Quick recap

We followed a fairly straightforward path to write our read_gff() function.

But you should note a few things that we did to make this process as easy as possible - here are a few:

1. We kept a fast iteration time (by using a small subset of the data to test on) - important to avoid getting bored waiting every time to change the code.
2. We used the specification to know what to aim for, and gave our function a single well-defined thing to do.
3. We created test case(s) with small datasets and used them to tell when we'd got it right.

And we also chose a pretty good name for our function - read_gff() makes this similar to read_tsv(), read_csv() and so on from other packages.