6. Error Handling

Tutorial

The reference tables throughout this doc note "raises KeyError/ValueError if missing" — this is how you actually catch those. Interview solutions rarely need heavy exception handling, but the syntax matters for safe lookups and for LeetCode "Design" problems that manually raise StopIteration (see §12).

# ══ TRY / EXCEPT ════════════════════════════════════════════════
try:
    val = d[key]
except KeyError:
    val = 0                       # equivalent effect to d.get(key, 0)

try:
    x = arr[i]
except IndexError:
    x = None

# Multiple exception types
try:
    n = int(s)
except (ValueError, TypeError):
    n = 0

# ── else / finally ──────────────────────────────────────────────
try:
    result = 10 / divisor
except ZeroDivisionError:
    result = float("inf")
else:
    print("no error occurred")    # runs only if try succeeded
finally:
    print("always runs")          # cleanup, runs no matter what

# ══ RAISE ═══════════════════════════════════════════════════════
def safe_divide(a: int, b: int) -> float:
    if b == 0:
        raise ValueError("division by zero")
    return a / b

# Re-raise with context (debugging chained failures)
try:
    process()
except KeyError as e:
    raise RuntimeError("processing failed") from e

Reference

Exception	Raised by	Common DSA cause
KeyError	d[missing_key]	Use d.get(k, default) instead
IndexError	arr[out_of_range]	Off-by-one in loop bounds
ValueError	int("abc"), lst.index(missing)	Bad input parsing
ZeroDivisionError	x / 0	Guard divisor before dividing
RecursionError	Recursion too deep	See sys.setrecursionlimit (§15)
StopIteration	next(it) past the end	Manual iterator protocol (§12)
try/except/else/finally	—	else runs iff no exception; finally always runs
raise X("msg")	—	Explicitly signal an error
raise X from e	—	Chain exception context

---