prop/test/microwaveprop/weather/scalar_file_test.exs

defmodule Microwaveprop.Weather.ScalarFileTest do
  # async: false — mutates the global :propagation_scores_dir env to redirect
  # the on-disk artifact tree to a private tmp directory.
  use ExUnit.Case, async: false

  alias Microwaveprop.Weather.ScalarFile

  setup do
    dir =
      Path.join(
        System.tmp_dir!(),
        "weather_scalar_file_test_#{System.unique_integer([:positive])}"
      )

    File.mkdir_p!(dir)

    prev = Application.get_env(:microwaveprop, :propagation_scores_dir)
    Application.put_env(:microwaveprop, :propagation_scores_dir, dir)

    on_exit(fn ->
      File.rm_rf!(dir)

      if prev do
        Application.put_env(:microwaveprop, :propagation_scores_dir, prev)
      else
        Application.delete_env(:microwaveprop, :propagation_scores_dir)
      end
    end)

    %{dir: dir}
  end

  defp sample_row(lat, lon, temp \\ 22.0) do
    %{
      lat: lat,
      lon: lon,
      valid_time: ~U[2026-04-28 12:00:00Z],
      temperature: temp,
      dewpoint_depression: 5.0,
      surface_rh: 70.0,
      surface_pressure_mb: 1013.0,
      surface_refractivity: 330.0,
      refractivity_gradient: -45.0,
      bl_height: 800.0,
      pwat: 25.0,
      temp_850mb: 12.0,
      dewpoint_850mb: 4.0,
      temp_700mb: 0.0,
      dewpoint_700mb: -10.0,
      lapse_rate: 6.5,
      mid_lapse_rate: 6.0,
      inversion_strength: 0.5,
      inversion_base_m: nil,
      ducting: false,
      duct_base_m: nil,
      duct_strength: nil
    }
  end

  describe "base_dir/0" do
    test "is rooted under the configured scores dir", %{dir: dir} do
      assert ScalarFile.base_dir() == Path.join(dir, "weather_scalars")
    end
  end

  describe "write!/2 + read_bounds/2" do
    test "round-trips the rows that fall inside the requested bounds" do
      valid_time = ~U[2026-04-28 12:00:00Z]

      rows = [
        sample_row(33.0, -97.0, 22.0),
        sample_row(33.5, -97.5, 23.0),
        # Outside Texas bounds below
        sample_row(45.0, -90.0, 5.0)
      ]

      ScalarFile.write!(valid_time, rows)

      result =
        ScalarFile.read_bounds(valid_time, %{
          "south" => 32.0,
          "north" => 34.5,
          "west" => -98.0,
          "east" => -96.0
        })

      lat_lons = result |> Enum.map(&{&1.lat, &1.lon}) |> Enum.sort()
      assert lat_lons == [{33.0, -97.0}, {33.5, -97.5}]
    end

    test "returns [] when the file does not exist" do
      assert ScalarFile.read_bounds(~U[2026-04-28 12:00:00Z], nil) == []
    end

    test "read_bounds with nil bounds returns every persisted row" do
      valid_time = ~U[2026-04-28 12:00:00Z]
      rows = [sample_row(33.0, -97.0), sample_row(45.0, -90.0)]

      ScalarFile.write!(valid_time, rows)

      assert Enum.count_until(ScalarFile.read_bounds(valid_time, nil), 3) == 2
    end

    test "preserves the full set of derived layer fields on round-trip" do
      valid_time = ~U[2026-04-28 12:00:00Z]
      [row] = [sample_row(33.0, -97.0)]

      ScalarFile.write!(valid_time, [row])

      [round_tripped] = ScalarFile.read_bounds(valid_time, nil)

      for key <- Map.keys(row) do
        assert Map.fetch!(round_tripped, key) == Map.fetch!(row, key),
               "key #{inspect(key)} mismatched after round-trip"
      end
    end
  end

  describe "read_point/3" do
    test "returns the snapped row for the requested lat/lon" do
      valid_time = ~U[2026-04-28 12:00:00Z]
      row = sample_row(33.0, -97.0)

      ScalarFile.write!(valid_time, [row])

      assert {:ok, fetched} = ScalarFile.read_point(valid_time, 33.0, -97.0)
      assert fetched.temperature == 22.0
    end

    test "returns :miss when no scalar file exists" do
      assert ScalarFile.read_point(~U[2026-04-28 12:00:00Z], 33.0, -97.0) == :miss
    end

    test "returns :miss when the cell is not in the file" do
      valid_time = ~U[2026-04-28 12:00:00Z]
      ScalarFile.write!(valid_time, [sample_row(33.0, -97.0)])

      assert ScalarFile.read_point(valid_time, 40.0, -80.0) == :miss
    end
  end

  describe "exists?/1 and list_valid_times/0" do
    test "reflects what has been written" do
      vt1 = ~U[2026-04-28 12:00:00Z]
      vt2 = ~U[2026-04-28 13:00:00Z]

      refute ScalarFile.exists?(vt1)
      assert ScalarFile.list_valid_times() == []

      ScalarFile.write!(vt1, [sample_row(33.0, -97.0)])
      ScalarFile.write!(vt2, [sample_row(33.0, -97.0)])

      assert ScalarFile.exists?(vt1)
      assert ScalarFile.list_valid_times() == [vt1, vt2]
    end
  end

  describe "HRDPS sibling read merge" do
    test "read_bounds returns HRRR + HRDPS rows when both dirs exist" do
      vt = ~U[2026-04-29 12:00:00Z]

      ScalarFile.write!(vt, [sample_row(33.0, -97.0, 22.0)])
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)

      rows = ScalarFile.read_bounds(vt, nil)
      sorted = Enum.sort_by(rows, & &1.lat)

      assert Enum.count_until(sorted, 3) == 2
      assert Enum.at(sorted, 0).lat == 33.0
      assert Enum.at(sorted, 1).lat == 53.0
    end

    test "read_bounds returns HRDPS rows when only HRDPS dir exists" do
      vt = ~U[2026-04-29 12:00:00Z]
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)

      assert [%{lat: 53.0, lon: -60.0}] = ScalarFile.read_bounds(vt, nil)
    end

    test "read_point falls through to HRDPS when the HRRR chunk is absent" do
      vt = ~U[2026-04-29 12:00:00Z]
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)

      assert {:ok, %{lat: 53.0, lon: -60.0, temperature: 8.0}} =
               ScalarFile.read_point(vt, 53.0, -60.0)
    end

    test "exists?/1 is true when only the HRDPS dir has chunks" do
      vt = ~U[2026-04-29 12:00:00Z]
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)
      assert ScalarFile.exists?(vt)
    end

    test "merge prefers HRRR row when both dirs cover the same lat/lon" do
      vt = ~U[2026-04-29 12:00:00Z]

      ScalarFile.write!(vt, [sample_row(33.0, -97.0, 22.0)])
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 33.0, -97.0, 99.0)

      [row] = ScalarFile.read_bounds(vt, nil)
      assert row.temperature == 22.0
    end
  end

  describe "HRDPS-only reads (for the /weather-ca endpoint)" do
    test "read_bounds_hrdps returns only HRDPS chunks, ignoring co-located HRRR" do
      vt = ~U[2026-04-29 12:00:00Z]

      # Both dirs cover the same point. The HRRR-only read_bounds/2 would
      # merge them (preferring HRRR); read_bounds_hrdps/2 must return ONLY
      # the HRDPS row so /weather-ca shows Canadian-source data.
      ScalarFile.write!(vt, [sample_row(33.0, -97.0, 22.0)])
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)

      assert [%{lat: 53.0, lon: -60.0, temperature: 8.0}] =
               ScalarFile.read_bounds_hrdps(vt, nil)
    end

    test "read_bounds_hrdps respects bounds" do
      vt = ~U[2026-04-29 12:00:00Z]

      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 33.0, -97.0, 22.0)

      bounds = %{"south" => 50.0, "north" => 60.0, "west" => -70.0, "east" => -50.0}
      assert [%{lat: 53.0, lon: -60.0}] = ScalarFile.read_bounds_hrdps(vt, bounds)
    end

    test "read_bounds_hrdps returns [] when no HRDPS dir exists" do
      vt = ~U[2026-04-29 12:00:00Z]
      ScalarFile.write!(vt, [sample_row(33.0, -97.0, 22.0)])

      assert ScalarFile.read_bounds_hrdps(vt, nil) == []
    end

    test "list_valid_times_hrdps returns times from HRDPS-suffixed dirs only" do
      vt1 = ~U[2026-04-29 12:00:00Z]
      vt2 = ~U[2026-04-29 13:00:00Z]
      vt_hrrr_only = ~U[2026-04-29 14:00:00Z]

      hand_write_chunk(ScalarFile.dir_for_hrdps(vt1), 53.0, -60.0, 8.0)
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt2), 53.0, -60.0, 9.0)
      ScalarFile.write!(vt_hrrr_only, [sample_row(33.0, -97.0, 22.0)])

      assert ScalarFile.list_valid_times_hrdps() == [vt1, vt2]
    end
  end

  # The /weather (dual-source) timeline picks valid_times from HRRR's
  # hourly cadence. HRDPS publishes 4×/day with multi-hour latency, so
  # the requested time will frequently miss any HRDPS dir. Snapping to
  # the closest HRDPS dir within a window lets the Canadian overlay
  # render slightly stale rather than disappear.
  describe "read_bounds_hrdps_nearest/3" do
    test "returns rows from the exact time when its HRDPS dir exists" do
      vt = ~U[2026-04-29 12:00:00Z]
      hand_write_chunk(ScalarFile.dir_for_hrdps(vt), 53.0, -60.0, 8.0)

      assert [%{lat: 53.0, lon: -60.0, temperature: 8.0}] =
               ScalarFile.read_bounds_hrdps_nearest(vt, nil, 6 * 3600)
    end

    test "snaps to the closest HRDPS dir within the window when exact time is absent" do
      requested = ~U[2026-04-29 14:00:00Z]
      closer = ~U[2026-04-29 13:00:00Z]
      farther = ~U[2026-04-29 10:00:00Z]

      hand_write_chunk(ScalarFile.dir_for_hrdps(closer), 53.0, -60.0, 7.0)
      hand_write_chunk(ScalarFile.dir_for_hrdps(farther), 53.0, -60.0, 99.0)

      assert [%{temperature: 7.0}] =
               ScalarFile.read_bounds_hrdps_nearest(requested, nil, 6 * 3600)
    end

    test "snaps forward as well as backward (HRDPS forecast can lead requested time)" do
      requested = ~U[2026-04-29 12:00:00Z]
      future = ~U[2026-04-29 13:00:00Z]

      hand_write_chunk(ScalarFile.dir_for_hrdps(future), 53.0, -60.0, 11.0)

      assert [%{temperature: 11.0}] =
               ScalarFile.read_bounds_hrdps_nearest(requested, nil, 6 * 3600)
    end

    test "returns [] when no HRDPS dir exists inside the window" do
      requested = ~U[2026-04-29 14:00:00Z]
      stale = ~U[2026-04-29 02:00:00Z]

      hand_write_chunk(ScalarFile.dir_for_hrdps(stale), 53.0, -60.0, 8.0)

      # 12h gap, 6h window → not eligible.
      assert ScalarFile.read_bounds_hrdps_nearest(requested, nil, 6 * 3600) == []
    end

    test "returns [] when no HRDPS dirs exist at all" do
      requested = ~U[2026-04-29 14:00:00Z]
      assert ScalarFile.read_bounds_hrdps_nearest(requested, nil, 6 * 3600) == []
    end
  end

  defp hand_write_chunk(dir, lat, lon, temp) do
    File.mkdir_p!(dir)
    lat_band = (lat / 5) |> Float.floor() |> trunc()
    lon_band = (lon / 5) |> Float.floor() |> trunc()
    path = Path.join(dir, "#{lat_band}_#{lon_band}.mp.gz")

    payload =
      Msgpax.pack!(
        [
          %{
            "lat" => lat,
            "lon" => lon,
            "valid_time" => "2026-04-29T12:00:00Z",
            "temperature" => temp
          }
        ],
        iodata: false
      )

    File.write!(path, :zlib.gzip(payload))
  end

  describe "Rust-writer compatibility (the cross-language wire format)" do
    test "decodes a hand-rolled gzipped MessagePack chunk and atomizes whitelisted keys" do
      valid_time = ~U[2026-04-29 12:00:00Z]
      dir = Path.join(ScalarFile.dir_for(valid_time), "")
      File.mkdir_p!(dir)

      # Mirror exactly what `prop_grid_rs::weather_scalar_file::write_atomic`
      # emits: gzipped MessagePack, top-level array of string-keyed maps.
      # If this round-trip ever breaks, Elixir is reading something the
      # Rust pipeline cannot produce — a wire-format regression.
      payload =
        Msgpax.pack!(
          [
            %{
              "lat" => 33.0,
              "lon" => -97.0,
              "valid_time" => "2026-04-29T12:00:00Z",
              "temperature" => 22.5,
              "dewpoint_depression" => 10.0,
              "surface_rh" => 50.0,
              "ducting" => false,
              # An out-of-whitelist key — the reader must NOT atomize this.
              "unknown_extra_key" => 1.23
            }
          ],
          iodata: false
        )

      path = Path.join(dir, "6_-20.mp.gz")
      File.write!(path, :zlib.gzip(payload))

      [row] = ScalarFile.read_bounds(valid_time, nil)

      # Whitelisted keys are atoms.
      assert row.lat == 33.0
      assert row.lon == -97.0
      assert row.temperature == 22.5
      assert row.ducting == false

      # `valid_time` round-trips back to a DateTime.
      assert row.valid_time == valid_time

      # Out-of-whitelist keys stay as strings so a future field can be
      # added without a deploy-order coordination problem.
      assert row["unknown_extra_key"] == 1.23
    end
  end

  describe "prune_older_than/1 and retain_window/2" do
    test "prune_older_than removes files strictly before the cutoff" do
      old = ~U[2026-04-27 12:00:00Z]
      new = ~U[2026-04-28 12:00:00Z]

      ScalarFile.write!(old, [sample_row(33.0, -97.0)])
      ScalarFile.write!(new, [sample_row(33.0, -97.0)])

      assert ScalarFile.prune_older_than(~U[2026-04-28 00:00:00Z]) == 1
      refute ScalarFile.exists?(old)
      assert ScalarFile.exists?(new)
    end

    test "retain_window keeps only files inside [run_time, run_time + max_forecast_hour * 3600]" do
      run_time = ~U[2026-04-28 12:00:00Z]
      inside = ~U[2026-04-28 14:00:00Z]
      outside = ~U[2026-04-29 06:00:00Z]

      ScalarFile.write!(run_time, [sample_row(33.0, -97.0)])
      ScalarFile.write!(inside, [sample_row(33.0, -97.0)])
      ScalarFile.write!(outside, [sample_row(33.0, -97.0)])

      assert ScalarFile.retain_window(run_time, 4) == 1
      assert ScalarFile.exists?(run_time)
      assert ScalarFile.exists?(inside)
      refute ScalarFile.exists?(outside)
    end
  end
end