🚧 This library is currently under construction. 🚧

eff is a work-in-progress implementation of an extensible effect system for Haskell, a general-purpose solution for tracking effects at the type level and handling them in flexible ways. Compared to other effect systems currently available, eff differentiates itself in the following respects:

• eff is really fast. It is designed specifically to cooperate with the GHC specializer to allow effect handlers to be aggressively optimized, even in highly polymorphic code. Performance should be indistinguishable from mtl, slightly faster than fused-effects, and significantly faster than freer-simple, polysemy, and other libraries based on free or free-like constructions.

  Traditional effect system microbenchmarks fail to capture the performance of real code, as they are so small that GHC often ends up inlining everything. In real programs, GHC compiles most effect-polymorphic code via dictionary passing, not specialization, causing the performance of other effect systems to degrade beyond what microbenchmarks would imply. eff takes care to allow GHC to generate efficient code without the need for whole-program specialization.

• eff is low-boilerplate and easy to use, even without Template Haskell or any generic programming. Effects in eff are just ordinary typeclasses on monads, just like in mtl, and a small bit of extra information is maintained in the type system to guide instance selection to eliminate the need for any overlapping instances. Defining effect handlers in terms of existing handlers does not require defining any new monad transformers.

• eff is expressive, providing support for both first-order/algebraic effects and higher-order/scoped effects, like fused-effects and polysemy (but unlike freer-simple).

• eff integrates well with the existing transformers ecosystem. Effect handlers in eff are ordinary monad transformers, and many built-in effects are handled by the ordinary transformers from the transformers package. As in fused-effects, existing transformers can be easily adapted into effect handlers, but eff goes a step further by allowing existing mtl-style typeclasses to be used as effects, just by defining a couple small instances.

If you are already familiar with mtl and the “mtl-style” approach to handling effects, eff may be approximately described as “mtl but without the quadratic instances problem.”

To illustrate just how easy it is to define and handle effects in eff, the following code example includes 100% of the code necessary to define a custom FileSystem effect and two handlers, one that runs in IO and another that uses an in-memory virtual file system:

import Prelude hiding (readFile, writeFile)

import qualified System.IO as IO

import Control.Effect
import Control.Effect.Error
import Control.Effect.State
import Control.Monad.IO.Class
import Data.Functor.Identity

-- ------------------------------------------------------------------
-- the FileSystem effect

class Monad m => FileSystem m where
  readFile :: FilePath -> m String
  writeFile :: FilePath -> String -> m ()

instance (Monad (t m), Send FileSystem t m) => FileSystem (EffT t m) where
  readFile path = send @FileSystem (readFile path)
  writeFile path contents = send @FileSystem (writeFile path contents)

-- ------------------------------------------------------------------
-- the FileSystemIOT handler

data FileSystemIO
type FileSystemIOT = HandlerT FileSystemIO '[]
type instance Handles FileSystemIOT eff = eff == FileSystem

instance MonadIO m => FileSystem (FileSystemIOT m) where
  readFile path = liftIO $ IO.readFile path
  writeFile path contents = liftIO $ IO.writeFile path contents

runFileSystemIO :: EffT FileSystemIOT m a -> m a
runFileSystemIO = runHandlerT . runEffT

-- ------------------------------------------------------------------
-- the FileSystemPureT handler

-- | A mapping from file paths to file contents.
type VirtualFileSystem = [(FilePath, String)]

data FileSystemPure
type FileSystemPureT = HandlerT FileSystemPure '[StateT VirtualFileSystem]
type instance Handles FileSystemPureT eff = eff == FileSystem

instance Error String m => FileSystem (FileSystemPureT m) where
  readFile path = HandlerT $ do
    fileSystem <- get
    case lookup path fileSystem of
      Just contents -> pure contents
      Nothing       -> throw ("readFile: no such file " <> path)

  writeFile path contents = HandlerT $ do
    fileSystem <- get
    -- add the new file and remove an old file with the same name, if it exists
    put ((path, contents) : filter ((/= path) . fst) fileSystem)

runFileSystemPure :: Monad m => EffT FileSystemPureT m a -> m a
runFileSystemPure = evalState [] . runHandlerT . runEffT

That’s it. No Template Haskell, no need to define any new monad transformers, and no need to define any lifting instances for each handler. For a thorough explanation of how the above example works, see the eff documentation.

eff is a work in progress, and many things are not yet implemented. The set of built-in effects currently provided is very small. However, all the existing functionality works, and it has been designed to support those extensions, so I do not anticipate any difficulty supporting them.

This library is also missing a benchmark suite. I have a small set of microbenchmarks I have been using to test out various scenarios and edge cases of different effect libraries, but they need to be cleaned up and added to this repository, and a set of less synthetic benchmarks are also important to assess real-world performance. If you have a small but non-trivial program where differences in effect system performance are significant, I would be much obliged if you could share it to build a more realistic set of effect system benchmarks.

All code in eff is original in the sense that it was not taken directly from other libraries, but the vast majority of it is directly inspired by the existing work of many others. The following is a non-exhaustive list of people and works that have had a significant impact, directly or indirectly, on eff’s design and implementation:

• Oleg Kiselyov, Amr Sabry, and Cameron Swords — Extensible Effects: An alternative to monad transfomers
• Oleg Kiselyov and Hiromi Ishii — Freer Monads, More Extensible Effects
• Nicolas Wu, Tom Schrijvers, and Ralf Hinze — Effect Handlers in Scope
• Nicolas Wu and Tom Schrijvers — Fusion for Free: Efficient Algebraic Effect Handlers
• Andy Gill and other contributors — mtl
• Rob Rix, Patrick Thomson, and other contributors — fused-effects
• Sandy Maguire and other contributors — polysemy