How does a + b dispatch in Python? | Interview Question

Q: How does a + b dispatch in Python?

**Target Audience:** Beginners to intermediate Python developers who have written a class and want to know how operators work on it. **Difficulty:** Medium **Python version:** Behaviour and bytecode checked on CPython 3.11 (outputs identical on 3.12 and 3.13); both examples also run in the page on Pyodide 0.25.1 (CPython 3.11.3). > **How to read this doc:** Plain language first; terms an interviewer expects are tagged `📌 Interview term:` and defined where they appear, then repeated in the Quick Glossary. Every output shown was printed by a real run. ## 1. Why This Even Matters — A Story First Two people from different countries want to combine their savings. One holds dollars, the other euros. The rule at the bank is polite: **the person on the left speaks first**. "Can you add my dollars to those euros?" If the left person says "sorry, I only know dollars", the clerk turns to the **person on the right** and asks the mirrored question: "can *you* add those dollars to your euros?"

The problem

Question presented to candidate: "When Python evaluates a + b, what actually happens? Which object's method gets called, what happens if it does not know how to add the other type, and how is a += b different?"

What a strong answer should cover:

a + b calls type(a).__add__(a, b); if that is missing or returns NotImplemented, Python calls type(b).__radd__(b, a); if that also fails, TypeError: unsupported operand type(s) for +.
NotImplemented is a special value to return (not NotImplementedError to raise), meaning "I cannot handle this type, ask the other side".
The subclass rule: if type(b) is a subclass of type(a) and overrides __radd__, b.__radd__ is tried first; if both operands have the same type, __radd__ is never tried.
Methods are looked up on the type, not the instance, so patching obj.__add__ has no effect on obj + x.
a += b tries __iadd__ first (in-place, e.g. list mutates and aliases see it), falling back to a = a + b (e.g. tuple, int, str); the t[0] += [x] tuple gotcha.
CPython detail: since 3.11 all binary operators compile to a single BINARY_OP instruction with the operator in its argument.

Code / implementation expected: Optional — a small class with __add__ and __radd__ returning NotImplemented for unsupported types.

operators
data-model

Try it yourself

Sketch your approach before you read the solution. That is what the interview tests.

Stuck? Get a hint.

A short nudge toward the approach. It will not give the answer away.

Solution

Target Audience: Beginners to intermediate Python developers who have written a class and want to know how operators work on it. Difficulty: Medium Python version: Behaviour and bytecode checked on CPython 3.11 (outputs identical on 3.12 and 3.13); both examples also run in the page on Pyodide 0.25.1 (CPython 3.11.3).

How to read this doc: Plain language first; terms an interviewer expects are tagged 📌 Interview term: and defined where they appear, then repeated in the Quick Glossary. Every output shown was printed by a real run.

1. Why This Even Matters — A Story First

Two people from different countries want to combine their savings. One holds dollars, the other euros.

The rule at the bank is polite: the person on the left speaks first. "Can you add my dollars to those euros?" If the left person says "sorry, I only know dollars", the clerk turns to the *

The solution is hidden so you can try first

6 min read · 10 sections

Run the code

Edit the code and run it here. Each console line with an expected value becomes a test.

PythonTracing __add__, __radd__, NotImplemented and the subclass rule

Practice

Python+= and __iadd__, the tuple gotcha and the BINARY_OP bytecode

Practice

Discussion

Other approaches, follow-ups and how it went in real interviews.

Sign in to join the discussion.

No comments yet. Share how you would answer it.

The solution is hidden so you can try first

Keep your momentum going.