How to avoid repetition in an Ecto select?

Questions / Help
1

I have a query that sums up totals from within a date range using Postgres’ filter (e.g. select sum(p.total) filter (where "date" >= first and "date" <= last)), and I’d like to avoid some of the repetition if I can, I have to fetch a lot of columns so the code gets difficult to read fairly quick.

The best I have so far is

      select: %{
        total: sum(p.total) |> in_range(p.date, ^first, ^last),
        other: sum(p.other) |> in_range(p.date, ^first, ^last),
      }

And the in_range macro looks like

  defmacro in_range(sum_expression, date_col, start_date, end_date) do
    quote do
      unquote(sum_expression)
      |> filter(
        unquote(date_col) >= unquote(start_date) and unquote(date_col) <= unquote(end_date)
      )
      |> coalesce(0)
    end
  end

Is there some way I can restructure the code so that I can avoid the repetition? I was thinking something like the block below, but I know it is incorrect

      sum_in_range = in_range(p.date, ^first, ^last)

      select: %{
        total: sum(p.total) |> sum_in_range,
        other: sum(p.other) |> sum_in_range
      }