
ghc and ghci 8.6.5 accept 1=2 at top level. It seems to have no effect. What does it mean?

It's the same in do-blocks: main :: IO () main = do let 3 = 2 + 2 putStrLn "Oh fiddlesticks" On Mon, Dec 14, 2020 at 12:14:53PM -0500, M Douglas McIlroy wrote:
ghc and ghci 8.6.5 accept 1=2 at top level. It seems to have no effect. What does it mean? _______________________________________________ Beginners mailing list Beginners@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/beginners

Il 14 dicembre 2020 alle 20:05 amindfv@mailbox.org ha scritto:
It's the same in do-blocks:
main :: IO () main = do let 3 = 2 + 2 putStrLn "Oh fiddlesticks"
What happens exactly when I type this? λ> "prova" = "foo" λ> 'c' = 'd' λ> 'c' 'c'
From the Report I read:
lexp → let decls in exp decls → { decl1 ; … ; decln } (n ≥ 0) ⁝ decl → (funlhs | pat) rhs ⁝ Am I correct in saying `pat rhs` is the rule being used here? I do not understand how/when it comes useful in a let —F

On 2020-12-14 18:14, M Douglas McIlroy wrote:
ghc and ghci 8.6.5 accept 1=2 at top level. It seems to have no effect.
On 2020-12-15 02:29, Francesco Ariis wrote:
Il 14 dicembre 2020 alle 20:05 amindfv@mailbox.org ha scritto:
It's the same in do-blocks:
main :: IO () main = do let 3 = 2 + 2 putStrLn "Oh fiddlesticks"
What happens exactly when I type this?
λ> "prova" = "foo" λ> 'c' = 'd' λ> 'c' 'c'
These are examples of pattern bindings, but since they don't bind any variables, they are not very useful. As examples of more useful pattern bindings, consider > (xs,ys) = splitAt 3 [1..8] > [1,x,y] = [1..3] These are both examples where the value of the expression on the rhs matches the pattern on the lhs, so you can obtain the values of the variables bound by the pattern: > xs [1,2,3] > ys [4,5,6,7,8] > x 2 > y 3 If the value of the expression on the rhs doesn't match the pattern on the lhs, you get an error, but because of lazy evaluation the value of the rhs is not computed and matched against the pattern until you use one of the variables in the pattern: > [1,z,2] = [1..3] > z *** Exception: <interactive>:7:1-16: Irrefutable pattern failed for pattern [1, z, 2] This means that when the pattern does not contain any variables, the value of the rhs is never computed and matched against the pattern, so even if the value does not match the pattern, there is no error message: > [] = [1..3] > [_,_,_] = [] > True = False > 1 = 2 If you turn on -Wunused-pattern-binds, you get a warning for pattern bindings like these: > :set -Wunused-pattern-binds > True=False <interactive>:2:1: warning: [-Wunused-pattern-binds] This pattern-binding binds no variables: True = False Best regards, Thomas H
participants (4)
-
amindfv@mailbox.org
-
Francesco Ariis
-
M Douglas McIlroy
-
Thomas Hallgren