Add {-# LANGUAGE Data #-} to the top of the file? (y or n)

[andreasabel]

When I load a hs file without project (cabal or stack) I get asked

Add {-# LANGUAGE Data #-} to the top of the file? (y or n)

This question can be repeated several times in a row, regardless whether I answer y or n.
Each time I answer y the {-# LANGUAGE Data #-} line will be added (which does not make any sense since LANGUAGE pragmas are idempotent).

The Data extension isn't even listed here: https://ghc.gitlab.haskell.org/ghc/doc/users_guide/exts/table.html
I don't know what purpose it should serve.

I just upgraded to the latest version of haskell-mode (v20231115) but the problem persists.

[andreasabel]

Seems to be originating from here:

haskell-mode/haskell-load.el

Lines 210 to 215 in 79eaf44

	(when haskell-process-suggest-language-pragmas
	(haskell-process-suggest-pragma
	session
	"LANGUAGE"
	(match-string 1 msg)
	file)))

It does not seem to limit suggestion to valid pragmas, the name of the pragma is here obtained by (match-string 1 msg), maybe coming from this match:

haskell-mode/haskell-load.el

Lines 199 to 209 in 79eaf44

	(or
	(and (string-match " -X\\([A-Z][A-Za-z]+\\)" msg)
	(not (string-match "\\([A-Z][A-Za-z]+\\) is deprecated" msg)))
	(string-match "Use \\([A-Z][A-Za-z]+\\) to permit this" msg)
	(string-match "Use \\([A-Z][A-Za-z]+\\) to allow" msg)
	(string-match "Use \\([A-Z][A-Za-z]+\\) to enable" msg)
	(string-match
	"Use \\([A-Z][A-Za-z]+\\) if you want to disable this"
	msg)
	(string-match "use \\([A-Z][A-Za-z]+\\)" msg)
	(string-match "You need \\([A-Z][A-Za-z]+\\)" msg)))

Ah, I guess this is thanks to our wonderful new x-partial warning:

• https://gitlab.haskell.org/ghc/ghc/-/issues/23934
• Add {-# WARNING #-} to Data.List.{head,tail} core-libraries-committee#87

In the *haskell* buffer I see:

Lect1.hs:111:13-16: warning: [GHC-63394] [-Wx-partial] …
In the use of ‘head’
(imported from Prelude, but defined in GHC.List):
"This is a partial function, it throws an error on empty lists. Use pattern matching or Data.List.uncons instead. Consider refactoring to use Data.List.NonEmpty."

Can you spot which regex matches here and extracts a Data hypothetical language extension?

[andreasabel]

In PR #1834 this issue is fixed by removing the bad regex.