Monorepo is short for monorepository. With it, code for many projects is stored in the same repository. I like to use monorepo for NodeJS and React projects but sometimes it’s difficult to use monorepo when performing static code analysis (SCA), particularly when it comes to code coverage and code patterns. However, Codacy’s tool helps with this. Let me show you how below:
Because many projects are stored in the same repository, many coding languages co-exist. Therefore, code coverage (checking code for errors) becomes complicated.
A tool that enables reports in multiple programming languages is helpful to address these concerns. With Codacy, I receive backend and frontend reports for each language in each repo. The weight is also calculated.
You can see the documentation here and an excerpt below:
I also worry about code patterns when using monorepos with static code analysis. Linting tools detect code patterns. However, these tend to be specific to coding languages. With multiple projects in multiple coding languages things can get tricky.
Example with ESLint
In the src folder of the project create a client and server folder. This is where it makes sense to have the .eslintrs.js file. In this, common rules can be set (or left empty).
Then, inside each client or server folder, an additional .eslintr.js, extending the src file can be added. This overrides the previously defined rules.
Locally seems to be going as expected. Now, to Codacy. Under the “Code Patterns” tab, I chose “Configuration File” instead of the Rules from the UI.
We get the same, expected results:
As you can see, with these techniques (and some help from Codacy) monorepo should no longer be an obstacle with static code analysis.
Want more from Helio, a Solutions Engineer at Codacy? Follow him on twitter.