SeleniumHQ/selenium · error · Error::WebDriverError

file not added for reaping: #{file.inspect}

Error message

file not added for reaping: #{file.inspect}

What it means

Raised as Error::WebDriverError by FileReaper.reap(file) when reap is enabled (default true) but the given file was never registered via FileReaper << (file). FileReaper is @api private, tracks temp files per Process.pid, and is invoked internally during driver teardown — end users essentially never call it directly.

Source

Thrown at rb/lib/selenium/webdriver/common/file_reaper.rb:46

        attr_writer :reap

        def reap?
          @reap = defined?(@reap) ? @reap : true
        end

        def tmp_files
          @tmp_files ||= Hash.new { |hash, pid| hash[pid] = [] }
          @tmp_files[Process.pid]
        end

        def <<(file)
          tmp_files << file
        end

        def reap(file)
          return unless reap?

          raise Error::WebDriverError, "file not added for reaping: #{file.inspect}" unless tmp_files.include?(file)

          FileUtils.rm_rf tmp_files.delete(file)
        end

        def reap!
          if reap?
            tmp_files.each { |file| FileUtils.rm_rf(file) }
            true
          else
            false
          end
        end
      end

      # we *do* want child process reaping, so not using Platform.exit_hook here.
      at_exit { reap! }
    end # FileReaper
  end # WebDriver

View on GitHub (pinned to aa36b38e69)

Solutions

1. If you call FileReaper directly, always push before reaping: FileReaper << file; ... ; FileReaper.reap(file).
2. In forked workers, do not reap files registered in the parent — either disable reaping (FileReaper.reap = false) or re-register the file in the child pid context first.
3. If this surfaces from library internals, file a Selenium issue with the backtrace; do not work around it by mutating @tmp_files.

Example fix

# before
FileReaper.reap(tmp) # => WebDriverError: file not added for reaping

# after
FileReaper << tmp
FileReaper.reap(tmp)

# Before calling FileReaper.reap(file) (internal use):
if Selenium::WebDriver::FileReaper.reap? && Selenium::WebDriver::FileReaper.tmp_files.include?(file)
  Selenium::WebDriver::FileReaper.reap(file)
end

When it happens

Trigger: Calling FileReaper.reap(path) for a path that was not previously pushed with FileReaper << (path), or reaping after a fork where Process.pid differs (tmp_files is keyed by pid, so a child sees an empty list). Also reachable if internal code reaps a path twice.

Common situations: Almost always an internal-library bug or a forked-process teardown mismatch, not user-facing. Surfaces in test suites that fork workers, or when custom Service subclasses try to manage temp files themselves.

Related errors

• #{action.inspect} is not a valid action
• #{source.type} is not a valid input type
• #{source.type} is not a valid input type
• #{source.type} is not a valid input type
• #{source.type} is not a valid input type

AI-assisted analysis of SeleniumHQ/selenium@aa36b38e69 (2026-08-14). Data as JSON: /api/errors/a40db644447cb019. Report an issue: GitHub.