feat(nixos): manage scattered trash dirs with trash-cli retention timer (#265)

* feat(nushell): add trash command using the home trash can

On a tmpfs root with persistent dirs bind-mounted in, `rm --trash`
scatters items into per-mount .Trash-$uid dirs that file managers never
show, and gio refuses to trash on those internal mounts.

Add a `trash` command implementing the freedesktop trash spec
'failsafe' mode: items are moved to ~/.local/share/Trash with
spec-compliant .trashinfo entries, so Thunar can list and restore them.

Also preserve ~/.local/share/Trash on idols-ai so the trash survives
reboots.

* feat(nixos): manage scattered trash dirs with trash-cli retention timer

Replace the hand-written nushell trash command with the established
trash-cli tooling. On a tmpfs root with persistent dirs bind-mounted
in, the trash crate (nushell rm --trash) scatters items into per-mount
.Trash-$uid dirs that file managers never show and nothing cleans up.

- add trash-cli to system packages: trash-list scans/trash-restore
  handles every mount point's trash dir
- daily systemd timer runs 'trash-empty 30 -f', purging items older
  than 30 days across the home trash and all mount points
- drop the hand-written trash.nu module
- note in preservation.nix: do NOT persist ~/.local/share/Trash; a
  bind-mounted trash dir breaks the trash crate's home-topdir match
  for files straight under $HOME (it would try /.Trash-$uid, EACCES)

* fix(nixos): order trash-empty after preservation.target

The scattered .Trash-$uid dirs live inside the preservation bind
mounts, which use DefaultDependencies=no and are therefore NOT ordered
after local-fs.target. Add preservation.target to After= so trash-empty
only runs once all those mounts are up.

hosts/idols-ai/preservation.nix

@@ -110,6 +110,14 @@ in
         # Keep .cache off tmpfs to avoid high RAM usage; many apps use it and it is storage-heavy.
         ".cache"
 
+        # NOTE: do NOT persist ~/.local/share/Trash here. The trash crate (nushell
+        # `rm --trash`) picks the home trash only when the file's mount topdir equals
+        # the trash dir's topdir; a bind-mounted Trash dir would make files straight
+        # under $HOME fail with EACCES (it would try /.Trash-$uid). The home trash on
+        # tmpfs is lost on reboot, which is acceptable: scattered per-mount
+        # .Trash-$uid dirs on persistent volumes are covered by the trash-empty
+        # retention timer (modules/nixos/base/trash.nix).
+
         # ======================================
         # Codes / Work / Playground
         # ======================================

modules/base/packages.nix

@@ -86,5 +86,6 @@
     which
     tree
     tealdeer # a very fast version of tldr
+    trash-cli # freedesktop trash can CLI: trash-list / trash-restore / trash-put / trash-empty
   ];
 }

modules/nixos/base/trash.nix

@@ -0,0 +1,61 @@
+{
+  # ==================================================================
+  #
+  # Trash retention cleanup (nixpkgs `trash-cli`)
+  #   https://github.com/andreafrancia/trash-cli
+  #
+  # On a tmpfs root with persistent dirs bind-mounted in, the trash
+  # crate (nushell `rm --trash`, GUI file managers) scatters items into
+  # per-mount .Trash-$uid dirs that file managers never show. This timer
+  # purges trash items older than RETENTION_DAYS across the home trash
+  # and all mount points.
+  #
+  # Usage:
+  #   trash-list             # view items in all trash dirs
+  #   trash-restore <file>   # restore an item
+  #   trash-rm <file>        # remove a single item from the trash
+  #   trash-empty 30 -f      # run the retention purge manually
+  #
+  # ==================================================================
+
+  myvars,
+  pkgs,
+  ...
+}:
+let
+  retentionDays = 30;
+in
+{
+  systemd.services.trash-empty = {
+    description = "Purge trash items older than ${toString retentionDays} days";
+    script = ''
+      ${pkgs.trash-cli}/bin/trash-empty ${toString retentionDays} -f
+    '';
+    # The scattered .Trash-$uid dirs live inside the preservation bind mounts, so we
+    # must run after preservation.target (reached only once all those mounts are up).
+    # preservation.mount units use DefaultDependencies=no, so they are NOT ordered
+    # after local-fs.target on their own; list both to be safe on every host.
+    # On hosts without preservation, the After=preservation.target edge is a no-op.
+    after = [
+      "local-fs.target"
+      "preservation.target"
+    ];
+    serviceConfig = {
+      Type = "oneshot";
+      User = myvars.username;
+      Environment = [
+        "HOME=/home/${myvars.username}"
+        "XDG_DATA_HOME=/home/${myvars.username}/.local/share"
+      ];
+    };
+  };
+
+  systemd.timers.trash-empty = {
+    description = "Timer for the trash-empty service";
+    wantedBy = [ "timers.target" ];
+    timerConfig = {
+      OnCalendar = "daily";
+      RandomizedDelaySec = "15min";
+    };
+  };
+}