Pipe cannot interact with other macros

[bertulli]

Hi! Thanks for you work, loved the package so far. I was working with some DataFrames and I got this problem with StatsPlots' @df macro working with @pipe:

julia> @pipe @df df_pair_temp zip(cols(instruction_pair_first_sym), cols(:first_encoding))

zip(["adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0", "adc_r0_r0_r0"  …  "udiv_r3_r2", "udiv_r3_r2", "umlal_r0_r1_r2_r3", "umlal_r0_r1_r2_r3", "umlal_r0_r1_r2_r3", "umlal_r0_r1_r2_r3", "umull_r0_r1_r2_r3", "umull_r0_r1_r2_r3", "umull_r0_r1_r2_r3", "umull_r0_r1_r2_r3"], String15["eb400000", "eb400000", "eb400000", "eb400000", "eb400000", "eb400000", "eb400000", "eb400000", "eb400000", "eb400000"  …  "fbb3f3f2", "fbb3f3f2", "fbe20103", "fbe20103", "fbe20103", "fbe20103", "fba20103", "fba20103", "fba20103", "fba20103"])

works, as you can see. df_pair_temp is my DataFrame, and instruction_pair_first_sym is a Symbol. Adding the first function

julia> @pipe @df df_pair_temp zip(cols(instruction_pair_first_sym), cols(:first_encoding)) |>
               collect
706-element Vector{Tuple{String, String15}}:
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r0_r0", "eb400000")
 ("adc_r0_r1_r2", "eb410002")
 ("adc_r0_r1_r2", "eb410002")
 ("adc_r0_r1_r2", "eb410002")
 ("adc_r0_r1_r2", "eb410002")
 ("adc_r0_r1_r2", "eb410002")
 ⋮
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("udiv_r3_r2", "fbb3f3f2")
 ("umlal_r0_r1_r2_r3", "fbe20103")
 ("umlal_r0_r1_r2_r3", "fbe20103")
 ("umlal_r0_r1_r2_r3", "fbe20103")
 ("umlal_r0_r1_r2_r3", "fbe20103")
 ("umull_r0_r1_r2_r3", "fba20103")
 ("umull_r0_r1_r2_r3", "fba20103")
 ("umull_r0_r1_r2_r3", "fba20103")
 ("umull_r0_r1_r2_r3", "fba20103")

BUT

julia> @pipe @df df_pair_temp zip(cols(instruction_pair_first_sym), cols(:first_encoding)) |>
               collect(_)
ERROR: syntax: all-underscore identifier used as rvalue
Stacktrace:
 [1] top-level scope
   @ REPL[31]:1

even though they should be the same. Expanding with @macroexpand also gives

julia> @macroexpand @pipe @df df_pair_temp zip(cols(instruction_pair_first_sym), cols(:first_encoding)) |>
               collect(_)
:(((var"##374"->begin
          ((var"##range#375", var"##range#376"), var"##377") = (StatsPlots).extract_columns_and_names(var"##374", instruction_pair_first_sym, :first_encoding)
          (StatsPlots).add_label(["zip(cols(instruction_pair_first_sym), cols(first_encoding))", "collect(_)"], |>, zip(var"##range#375", var"##range#376"), collect(_))
      end))(df_pair_temp))

Maybe this is related to #20 and #21 . Can I do anything to fix this?

Thanks!