Julien Delange, Founder and CEO
Julien is the CEO of Codiga. Before starting Codiga, Julien was a software engineer at Twitter and Amazon Web Services.
Julien has a PhD in computer science from Universite Pierre et Marie Curie in Paris, France.
Code reviews help developers write better code and are a key element of improving software security. The purpose of a code review is to enable developers to find and fix bugs before they reach production. It also helps to ensure that all the developers in an organization share a common coding style. Code review tools can automate the process of reviewing code, making it faster and easier for a team to review the code.
In this article, we review the best tools to help you be more productive with your code reviews.
Pull Requests should always be reviewed by at least one peer. But once the code is approved and all verifications passed, the code should be merged to ensure all code is merged in the default branch and unblock developers waiting for this change.
Some developers merge their pull requests even if some verifications do not pass. And some others take a few hours before merging their pull request, blocking other developers from waiting for the new code to be merged.
Mergify automates your pull requests merge based on rules. You define rules (see example) that trigger an automatic merge. For example, when a pull request has no conflict, at least one review, and some check (such as Codiga) pass, the pull request is automatically merged in the default branch.
Mergify ensures that all necessary checks pass before a pull request is merged, unblocking developers (or users) waiting for a new change.
Making sure that your code does not have any errors is hard. Developers reviewing your code ensure that your logic is correct but can miss important issues that may need additional work later.
Codiga is a code analysis and automated code reviews platform that gives you feedback on your code within seconds. Codiga flags any coding error in your pull requests (such as security, safety, performance, or design issues). It also detects if a function is too complex or if the pull request introduces code duplicates that could be refactored into functions.
Code Coverage is one important measure of how well your code is tested and proven correct. I have talked about it in my recent book Technical Debt in Practice: it's always important to keep an eye on your code coverage and ensure new code does not reduce code coverage.
Codecov is a code coverage analysis platform. When used with Pull Request, it reports the coverage for each added/modified file and highlights code coverage regression.
Commit messages shoud be descriptive and reference what the new code is solving or fixing.
A commit message such as
"Fix problem we found two days ago" does not help
reviewers to have context about the changes or developers who will eventually
look at the Git history and wonder what this code is doing.
Instead, the commit message should clearly reference the issue it solves. A message
[CI-1234] Fix authentication errors from service ABC has a clear reference
to the corresponding ticket (
CI-1234) and a short description of what the code
Developers often forget to reference issues in their commit message. And often write messages that do not clearly describe what the change is doing. Commitcheck comes to the rescue. It checks commit messages and ensures they reference tickets and satisfy developers' requirements.
If you are working in a large team, making sure that commit messages are consistent is an important code hygiene rule, and we highly recommend you ensure that commit messages and descriptions are consistent.
Have you ever committed private SSH keys, certificates, or other sensitive information in your Git repository? Adding such data is a problem, poses a significant security risk, and requires immediate remediation. What is worse is when you do not know that you or one of your colleagues added credentials or certificates in a Git repository, exposing sensitive data without knowing. Doing this is like dropping your keys in front of your doorstep: it's an invitation for any intruder to come and steal data.
For this reason, developers should use tools to prevent such situations. They should carefully check every file being added to their Git repository and also install tools to prevent them from doing so.
GitGuardian is a service that scans your source code and finds API keys, passwords, certificates, or encryption keys before it is deployed or merged. The service automatically alerts you when sensitive materials are found in your source code.