Ensure the engine is idle even when multiple requests are scheduled at once

[milk1000cc]

If the crawler has multiple start_urls or yields multiple requests at once, the engine does not seem to terminate even if the process finishes successfully.

class ExampleCrawler < Vessel::Cargo
  start_urls(
    "https://example.com/1" => :parse,
    "https://example.com/2" => :parse
  )

  def parse
    yield [
      request(url: "https://example.com/3", handler: :parse),
      request(url: "https://example.com/4", handler: :parse)
    ]
  end
end

The number of req_enqueued to be increased when scheduling seems to be wrong.

[route]

Right you are) thanks!