Refactor how nutrients are stored and processed

[jbytheway]

Summary

SUMMARY: Infrastructure "Refactor how nutrients are stored and processed"

Purpose of change

While working towards a fix for #28180, I found the code working with nutrients to be difficult to follow and with different aspects mixed in the same function. This is an attempt to separate concerns to make it easier to work on #28180.

Describe the solution

• Existing struct nutrients renamed food_summary and a new nutrients struct introduced for a subset of its members.
• Use that new nutrients struct in islot_comestible.
• Overload arithmetic operators on nutrients to make it easy to perform nutrient-related calculations.
• Refactor the code which computes item nutrients and deals with digestion to use this new struct.
• In particular, separate out the logic for default base nutrients from the code which combines nutrients for components of crafted food.

Most of this is intended to have no observable change to game behaviour, with one exception: previously the tapeworm parasite halved all vitamins a Character consumed. Now it halves both the vitamins and the calories.

While testing, I discovered that the stave_test was buggy -- it didn't actually test anything, and it had the potential to get stuck in an infinite loop. I fixed it, but in so doing found that it was already failing on master. Players are starving faster than the test indicates they ought to (30 days vs 46). This may be a problem, but it's orthogonal to this PR. @KorGgenT might want to look into that.

Describe alternatives you've considered

Could have included water content in the nutrients struct, but currently water is not calculated based on components so it didn't really fit with the code structure.

Testing

Unit tests only.

[Davi-DeGanne]

Regarding this:

The only intended observable behaviour change from this commit is that
tapeworms now consume half of the calories you ingest, as well as half
the vitamins (they used to only affect vitamins)

I actually had the exact same line of reasoning when I did the stomach refactor (#35143) but I discovered I was wrong -- tapeworm does effect calories AFAICT, so make sure this refactor doesn't make tapeworms twice as effective as intended.

Other than that, I think this change is awesome. Adding arithmetic functions to nutrients was a great idea, it will really help my efforts when I implement my planned changes to stomach processing speeds.

[jbytheway]

I actually had the exact same line of reasoning when I did the stomach refactor (#35143) but I discovered I was wrong -- tapeworm does effect calories AFAICT, so make sure this refactor doesn't make tapeworms twice as effective as intended.

I'm fairly sure hunger_chance only affects how hungry the player feels, not how much calories they get. Hard to say for sure because the relevant code is quite abstracted. I'll see if I can easily add a test for this.

[jbytheway]

Thanks for the review BevapDin. I was only moving that code from one place to another, but indeed there was a lot worth improving in it.

[jbytheway]

I think I've now resolved all the above comments.

[Davi-DeGanne]

Well, your test as written wouldn't have properly discovered whether my concern was valid or not, since effects are processed over time, and you don't pass any time at all in the test.

That said, I have read through the code and determined that unless there's something really weird going on, you are right -- hunger_chance and hunger_min don't effect calorie intake/absorption.

[jbytheway]

Yes, I realise it was a bit of a weak test. Thanks for taking the time to check and reassure us both.