growth-ring: Add remaining lint checks

[rkuris]

Removed a couple of index dereferences and unwraps.

Unfortunately, there's one spot where it isn't easy to add the lint check, so there is a block that disables it. It's still a lot better having it in most of the project and not in this block. (search for TODO: not allowed for the spot).

Rewriting the loop to be more idiomatic is probably the right move here, but that's outside the scope of this fix.

[xinifinity]

Maybe do not enable warning on indexing_slicing until we have fix most of them, as there are a bunch of places we need to explicitly disable them.

[rkuris]

The idea here is that we want to prevent future code from doing index slicing, so this is a stopgap until we can get rid of it.

The plan is to do the same thing with firewood, which has a lot more of these. I have a tool to do that. I think having them marked as such means someone will be able to easily find and fix them later.

[xinifinity]

Sounds good, not feeling strong about this.

[richardpringle]

Index slicing is a little less trivial than disallowing unwrap.

There are times when you call a method to get a slice of a specific length and as long as your index is smaller than said length, you're fine indexing into the slice. It's these cases where the compiler wouldn't even do a bounds check.

Do we want to keep this type of behaviour and add the lint-ignore attribute?

It's possible that this never comes up. I can already think of places I do this but could do it without indexing. But in the case where it does, it would suck to make a function return a Result if it didn't need to.

I'm approving the PR because I agree with the change in general, but I would like to have a strategy in case the aforementioned scenario does come up.

[rkuris]

Index slicing is a little less trivial than disallowing unwrap.

Please have a look at the other 2 proposed lint changes going forward that were approved in #356 to ensure we're all in agreement on direction before I continue with these changes.