Exercise 1: A Second Independently-Testable Formatter Method — Possible Solution
====================================================================

THE NEW METHOD
------------------------------
    class PayrollReportFormatter:
        def format_report(self, total):
            return f"Total: ${total:.2f}"
        def format_summary_line(self, total, employee_count):
            return f"{employee_count} employees, total ${total:.2f}"

Follows this chapter's own format_report exactly - takes only the
values it needs as plain arguments, with no reference to PayCalculator
anywhere in its own signature or body.

VERIFYING IT TESTS WITH FAKE VALUES, ZERO DEPENDENCY
------------------------------
    format_summary_line tested with fake values, zero PayCalculator dependency:
      42 employees, total $9999.99

Calling formatter.format_summary_line(9999.99, 42) - two entirely
fabricated values, no PayCalculator constructed anywhere - correctly
produces the formatted string. Exactly like this chapter's own
format_report, this new method never needed real payroll data to be
verified as working correctly.

WHY THIS CONFIRMS THE SPLIT GENERALIZES, NOT JUST TO ONE METHOD
------------------------------
This chapter's own finding showed ONE formatter method staying
independently testable after the SRP split. This exercise confirms a
SECOND, different formatter method inherits the identical property
automatically - not because of anything special about this specific
method, but because PayrollReportFormatter's own class boundary
already excludes any dependency on PayCalculator. Every future method
added to this class, as long as it only formats values it's handed
directly, gets this same testability for free.

WHY THIS WORKS AS AN ANSWER
------------------------------
The new method follows this chapter's own established formatter
pattern exactly, and it's verified working correctly with fake inputs
rather than assumed to inherit the property just because it lives in
the same class.