Return to Answer

I am building on bisserlis' answer here. interpret uses a lot of repetition that can be reduced with case..of. That allows a where clause which factors out common code that modifies memory and simply advances the instruction pointer:

interpret m@(ZL _ x _) i@(ZL _ cmd _) =
 case cmd of
 '>' -> memory znxt
 '<' -> memory zprv
 '+' -> memory zinc
 '-' -> memory zdec
 '[' -> interpret m (if x /= 0 then znxt i else zjmp znxt 0 i)
 ']' -> interpret m (if x == 0 then znxt i else zjmp zprv 0 i)
 '.' -> putChar (toEnum x)>> memory id
 ',' -> getChar>>= memory . zapp . const . fromEnum
 _ -> memory id -- Comment `Char`
 where
 memory :: (ZL Int -> ZL Int) -> IO ()
 memory f = interpret (f m) (znxt i)

I would use longer names for znxt etc., then you might consider renaming zapp to underHead or atPtr and then your code might read like this (I am mixing styles here):

 '>' -> memory advancePtr
 '<' -> memory moveHeadLeft
 '+' -> memory (underHead (+1))
 '-' -> memory (atPtr (subtract 1))

Reads almost like the specification!

I am building on bisserlis' answer here. interpret uses a lot of repetition that can be reduced with case..of. That allows a where clause which factors out common code that modifies memory and simply advances the instruction pointer:

interpret m@(ZL _ x _) i@(ZL _ cmd _) =
 case cmd of
 '>' -> memory znxt
 '<' -> memory zprv
 '+' -> memory zinc
 '-' -> memory zdec
 '[' -> interpret m (if x /= 0 then znxt i else zjmp znxt 0 i)
 ']' -> interpret m (if x == 0 then znxt i else zjmp zprv 0 i)
 '.' -> putChar (toEnum x)>> memory id
 ',' -> getChar>>= memory . zapp . const . fromEnum
 _ -> memory id -- Comment `Char`
 where
 memory :: (ZL Int -> ZL Int) -> IO ()
 memory f = interpret (f m) (znxt i)

I would use longer names for znxt etc., then you might consider renaming zapp to underHead or atPtr and then your code might read like this (I am mixing styles here):

 '>' -> memory advancePtr
 '<' -> memory moveHeadLeft
 '+' -> memory (underHead (+1))
 '-' -> memory (atPtr (subtract 1))

Reads almost like the specification!

I am building on bisserlis' answer here. interpret uses a lot of repetition that can be reduced with case..of. That allows a where clause which factors out common code that modifies memory and simply advances the instruction pointer:

interpret m@(ZL _ x _) i@(ZL _ cmd _) =
 case cmd of
 '>' -> memory znxt
 '<' -> memory zprv
 '+' -> memory zinc
 '-' -> memory zdec
 '[' -> interpret m (if x /= 0 then znxt i else zjmp znxt 0 i)
 ']' -> interpret m (if x == 0 then znxt i else zjmp zprv 0 i)
 '.' -> putChar (toEnum x)>> memory id
 ',' -> getChar>>= memory . zapp . const . fromEnum
 _ -> memory id -- Comment `Char`
 where
 memory :: (ZL Int -> ZL Int) -> IO ()
 memory f = interpret (f m) (znxt i)

I would use longer names for znxt etc., then you might consider renaming zapp to underHead or atPtr and then your code might read like this (I am mixing styles here):

 '>' -> memory advancePtr
 '<' -> memory moveHeadLeft
 '+' -> memory underHead (+1)
 '-' -> memory atPtr (subtract 1)

Reads almost like the specification!

I am building on bisserlis' answer here. interpret uses a lot of repetition that can be reduced with case..of. That allows a where clause which factors out common code that modifies memory and simply advances the instruction pointer:

interpret m@(ZL _ x _) i@(ZL _ cmd _) =
 case cmd of
 '>' -> memory znxt
 '<' -> memory zprv
 '+' -> memory zinc
 '-' -> memory zdec
 '[' -> interpret m (if x /= 0 then znxt i else zjmp znxt 0 i)
 ']' -> interpret m (if x == 0 then znxt i else zjmp zprv 0 i)
 '.' -> putChar (toEnum x)>> memory id
 ',' -> getChar>>= memory . zapp . const . fromEnum
 _ -> memory id -- Comment `Char`
 where
 memory :: (ZL Int -> ZL Int) -> IO ()
 memory f = interpret (f m) (znxt i)

I would use longer names for znxt etc., then you might consider renaming zapp to underHead or atPtr and then your code might read like this (I am mixing styles here):

 '>' -> memory advancePtr
 '<' -> memory moveHeadLeft
 '+' -> memory (underHead (+1))
 '-' -> memory (atPtr (subtract 1))

Reads almost like the specification!