Question

I’m trying to figure out how to create a Sequel Dataset that contains function calls and constants as part of its select and uses aliases. I’m stuck on the proper syntax.

I’m trying to avoid inserting raw SQL but I will if I have to as long as I can chain the resulting dataset and refer to column aliases inside the raw SQL.

My actual query is more complicated (and I’m not actually calling sin, it’s just an example of a function call), but for illustrative purposes, here’s a stripped down version of what I’m trying to do in SQL:

SELECT
  0 AS a,               -- constant
  sin(t.x) AS b,        -- function call with aliased column value argument
  t.x AS c
FROM
  T AS t
;

What would the equivalent, idomatic, Sequel be? So far I have:

DB[:T.as(:t)].
    select(
        #0 as a,        # How? constant
        #sin(t.x) as b, # How? function call with aliased column value argument
        :t__x.as(:c)
    )

Answer 1

If you want a more concise way, you can use implicit aliasing using symbols with a triple underscore, use just the select method with a virtual row block:

DB[:T___t].select('0'.lit.as(:a), :t__x___c){sin(t__x).as(b)}

This isn’t exactly the same as the order of the selected columns differs, but that usually doesn’t matter.