bytestring

The documentation is itself fairly funny, for those who don’t care to click ahead: > This "function" has a superficial similarity to ‘unsafePerformIO’ but it is in fact a malevolent agent of chaos. It unpicks the seams of reality (and the IO monad) so that the normal rules no longer apply. It lulls you into thinking it is reasonable, but when you are not looking it stabs you in the back and aliases all of your mutable buffers. The carcass of many a seasoned Haskell programmer lie strewn at its feet. > Witness the trail of destruction: https://github.com/haskell/bytestring/commit/71c4b438c675aa360c79d79acc9a491e7bbc26e7 https://github.com/haskell/bytestring/commit/210c656390ae617d9ee3b8bcff5c88dd17cef8da https://ghc.haskell.org/trac/ghc/ticket/3486 https://ghc.haskell.org/trac/ghc/ticket/3487 https://ghc.haskell.org/tra

In fairness, that comment refers to using the function accursedUnutterablePerformIO The comment for accursedUnutterablePerformIO: This "function" has a superficial similarity to 'unsafePerformIO' but it is in fact a malevolent agent of chaos. It unpicks the seams of reality (and the 'IO' monad) so that the normal rules no longer apply. It lulls you into thinking it is reasonable, but when you are not looking it stabs you in the back and aliases all of your mutable buffers. The carcass of many a seasoned Haskell programmer lie strewn at its feet. Witness the trail of destruction: https://github.com/haskell/bytestring/commit/71c4b438c675aa3... https://github.com/haskell/bytestring/commit/210c656390ae617... https://ghc.haskell.org/trac/ghc/ticket/3486 https://ghc.haskell.org/trac/ghc/ticket/3487 https://ghc.haskell.org/trac/ghc/ticket/7270 D

Technically the escape hatches are only the stuff here: https://hackage.haskell.org/package/ghc-prim-0.5.0.0/docs/GH... So you could write case readIntArray# v (k +# 1#) realWorld# of (# vkp, _ #) -> case readIntArray# v (k -# 1#) realWorld# of (# vkm, _ #) -> case vkp <# vkm of 0# -> ... 1# -> ... with syntax sugar reasonably close to V ! (k+1) < V ! (k-1). The problem is the state tokens; I ignored them by using realWorld# and _ above, but that makes the code not referentially transparent. So in practice, Haskell always passes them around in a monad (Except in all the core libraries, which use things like accursedUnutterablePerformIO to go back down to the thunk level: https://github.com/haskell/bytestring/blob/master/Data/ByteS...). But reads are (mostly) pure so it shouldn't break anything to provide a pure interface. A

The unsafe low-level stuff gives rise to some great function names like `reallyUnsafePtrEquality` and `accursedUnutterablePerformIO` https://github.com/haskell/bytestring/blob/dd3c07d115840d134...