nyc VS Bazel

Native code coverage via v8 or istanbul.

Writing tests is essential, and knowing whether you test all the required cases for your logic is even more critical. The most common testing coverage tool is Istanbul, where you can see how well your tests exercise your code by lines, functions, and branches. Below is an example of how the test coverage report looks in your terminal:

Here is a quote from istanbul, one of the most used code coverage tool:

Unit testing framework was already implemented, using Vitest so I started hacking by setting up a coverage provider to explicitly identify the covered/uncovered lines and mentioned this to the maintainer in the comments. I used Istanbul 🇹🇷 for this purpose.

Your project probably has a coverage report. If you’re using Jest as your unit test runner, generating a coverage report is embedded in it. It is done with Istanbul under the hood, which generates a nice HTML page presenting the entire project unit test coverage.

> This is a rant written by someone with just enough understanding to be dangerous, but not quite enough wisdom to know why things are still the way they are. Most of the complaints raised are subtly inaccurate.

Author seems aware of CAP_NET_BIND_SERVICE: https://source.small-tech.org/site.js/app/-/issues/169 and https://github.com/istanbuljs/nyc/issues/1281 – the "side effects" are NodeJS explicitly checking for it, so that's a NodeJS thing and not a Linux thing.

Yet curiously it's completely unmentioned in this article, in spite that this is probably what started the author's dislike of privileged ports. I guess it was inconvenient as it got in the way of angrily ranting.

This approach will create two json coverage files, which will be merged together by NYC. Therefore the results will be purely local. If You don't mind using online tools like Codecov or Coveralls for merging data from different tests, then go ahead and use them. They will probably also be more accurate. But if You still want to learn how to get coverage from E2E, then please read through

https://istanbul.js.org/ measures how much of your code is covered by tests

Jest uses a package called Istanbul to provide test coverage metrics such as statement, branch, function, and line coverage so that you can understand and enforce the quality of your test suite, providing more confidence in releases.

So let’s see if nyc (the code coverage generator) can help with that. Hmm… this documentation seems interesting! So basically what I understand from it is that I need to collect all the reports from the different packages and then run the nyc report over it. The flow should be like this:

Wow, if you curl it, there's a lot of boilerplate code there.

Maybe built using Bazel?

https://bazel.build

Bazel by Google

Luckily a feature to limit the disk cache size is in development: https://github.com/bazelbuild/bazel/issues/5139

This is a problem that Bazel (https://bazel.build) solves in a very convenient way. You can just keep using the paths relative to the repository root, and as long as you properly declare your test needs that file it will access it without problems. Or you can use the runfile libraries to access them too.

When doing research for this lab exercise I looked at both vcpkg and conan. Both are package managers that would automate the installation and configuration of my program with its dependencies. However, when it came to releasing and sharing my program my options were limited. For example, the central public registry for conan packages is conan-center, but these packages are curated and the process is very involved. There was no way conan-center would accept a class project like mine. Alternatively, I could host a conan package on a public Artifactory repository, but accessing the package requires users to add the repository to their conan remote. This already sounded like too many steps to expect regular users to follow - I already haven't setup any conan remotes, there's no way I could expect regular users to know about conan remotes, let alone have conan installed on their system. After discussing with people online and consulting my instructor, I ultimately decided to do a GitHub release. However, in the future I was encouraged to look into using CMake or bazel.

NOTE: I won’t mention SBT and Leiningen here because, with all due respect, they are niche build tools. I also won’t discuss Kobalt for the same reason (besides, it’s no longer actively maintained). Additionally, I won’t touch upon Bazel and Buck in this context, mainly because I’m not very familiar with them. If you have insights or comments about these tools, please feel free to share them in the comments 👇

> None of this solves C's only REAL problem (in my opinion) which is the lack of dependency management.

Bazel solves this really nicely, I know some people have strong opinions on it but I cannot recommend it enough

https://bazel.build/