Roo && concurrent-ruby

roo_with_concurrency.rb

require 'concurrent'
require 'concurrent-edge'
# require 'benchmark'
# require 'pry'

# rails...
#require './config/environment'

class Roo::Base
  def each_sheet_concurrently(&block)
    jobs = self.sheets.map do |name|
      Concurrent.future do
        block.call(self.sheet(name), name)
      end
    end
    Concurrent.zip(*jobs).value
  end

  def each_row_concurrently(name, &block)
    jobs = 1.upto(last_row(name)).map do |line|
      Concurrent.future do
        block.call(row(line, name))
      end
    end
    Concurrent.zip(*jobs).value
  end

  # alternate method...
  def each_concurrently(&block)
    jobs = Concurrent::Array.new

    self.sheets.each do |name|
      1.upto(sheet(name).last_row) do |line|
        jobs << Concurrent.future do
          block.call(sheet(name).row(line, name), name)
        end
      end
    end
    Concurrent.zip(*jobs).value
  end
end


# file_path = some_excelx_spreadsheet...
spreadsheet = Roo::Spreadsheet.open(file_path)

rows_per_sheet = Array.new
all_rows = Array.new

spreadsheet.each_sheet_concurrently do |sheet, name|
  # some operation on the sheet
  rows_per_sheet << sheet.last_row

  spreadsheet.each_row_concurrently(name) do |row|
    # some operation on rows
    all_rows << row
  end
end

puts all_rows.count

# spread = Concurrent::Hash.new
# spreadsheet.each_concurrently do |row, sheet|
#   spread[sheet] ||= Concurrent::Array.new
#   spread[sheet] << row
# end

# binding.pry
# so we don't advance and not hit the debugger...
# foo = "bar"