schmittjoh
So far, scrutinizer has already provided you with condensed output in the form of comments, and auto-generated patches. As of now, it also gives you access to the raw metric data that is gathered for your project. You can use these metrics to spot some of the most problematic areas and prioritize refactoring efforts, and also a couple of cool features that are in the pipeline (see below).
As a start, we support metrics generated by open-source tools like php code coverage (top project risks, least tested methods), php lines of code (general size metrics), and pdepend (complexity, and coupling information). A typical build configuration for these tools looks like:
before_commands:
- composer install --dev --prefer-source
tools:
php_code_coverage: true
php_pdepend:
excluded_dirs: [vendor]
php_loc:
excluded_dirs: [vendor]
Also, you can gather metrics for your own code. For example, scrutinizer benchmarks the serialization performance of jms/serializer on each code change. The benchmark script is located in the serializer repository and is then invoked as a custom command. On the website, this is shown as a nice column chart. The configuration looks like this:
before_commands:
- composer install --dev --prefer-source
tools:
custom_commands:
-
command: php tests/benchmark.php json 10 tests/benchmark.json
scope: project
output_file: tests/benchmark.json
iterations: 5
This also lays the foundation for some future features which will come shortly. One of these is statistical graphs which add a time dimension to all the gathered metrics and allow you to track development over time. The other feature is badges which will allow you to pose with your code coverage percentage for example :)
Stay tuned for these, and do not forget to enable the tools mentioned above so you can take full advantage of the upcoming features!