Some days ago, we decided to have a look at scrutinizer and how this service can help us to improve our code. To explain what scrutinizer is about, here is a short explanation from their website:

scrutinizer ci is a hosted continuous inspection service for open-source as well as private code. It runs static code analysis tools, runtime inspectors, and can also run your very own checks in your favorite language.

For more info check their website.

For now, we just implemented their service and setup the basic configurations.

As a first usage, we can see if the new pushed code creates new issues/errors during the scrutinizer analysis and how they recommend to change the code.

How to get the info about new errors

If you create a new pull-request at github, scrutinizer immediately starts analysing the code. Scrutinizer will add a comment to your pull-request with a link to the ongoing analysis process. After the analysis is finished, scrutinizer will update the comment with the actual test result.

If you click the link, you will see how your code-changes affect the master-branch.

In the above example, you see that the code-changes resolved some issues, but also created a new one. By clicking the “1 issue” link, you can have a detailed look at what changes caused this error and how you could resolve it.

As you can see, a new Bug was found in the code. Follow the link to the affected file and see the error with a suggestion how to fix it.

Once you have resolved the error and pushed your new code, scrutinizer will restart the process and verify your fix.

This was just a short overview of how we could benefit from this powerful tool. There are a lot more interesting things to discover. E.g. analysing existing code, code coverage by phpunit, …

Just have a look and try it out. Let me know how this works out for you. What additional experience do you have with scrutinizer? Could you cover all bugs/errors? Does this tool make sense to you at all?

Like to hear your thoughts.

Cheers, Marian