proposal! I think a rule checking if records are initialised during their constructors would be a really solid addition to the default quality profile - it's...

Hmm, it could be a little confusing to be named so similarly to `VariableInitialization`. > It's actually not specific to constructors though, as this rule should also target the Initialize...