Python Lambda Late Binding Bug Charged 500 Wrong

# ── Understanding what a lambda actually is ──

# A full named function — has a name, lives permanently in the module namespace,
# appears by name in tracebacks, can carry a docstring and type hints
def apply_discount_named(price: float) -> float:
    """Apply a 10% discount to the given price."""
    return price * 0.9

# The exact same logic as a lambda — no name, no docstring, no type hints
# Assigning it to a variable like this is actually a PEP 8 anti-pattern,
# but it illustrates what a lambda IS before we get to where it BELONGS
apply_discount_lambda = lambda price: price * 0.9

# Both are callable objects — Python's type system treats them identically
original_price = 49.99

print(apply_discount_named(original_price))   # 44.991
print(apply_discount_lambda(original_price))  # 44.991 — same result, same computation

# Python reports the same type for both — there is no 'lambda type'
print(type(apply_discount_named))   # <class 'function'>
print(type(apply_discount_lambda))  # <class 'function'> — identical type

# The difference shows up in tracebacks and introspection
print(apply_discount_named.__name__)   # 'apply_discount_named' — useful in debugger
print(apply_discount_lambda.__name__)  # '<lambda>' — useless in a 3 AM incident

print()

# Lambda with multiple parameters — comma-separated, same as def
calculate_line_total = lambda unit_price, quantity, tax_rate: unit_price * quantity * (1 + tax_rate)

print(calculate_line_total(9.99, 3, 0.08))  # 32.3676

# Lambda with no parameters — valid but rarely needed outside tests
get_placeholder = lambda: "N/A"
print(get_placeholder())  # N/A

# Conditional expression (ternary) inside a lambda — the right way to branch
classify_score = lambda score: "pass" if score >= 50 else "fail"
print(classify_score(72))  # pass
print(classify_score(31))  # fail

# ── Real production patterns where lambda earns its place ──

# EXAMPLE 1: Sorting a list of employee records by various keys
employees = [
    {"name": "Priya",   "department": "Engineering", "salary": 95000,  "tenure_years": 4},
    {"name": "Marcus",  "department": "Marketing",   "salary": 72000,  "tenure_years": 7},
    {"name": "Chen",    "department": "Engineering", "salary": 110000, "tenure_years": 2},
    {"name": "Fatima",  "department": "HR",          "salary": 68000,  "tenure_years": 9},
    {"name": "Jordan",  "department": "Engineering", "salary": 88000,  "tenure_years": 3},
]

# Single-field sort — classic lambda use case
by_salary_desc = sorted(employees, key=lambda emp: emp["salary"], reverse=True)
print("By salary (highest first):")
for emp in by_salary_desc:
    print(f"  {emp['name']:<10} ${emp['salary']:,}")

print()

# Multi-key sort using a tuple — lambda returning a tuple sorts by first element,
# then by second element for ties. Clean and self-explanatory.
by_dept_then_salary = sorted(
    employees,
    key=lambda emp: (emp["department"], -emp["salary"])  # dept alphabetically, salary descending
)
print("By department, then salary within department:")
for emp in by_dept_then_salary:
    print(f"  {emp['department']:<15} {emp['name']:<10} ${emp['salary']:,}")

print()

# EXAMPLE 2: map() — apply a transformation to every element
# Computing total compensation (salary + 15% bonus) for each employee
total_compensation = list(
    map(lambda emp: {**emp, "total_comp": round(emp["salary"] * 1.15)}, employees)
)
print("Total compensation (salary + 15% bonus):")
for emp in total_compensation:
    print(f"  {emp['name']:<10} ${emp['total_comp']:,}")

print()

# EXAMPLE 3: filter() — keep only elements matching a condition
# Senior engineers: Engineering department with 3+ years tenure
senior_engineers = list(
    filter(
        lambda emp: emp["department"] == "Engineering" and emp["tenure_years"] >= 3,
        employees
    )
)
print("Senior engineers (Engineering dept, 3+ years):")
for emp in senior_engineers:
    print(f"  {emp['name']} — {emp['tenure_years']} years")

print()

# EXAMPLE 4: Chaining map and filter — keep lazy for efficiency on large datasets
# Filter to affordable prices, then format as currency strings
raw_prices = [29.99, 49.99, 9.99, 149.99, 89.99, 19.99]
tax_rate = 0.08

# Chain stays lazy until list() materializes it — efficient for large datasets
formatted_affordable = list(
    map(
        lambda price: f"${price:.2f}",
        filter(
            lambda price: price <= 50.00,
            map(lambda p: round(p * (1 + tax_rate), 2), raw_prices)
        )
    )
)
print("Affordable prices after tax (formatted):", formatted_affordable)

# ── The decision framework in practice ──

users = [
    {"username": "alex_92",    "age": 31, "verified": True,  "score": 88, "plan": "pro"},
    {"username": "beta_tester","age": 17, "verified": False, "score": 72, "plan": "free"},
    {"username": "carol_dev",  "age": 25, "verified": True,  "score": 95, "plan": "enterprise"},
    {"username": "dan_h",      "age": 19, "verified": True,  "score": 61, "plan": "pro"},
    {"username": "eve_ml",     "age": 28, "verified": True,  "score": 79, "plan": "free"},
]


# ✅ CORRECT USE OF LAMBDA — single expression, used once, inline at the call site
# The key function is self-explanatory. No def required, no namespace pollution.
top_scorers = sorted(users, key=lambda user: user["score"], reverse=True)
print("Top scorers:", [u["username"] for u in top_scorers])

# ✅ CORRECT USE OF LAMBDA — sorting by a derived tuple key, still one expression
by_plan_then_score = sorted(
    users,
    key=lambda user: (user["plan"], -user["score"])
)
print("By plan then score:", [(u["plan"], u["username"]) for u in by_plan_then_score])

print()


# ✅ CORRECT USE OF DEF — complex predicate, multiple conditions, needs documentation
# This logic has business meaning that deserves to be named, tested, and documented.
# Burying it in a lambda would make it untestable and undiscoverable.
def is_eligible_for_beta(user: dict) -> bool:
    """
    A user qualifies for beta access if they are an adult (18+),
    have a verified account, and have achieved a quality score above 70.
    Free plan users are excluded from this cohort regardless of score.
    """
    return (
        user["age"] >= 18
        and user["verified"]
        and user["score"] > 70
        and user["plan"] != "free"
    )

eligible_users = list(filter(is_eligible_for_beta, users))
print("Eligible for beta:", [u["username"] for u in eligible_users])

print()


# ❌ ANTI-PATTERN — forcing multi-condition logic into a lambda
# Same output, but zero readability benefit. Cannot be documented, cannot be unit tested
# by name, and produces a '<lambda>' traceback if it raises a KeyError.
bad_filter = list(filter(
    lambda u: u["age"] >= 18 and u["verified"] and u["score"] > 70 and u["plan"] != "free",
    users
))
print("Same result, worse code:", [u["username"] for u in bad_filter])

print()


# ── THE LATE BINDING TRAP — the most common lambda bug in production ──

print("--- Late binding demonstration ---")

# WRONG: all lambdas see the final value of i after the loop completes
bad_lambdas = [lambda: i for i in range(5)]
print("Without binding (WRONG):", [f() for f in bad_lambdas])  # [4, 4, 4, 4, 4]

# CORRECT: i=i captures the current value at definition time via default arg
good_lambdas = [lambda i=i: i for i in range(5)]
print("With binding (CORRECT): ", [f() for f in good_lambdas])  # [0, 1, 2, 3, 4]

print()

# Realistic production scenario: building per-user report generators
# The late binding trap is easy to hit when the loop variable is meaningful
user_ids = ["usr_101", "usr_102", "usr_103"]

# WRONG — every report generator will fetch usr_103's data
bad_generators = [lambda: f"Generating report for {uid}" for uid in user_ids]
print("Bad generators (all same):", [g() for g in bad_generators])

# CORRECT — each generator captures its own uid
good_generators = [lambda uid=uid: f"Generating report for {uid}" for uid in user_ids]
print("Good generators (distinct):", [g() for g in good_generators])

print()


# ── FUNCTOOLS.PARTIAL AS AN ALTERNATIVE TO LAMBDA ──
# For some cases, functools.partial is more readable than lambda
import functools

def apply_rate(value: float, rate: float) -> float:
    """Apply a multiplier rate to a value."""
    return round(value * rate, 2)

prices = [10.00, 25.00, 50.00, 100.00]

# Lambda approach — fine, readable
with_tax_lambda = list(map(lambda p: apply_rate(p, 1.08), prices))

# functools.partial approach — arguably clearer about what is being fixed
apply_tax = functools.partial(apply_rate, rate=1.08)
with_tax_partial = list(map(apply_tax, prices))

print("With tax (lambda):  ", with_tax_lambda)
print("With tax (partial): ", with_tax_partial)  # identical output