Exercise 1: Tag vs. Class vs. Id, in Both Source Orders — Possible Solution
====================================================================

THE TEST
------------------------------
    elem = <p id="x" class="y">

    r_tag   = Rule(['p'],  [Declaration('color', 'black')])
    r_class = Rule(['.y'], [Declaration('color', 'blue')])
    r_id    = Rule(['#x'], [Declaration('color', 'red')])

    sheet_a = Stylesheet([r_tag, r_class, r_id])   # id rule LAST
    sheet_b = Stylesheet([r_id, r_class, r_tag])   # id rule FIRST

    cascade(elem, sheet_a)['color']
    cascade(elem, sheet_b)['color']

RESULT
------------------------------
    sheet_a -> 'red'
    sheet_b -> 'red'

Both stylesheets produce the exact same winner, regardless of where
the id rule sits in source order.

WHY SOURCE ORDER NEVER GETS A CHANCE TO MATTER HERE
------------------------------
cascade() sorts every matching rule by the key `(specificity, order)` --
specificity FIRST, source order only as a secondary key. The three
rules here have genuinely different specificities:

    specificity('p')  -> (0, 0, 1)
    specificity('.y') -> (0, 1, 0)
    specificity('#x') -> (1, 0, 0)

Comparing tuples lexicographically, (1,0,0) is strictly greater than
both of the others, in EITHER stylesheet -- the id rule's own
specificity tuple has a 1 in the very first position, which no
combination of classes or tags can ever match or exceed (short of
also using an id). Because sort() is comparing (specificity, order)
pairs, and the id rule's specificity component alone already makes it
the largest of the three tuples, the order component of that
comparison is never even consulted to break a tie -- there simply
isn't a tie to break.

WHY THIS WORKS AS AN ANSWER
------------------------------
Source order is only a TIE-BREAKER, the second element of the sort
key -- it only has any influence over the outcome when the first
element (specificity) is identical between two competing rules. Here,
all three rules have distinct, unequal specificity tuples, so the
"loudest" rule (the id selector) always wins the comparison on the
first key alone, and where it happened to be written in the
stylesheet is irrelevant to the result. This is exactly why real CSS
authors can't reliably "win" a specificity fight just by writing their
rule later in the file -- source order is a fallback, not a trump
card.