Pull Request Review Guidelines
To maintain high code quality standards while ensuring efficiency and meeting deadlines when reviewing our pull requests, we compiled this list of guidelines.
First and foremost
No blaming, but constructive advice and suggestions. No one is perfect.
Goals and purpose
Checking whether the PR actually does what it’s supposed to do
Avoiding the introduction of new bugs
Keeping the team up to date about changes in the code base
Knowledge transfer between developers
Tip
Have a personal goal of a minimum of reviews. You should aim to make two reviews for every own PR that you open.
Checklist
1. General overview
Read issue and PR description carefully (including background and context) and get an overview of the changed files.
Does the PR match the issue description? Are there any unrelated changes or additional resolved issues?
Note
When the changes are out of the issue’s scope, suggest to open another separate PR
2. Functional tests
Checkout the branch, open the CMS and test the main functionality, that the PR implements
Test for edge cases in user inputs etc.
Have a look at the dev server log and keep an eye on any errors or warnings
Test the GUI:
Test potential breaks in the GUI
Is the design responsive?
Are all hard coded strings translated?
Are UI elements spelled correctly?
Is the wording consistent?
Is gender-sensitive language used in the German translations?
Does the GUI make sense from a user perspective?
3. Code review
Now read the code in detail and check whether the PR fulfills the requirements:
Try to find potential problems in edge cases by following along the code and think about breaks in conditions
Check tests, that already exist. Do they miss edge cases or can they be extended for the current PR?
Are exceptions handled gracefully?
Does it have any negative side effects (e.g. slower execution)?
Is the logic correct (endless loop, deadlock, etc…)?
Is the code understandable/retraceable?
Are the files & functions reasonably large or would it be better so split them?
Is the code nesting sensible or does it have too many levels?
Do the variables have unambiguous naming?
Has redundant code been avoided?
Is the change covered by the test?
Is the styling consistent?
Are there enough inline comments on code?
Can complicated loops be replaced by syntactic sugar?
4. Feedback
Be kind.
Explain your reasoning.
Balance giving explicit directions with just pointing out problems and letting the developer decide.
Encourage developers to simplify code or add code comments instead of just explaining the complexity to you.
When you feel a comment is purely subjective, mark it as “nit picky” and leave it up to the PR author whether they want to implement your feedback.
If you are unsure about the quality of your review, give a short review report - describe how and what you have tested.
More information: How to do a code review - Google’s Engineering Practices documentation