Question

It seems that Haskell has established several naming conventions around monads.

Examples:

• appending T to the end to obtain the name of the monad transformer (e.g. Reader -> ReaderT)
• using runXXX to perform a monad computation (e.g. runST, runReader)
• liftXXX for various values of XXX

Are there other naming conventions?

Answer 1

• runX m where m :: X a will run the X monad and return the “side effect” along with the monad result, a.

• evalX m will run the computation and return the result, a.

• execX m will run the computation and return the “side effect” but not the result.

• The lifts come in various flavors that can be a bit too tricky for me to want to explain them in a SO answer. You should probably know lift and liftIO and be aware of / eventually seek out the other variants such as liftWith and liftBaseWith. See, for example, EZYang’s posting on the topic.

• appending a T after the monad name implies transformer. Appending an M after a function name implies it is monadic. Appending an _ implies the result is ignored.

• All other suffixed letters mean “use hoogle”.