Polysemy: Introduction to interpreters

As seen in the previous log we have an expression as follows:

import Polysemy
import Polysemy.Trace

displayFile :: FilePath -> Sem '[Trace, Embed IO] Int
displayFile path = do
  trace $ "Displaying " <> path
  content <- embed $ readFile path
  embed $ putStr content
  return $ length content

In order to do something useful we need to interpret them in term of IO:

intrDisplayFile :: Sem '[Trace, Embed IO] a -> IO a
intrDisplayFile =
  runM
    . traceToStdout

It can be read that way:

• traceToStdout interprets Trace (after the application we have Sem '[Embed IO] a)
• runM lowers the Embed IO

One of the benefit of effects systems is to be able to change the behavior (via the interpretation) without changing the expression.

As such Trace could be:

• Redirected to stderr with traceToStderr
• Redirected to a Handle with traceToHandle
• Dropped with ignoreTrace
• Accumulated with runTraceList
• etc.

See the full the code here.