How do write this in do notation?

I'm working through "Write yourself a scheme interpreter in 48 hours" and one exercise is to write a function using do notation. This is the function:开发者_开发知识库

parseNumber :: Parser LispVal
parseNumber = liftM (Number . read) $ many1 digit

I tried the following, but that didn't work:

parseNumber2 :: Parser LispVal
parseNumber2 = do digits <- many1 digit
                  return $ liftM $  (Number . read) digits

Never mind. I indented the code incorrectly

---

Once you've unpacked digits, you don't need the liftM anymore. So it's just:

return $ Number $ read digits

---

I hopped on the #haskell IRC channel and asked lambdabot:

11:25 < dmwit> ?src liftM
11:25 < lambdabot> liftM f m1 = do { x1 <- m1; return (f x1) }
11:27 < dmwit> ?src ($)
11:27 < lambdabot> f $ x = f x

(If you don't want to ask lambdabot, you can also ask Hoogle and click the source link in the documentation.)

So, let's fill in f and m1:

liftM (Number . read) $ many1 digit
    ={definition of ($)}
liftM (Number . read) (many1 digit)
    ={definition of liftM}
do { x1 <- (many1 digit); return ((Number . read) x1) }
    ={expand to layout form, eliminate redundant parentheses}
do
    x1 <- many1 digit
    return ((Number . read) x1)

You can write half a dozen variations of this of varying syntactic aesthetic, but here's my favorite, after expanding the definitions of (.) and ($) a few times:

    ={definitions of (.) and ($)}
do
    x1 <- many1 digit
    return . Number . read $ x1