add astable macro

[pdeffebach]

Not yet ready for a review.

To simulate @t from DataFrameMacros

julia> df = DataFrame(a = rand(5), b = rand(5));

julia> @transform df @astable begin 
           :x = :b
           $"hello" = :a
       end
5×4 DataFrame
 Row │ a         b         x         hello    
     │ Float64   Float64   Float64   Float64  
─────┼────────────────────────────────────────
   1 │ 0.495296  0.788492  0.788492  0.495296
   2 │ 0.935795  0.895012  0.895012  0.935795
   3 │ 0.340286  0.32458   0.32458   0.340286
   4 │ 0.911353  0.585055  0.585055  0.911353
   5 │ 0.74855   0.655892  0.655892  0.74855

[pdeffebach]

A note on the API: As you may know I also maintain AddToField.jl with syntax

@addnt begin 
    @add x = 1
    @add y = 5
end

To make the named tuple (x = 1, y = 5).

This functionality is similar, but a bit more DataFrames specific. It allows for consistency by keeping :x and $x referring to columns in the DataFrame. We can't do that in an external macro because we need to treat the column identifiers on the LHS in a special way.

[pdeffebach]

Okay I'm thinking I will need to port over more of DataFrameMacros.jl's parsing code, in particular the stuff that uses MacroTools.postwalk. I will think more about this tomorrow.

[bkamins]

Thank you for working on this!

[pdeffebach]

Closed in favor of #298