3.1 Framework: behaviors built on draws

Chapter 3: Framework helpers · Section 1 · Guide map

RandEXom.Framework is the task layer. A framework class accepts or creates an IRandomR source, then adds rules around its draws: decide whether a probability succeeds, reorder or pick items, or constrain a number. It is not a replacement for RandomLib. If all you need is an integer in [min, max), call IRandomR.NextInt directly; SlotR is a wrapper around that operation.

Framework.Boolean.TruePercentageR ── compare a bounded draw to a probability
Framework.Item.Shuffles_FisherYates ── use draws to swap list positions
Framework.Item.GachaR<T> / GachaRBatched<T> ── use draws to remove items from a pool
Framework.Item.DistributedTreeR<T> ── choose branches, then keep or remove an item
Framework.Number.SlotR / PongR / DistanceR ── apply rules to numeric draws
                 │
                 └── each uses IRandomR ← RandomLib ← ISeedR ← SeedLib

3.1.1 Choose a behavior

You want	Type	Behavior and state
A true/false event with a specified chance	TruePercentageR	Independent probability on each call; no quota
A rearranged existing list	Shuffles_FisherYates	Mutates the list in place
A weighted bag that can refill automatically	GachaR<T>	Stores individual copies, removes one per pull
A weighted bag with large counts	GachaRBatched<T>	Stores counts, removes one per pull, manual refill
Items grouped at tree endpoints	DistributedTreeR<T>	Chooses branches randomly; optionally removes items and locks empty endpoints
One draw for each slot-machine reel	SlotR	Call Next() once per reel; delegates each draw to its random source
Alternating low/high halves of a range	PongR	Maintains alternation state
Numbers separated from the previous result	DistanceR	Remembers the previous draw

PerlinNoise remains an empty DEBUG-only placeholder; see research context.

3.1.2 Default versus supplied sources

Most helpers offer a default constructor, a numeric-seed constructor, and a constructor accepting IRandomR. The first two create a NetRandom internally. The third gives you control over the source and its seed algorithm:

using RandEXom.Framework.Number;
using RandEXom.RandomLib;

var source = new ModuloRandom(seed: 42);
var spaced = new DistanceR(source, distance: 3);
int next = spaced.Next(0, 10);

If two helpers receive the same source object, their draws interleave and influence each other's later results. If they each construct a source from the same numeric seed, they have separate state. Framework classes may also maintain their own state—such as a bag's remaining items or DistanceR's last result—which is distinct from random-source state. A helper's Reset() or Refill() does not generally rewind its source.

For operations and examples, continue to Boolean, Item, or Number. The class index maps every type, and how the library fits together explains the layer underneath.

---

← Previous: 2.2.4 Choosing a source · Guide map · Next: 3.2 Boolean →