Jeremy Fleischman

I haven't looked at the diff yet, but a couple of things perhaps worth checking: - https://github.com/NixOS/nixfmt/pull/291/ - The "merge driver" mentioned here: https://github.com/NixOS/nixpkgs/pull/395864

I haven't checked the code, but I thought treefmt implemented it's `--stdin` feature by creating a file under the hood. Are you *sure* fourmolu is actually being passed the file...

@MagicRB, I put together this experiment that shows treefmt creates a tempfile when invoked via `--stdin`: https://github.com/jfly/2025-04-26-treefmt-stdin-demo. Note that the treefmt spec doesn't expect formatters to understand how to process...

IIUC, this doesn't feel like a Haskell/Emacs problem. I'm not sure I totally grok the specific issue here (is this a monorepo with multiple "subprojects" with different formatting rules?), but...

> Is that a thing we believe all formatters should support? If their behavior could depend on the real filename, then they already support something like this, or they have...

@zimbatm, @brianmcgee, nixfmt would benefit from having a stdin spec to follow. We're doing some light bikeshedding over on https://github.com/NixOS/nixfmt/issues/301 ;)

We discussed this today. Here's the plan: 1. We'll add a "Stdin Spec" to the existing [Formatter Spec](https://treefmt.com/main/reference/formatter-spec/). In short, it'll be treefmt's cli: - If the cli accepts stdin,...

> In https://github.com/numtide/treefmt/issues/557, the user described adding a file to the git index (via git add) but not yet committed ("staged" in Git terms), and opined that the file they...

We absolutely care about making treefmt play nicely with precommit hooks. I don't think I understand the issue, though. Precommit hooks are notoriously difficult to write correctly (distinguishing between staged...

> Turns out, after some experimentation, prettier, for good reason, doesn't process files that are not descendants of the directory it's being invoked from. This reminds me of [the issue...