Gun spawn overhaul

[Holli-Git]

Summary

Content "Overhaul gun spawning"

Purpose of change

Gun spawns are all over the place. There's not much rhyme or reason why a gun would spawn in its quantity besides guesswork. This PR aims to bring gun spawns in-line with real-world numbers, so there's less of a headache when adding new guns. This information was based mostly on gunbroker results, and #37571.

Describe the solution

Uses gunbroker and #37571 to gauge how often a caliber should spawn, and how often a gun should spawn. This is by no means perfect and has lots of flaws, but currently its the best solution due to a lack of any other information.

Removes itemgroups like rare/obscure guns and smg groups. This is because before, "rare guns" had a fixed chance to spawn somewhere, despite rarity not really working like that. Instead, guns now have to get a good dice roll to spawn somewhere, instead of it being fixed. SMG spawns were removed as a civilian can't really own an actual SMG, and so it was better split into either pistols or rifles.

Adjusts some behavior of conversion gun spawns, as they could spawn in guns you could not easily legally own.

Describe alternatives you've considered

Not doing this, it was a whole lot of work.

Testing

Loaded up a world before changes, loaded into the same one after without any errors. Tested a few itemgroups in game, such as the display ones to make sure no ammo spawned with it, etc.

Additional context

Caused by #75925. Also requires a lot of reworking of guns, and ammo spawns. Yay.

[GuardianDll]

itemgroups looks okay

[tenehea]

SMG spawns were removed as a civilian can't really own an actual SMG, and so it was better split into either pistols or rifles.

This is a very bizarre assertion, unless rifles contains semi-auto pistol caliber rifles.

[Holli-Git]

This is a very bizarre assertion, unless rifles contains semi-auto pistol caliber rifles.

Yeah, they do now. Check the changes I've made

[Night-Pryanik]

This is a big chunk of changes. It's really hard to review it properly (for me, at least). May I suggest splitting this PR into a smaller pieces, for example first one for pistols, second one for shotguns etc?

[Holli-Git]

This is a big chunk of changes. It's really hard to review it properly (for me, at least). May I suggest splitting this PR into a smaller pieces, for example first one for pistols, second one for shotguns etc?

That would've been fairly harder to do, as since they cover a lot of the same areas it would introduce a lot of merge conflicts. Guns are also pretty interconnected, and I'm unsure if I could sort them into pistol/shotgun/rifle. Though I do understand where you're coming from, and am thankful for the review, so in the future I'll split it up into different PRs.

[kevingranade]

Ran my demographics test on top of this to see if anything skewed badly. tl;dr nothing horrific, rifles vs pistols is kind of out of whack, but my source for that is pretty sketchy to start with, and I don't know if this PR did it or if it was already there. I'll rerun on before-this-change to check, but it takes 8h to run so I'll have results tomorrow (EDIT: if I remember to start it, which I did not).

gun type	expected	actual
pistols	0.616f	0.47416f
rifles	0.357f	0.50239f
shotguns	0.027f	0.02345f

calibre	expected	actual
9mm	0.21502f	0.24853f
223	0.1286f	0.10938f
45	0.05041f	0.06391f
380	0.04938f	0.04019f
40	0.04115f	0.04782f
357mag	0.03498f	0.0169f
762	0.03498f	0.02165f
308	0.02572f	0.08697f
38	0.01955f	0.02545f
44	0.01543f	0.00827f
3006	0.00823f	0.01782f
45colt	0.00823f	0.00413f
762R	0.00412f	0.00629f
270win	0.00309f	0.00441f
410shot	0.00309f	0.00101f
300blk	0.00206f	0.0f
10mm	0.00103f	0.00154f
123ln	0.00103f	0.00069f
other	0.11728f	0.06595f

[kevingranade]

Belated rerun from before this PR:

gun type	expected	actual
pistols	0.616f	0.41017f
rifles	0.357f	0.55517f
shotguns	0.027f	0.03466f

In short, it was already skewed this direction, though this PR might have contributed a bit.