chore: Bump haskell grammar

[github-actions]

No description provided.

[liesnikov]

As of now, I'm not seeing any errors when I compile it locally, it looks like it passes on the ci too: https://github.com/liesnikov/tree-sitter-langs/actions/runs/15485334391/job/43599147943#step:13:148

[liesnikov]

I am not sure what happened with CI, but I'm not seeing any action runs on bump-haskell branch in this repo

[liesnikov]

ah, it needs a toggle. Thanks!
Let's see what it says now

[liesnikov]

Hmm, it seems like the tests aren't stable.
The fail is in queries/haskell for all four failing runs. The run in my repo on the same rev for ubuntu 24.04 28.2 succeeded (I cancelled the others), while here the exact same one failed.

 Test queries/haskell backtrace:
  tsc--make-query(#<user-ptr ptr=0x7f6d699ae760 finalizer=0x7f6d6c3241
  tsc-make-query(#<user-ptr ptr=0x7f6d699ae760 finalizer=0x7f6d6c32415
  (let ((language (tree-sitter-require lang-symbol)) (patterns (tree-s
  tree-sitter-langs-tests-check-queries(haskell)
  (let ((lexical-binding nil)) (let* ((fn-137 #'not) (args-138 (condit
  (lambda nil (let ((lexical-binding nil)) (let* ((fn-137 #'not) (args
  ert--run-test-internal(#s(ert--test-execution-info :test #s(ert-test
  ert-run-test(#s(ert-test :name queries/haskell :documentation nil :b
  ert-run-or-rerun-test(#s(ert--stats :selector t :tests [... ... ... 
  ert-run-tests(t #f(compiled-function (event-type &rest event-args) #
  ert-run-tests-batch(nil)
  ert-run-tests-batch-and-exit(nil)
  eval((ert-run-tests-batch-and-exit nil) t)
  command-line-1(("-l" "/home/runner/eask-snapshot/lisp/core/emacs.el"
  command-line()
  normal-top-level()
Test queries/haskell condition:
    (tsc-query-invalid-node-type "exp_negation"
				 (5 . 1)
				 126)

[jcs090218]

It seems like it can be compiled on aarch64 but not on x64. 🤔

[liesnikov]

edit: wait, nevermind, I was confused about the runs - you are right and the x64 runs also failed there

If you look at the CI run in this repo this seems to be the differentiating factor. But it wouldn't explain why the one in my repo succeeded - that run is on x64 and it did succeed.
It was triggered manually, but I don't see anything in the action file where that would make a difference.

[liesnikov]

Not sure how to make sense of the local succeeding tests though.
I am on Debian 12 with x86_64 CPU, the project shell with dependencies was constructed via nix.

[jcs090218]

My best guess is that the environment differs slightly—either due to Nix or GitHub Actions. 🤔