Georgi Lyubenov

Ah, sorry, I did not mean to say "please stop blocking this", but rather wanted to simply reinforce my interest in having such a function available.

Sounds like we need `Parsers à la Carte` :smile:

You should add the new flags to `print_usage()`.

Also encountering the same issue

It reproduces with `stack` on my end, the OP seems to mention a cabal project though.

`stack` *does* respect the `language` field, at least with hpack, because things fail to compile with stack, but don't show errors when I run `haskell-language-server-wrapper` (e.g. `ImportQualifiedPost`)

`haskell-language-server`s `--probe-tools` gives some minimal info in this spirit: ``` blademaster :: ~ » haskell-language-server-wrapper --probe-tools haskell-language-server version: 1.5.1.0 (GHC: 8.10.7) (PATH: /home/googleson78/.ghcup/bin/haskell-language-server-wrapper-1.5.1) (GIT hash: 745ef26f406dbdd5e4a538585f8519af9f1ccb09) Tool versions found on...

I'm all for being more sensitive and non-hurtful, but I believe changing "should" to "you may" or "you could" is changing the semantics of the sentence from a "this might...

You have source span info available in hlint, along with parens info, so it should be possible, I think? As long as ghc sets a "proper" source span on parens...

While I do agree with @zliu41's sentiment, I also know that there are people who are firmly in the "if it's more than one line put it in a `let`/`where`"...