Currently, the way #[test] is handled is proc macro deletes it when processing datatest attribute.
There are two reason #[test] attribute is desirable:
- Better integration with IDEs which don't know about
#[datatest], but can handle #[test] (for example, allow running test by clicking on it). I use IntelliJ as a test bed 😛
- Differentiation between test tests and bench tests.
The issue is that if #[test] goes first, it will be handled by Rust built-in test harness, which is undesirable. Need to figure out a way to disable that -- potentially, #4 will solve that, too.
Currently, the way
#[test]is handled is proc macro deletes it when processingdatatestattribute.There are two reason
#[test]attribute is desirable:#[datatest], but can handle#[test](for example, allow running test by clicking on it). I use IntelliJ as a test bed 😛The issue is that if
#[test]goes first, it will be handled by Rust built-in test harness, which is undesirable. Need to figure out a way to disable that -- potentially, #4 will solve that, too.