Hah yeah it's perhaps not the easiest introduction to procedural generation! Did you make any progress?
I've been thinking about how I would go about doing this, which is perhaps not the simplest way but I guess you could remove/modify some steps to make them simpler.
For positioning there's a technique I've been trying out recently called "Poisson disc sampling" which allows you to randomly place objects in an area, but ensure they aren't too close to each other. Here is a good animated explanation of how it works. There's also this video by Sebastian Lague.
For planets I used
round(random(min, max)) to get a count, then placed them at a random distance/angle from the star. We have a behaviour called "Orbit" that makes them move, and "pinned" it to the star so it follows it. Then used more random calculations for the properties.
Naming things is a little more complicated, there's a technique called Markov Chains which can be used for this. I've never written one but it looks quite easy, you can hold the name fragments in an array. There's a built in tool called the "array editor" which can be used to create and edit array files, which you can then load into an array object at runtime.
As you can see quite a lot of this can be done without arrays, you can just use loops and local variables. If you want your code to be able to show the relations between planetary systems, suns, planets, moons you might have to create an array for lookups.
There's plenty more that can be done from the demo project, hopefully it shows you enough to figure out the rest.