lib.types.anything: Add functionArgs

[roberth]

Description of changes

Make function reflection work even when types.anything mediates the value.

I'm not a big fan of function reflection, or types.anything, because both are broken concepts, but this at least fixes the use case where a function is expected and passed to, say, callPackage or evalModules, etc.

I have chosen not to apply this change to functionTo, because this solution has a performance cost, and perhaps usages of functionTo are better - by not using reflection.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 23.11 Release Notes (or backporting 23.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

[infinisil]

Where did this turn up to be necessary?

[roberth]

I think it was a solution to a problem with two solutions - one being this PR and the other one being:

• linux-builder: change from modules to config nix-darwin/nix-darwin#785

The original problem was in a private repo, but if I recall correctly, anything was interfering with the application of module arguments, which, as you know, behaves a bit like

• lib.lazyFunction: init #194514

as it relies on reflection.

[infinisil]

Minor suggestions, looks fine overall!

[rhendric]

Recently found myself in want of this, but it looks like this PR was abandoned. Any interest in reviving it, or should I open a new one?

[roberth]

Rebased, applied suggestions, tests pass locally. Can be merged if ofborg agrees.

EDIT: ignore nixfmt, because types.nix wasn't formatted

[qweered]

Needs rebase