0x50Lesson 6 of 14

Code review

Review pull requests kindly and effectively, use suggestions and review states, and route reviews with CODEOWNERS.

26 min 7-question quiz 2 code exercises

By the end of this lesson you can

Leave clear, kind, actionable review comments, including suggested changes
Use review states - comment, approve, request changes - and required reviews
Route reviews automatically with a CODEOWNERS file

Code review catches bugs, spreads knowledge across the team, and keeps a codebase consistent. On GitHub you review in a PR’s Files changed tab: click a line (or drag across several) to comment, and submit everything as one review with a state:

Comment - feedback without a verdict.
Approve - good to merge.
Request changes - must be addressed first. With branch protection, this blocks merging until the reviewer approves.

A suggested change is a comment containing a block marked suggestion. The author can apply it with one click, which commits it for them:

a review comment with a suggestion

1**nit:** this name hides what the number means.
2
3```suggestion
4MAX_DASH_DISTANCE = 3  # tiles
5```

Good reviews are about the code, never the person, and make each comment’s weight clear:

Prefix optional comments: nit: (tiny), suggestion:, question:. Make blocking ones explicit.
Explain why: “This runs once per frame, so the list copy will add up” beats “don’t copy the list”.
Ask rather than accuse: “What happens if enemies is empty?”
Praise good things - it tells people what to keep doing.
Review promptly. A PR waiting three days for review is a PR whose author has moved on.

And as the author: don’t take comments personally, reply to each one (or resolve it with a fix), and say thanks.

Try it

Review this pull request

A contributor’s PR adds a high-score saver to Dungeon Dash. Click every line you’d comment on, then check what you missed.

src/scores.py (new file) (1 of 1)Flags found 0/0

Click every part that looks suspicious. There are 5.

import json API_KEY = "sk_live_51HxQ..." def save_score(name, score): data = json.load(open("scores.json")) data.append({"name": name, "score": score}) data.sort(key=lambda entry: entry["score"]) print("DEBUG", data) json.dump(data, open("scores.json", "w")) def top_score(): return load()[0]

CODEOWNERS

A CODEOWNERS file (in .github/, the root or docs/) maps paths to the people or teams who own them. When a PR touches matching files, GitHub automatically requests their review - and branch protection can require a code owner’s approval.

Each line is a pattern and one or more owners. Patterns work like .gitignore, and the last matching line wins - so put general rules first and specific ones after.

.github/CODEOWNERS

1# Default owners for everything
2*                 @dungeon-dash/maintainers
3
4# Art and audio
5*.png             @sam-sprites
6/assets/audio/    @dungeon-dash/audio
7
8# The physics engine
9/src/engine/      @mira
10
11# Docs: anyone on the docs team, plus mira for the API reference
12/docs/            @dungeon-dash/docs
13/docs/api/        @mira @dungeon-dash/docs

Key takeaways

Submit reviews as Comment, Approve or Request changes; suggestions let authors apply fixes in one click.
Comment on the code, explain why, label nits, ask questions and point out what’s good.
CODEOWNERS requests reviews automatically; the last matching line wins.
Branch protection can require approvals and code owner reviews before merging.

Lesson quiz

7 questions · pass with 5 correct · up to 50 XP

Passing this quiz completes the lesson and keeps your streak going. Questions you miss come back in review sessions later.

Practice: automate GitHub chores with Python

Real GitHub work involves lots of small automation: matching CODEOWNERS, expanding build matrices, bumping versions, reading the API. Write those helpers in Python and run them against sample inputs - locally in your browser, with no GitHub account needed.

Exercise 1

Who owns this file?

+25 XP

The input is a CODEOWNERS file, a line ---, then changed file paths (starting with /). For each path print its owners from the last matching rule, or (no owners). Then print reviewers: ... - every distinct owner across all files, sorted.

Simplified matching: a pattern ending in / matches everything under that folder; a pattern starting with / is matched from the root; otherwise the pattern is matched (with fnmatch) against the file name. Skip blank lines and # comments.

Dungeon Dash
No default owner

main.py

Loading editor…

Python runs in a sandboxed browser worker with a 60 second time limit. Its runtime loads from the Pyodide CDN; your code stays in this browser.

Exercise 2

Can this PR merge?

+25 XP

Each input line is a review event in time order: user STATE, where STATE is APPROVED, CHANGES_REQUESTED or COMMENTED. Only each user’s latest approve or request-changes counts (a later COMMENTED doesn’t change their verdict). The branch requires 2 approvals and no outstanding change requests.

Print each reviewer’s verdict sorted by name (mira: APPROVED, or sam: none for a reviewer who only commented), then mergeable: yes or mergeable: no (1 of 2 approvals, changes requested by mira) - listing whichever reasons apply, separated by commas.

Approved after changes
Blocked

main.py

Loading editor…

Python runs in a sandboxed browser worker with a 60 second time limit. Its runtime loads from the Pyodide CDN; your code stays in this browser.

Questions about this lesson

Stuck? Ask. Figured something out? Share it. Explaining is one of the best ways to learn.

Loading posts…