chapter_06_object_references

Chapter 6 — Object References, Mutability, and Recycling

Theme: Names are labels stuck to objects — not boxes containing them. This single mental model explains aliasing, mutability bugs, copy semantics, and garbage collection.

What You'll Learn

Topic	Key insight
Variables as labels	Assignment binds a name to an object, not a copy
is vs ==	Identity vs value — use == for almost everything
Tuple immutability	Tuples can't change their references, but referents can mutate
Shallow copy	New outer container, shared inner objects
Deep copy	Full independent duplicate, handles cycles
Call-by-sharing	Parameters are aliases of arguments
Mutable defaults	Defaults evaluated once — shared across all calls
del and GC	del removes a name; CPython uses reference counting

Key Files

• examples.py — 6-part walkthrough of all book examples
• exercises.py — 5 exercises with verification
• mini_project.py — Document graph inspector (aliasing, deepcopy, weakref)
• benchmarks.py — is vs ==, copy vs deepcopy speed
• notes.md — Quick reference tables
• pitfalls.md — 7 production bugs
• interview_questions.md — 9 questions L3→L6
• architecture_notes.md — CPython refcount & GC internals

30-Second Rules

y = x              → alias (same object)
y = list(x)        → shallow copy (new outer, shared inners)
y = copy.deepcopy(x) → deep copy (everything new)

x is None          → correct (singleton check)
x == "hello"       → correct (value check)
x is "hello"       → WRONG (interning is an implementation detail)

def f(items=[]):   → BUG (mutable default shared)
def f(items=None): → CORRECT (create fresh list inside)
  items = list(items) if items else []

Reference: Fluent Python 2nd ed., Chapter 6 — pages 201–223