ModernUO/Projects/UOContent.Tests/Tests/Engines/Pathing/BitmapAStarAlgorithmTests.cs
Kamron Batman 6a3804addc
feat: Replace FastAStarAlgorithm with BitmapAStarAlgorithm (#2446)
## Summary

Replaces `FastAStarAlgorithm` with `BitmapAStarAlgorithm`: one cache lookup per cell expansion (8-direction mask + per-direction destination Z) instead of 8 separate `MovementImpl.CheckMovement` calls. Adds the supporting cache infrastructure to back it.

Public API unchanged — `MovementPath` / `Mobile.Move` / `CalcMoves.Find` return the same shapes; the algorithm swap is internal.

## What's in this PR

- **`BitmapAStarAlgorithm`** — A* that issues one `StepCache.TryGetMask` call per cell expansion. Inline fallthrough to the per-cell slow path for multi-Z, off-map, source-Z mismatch, and non-default walkers.
- **`StepCache`** — singleton chunk store keyed by `(mapId, chunkX, chunkY)`. Lazily built on first query, invalidated by `Sector.MultisVersion` mismatch, memory-bounded by sampled probabilistic LRU.
- **`StepProbe`** — computes static-only walkability for a single cell, mirroring `MovementImpl.Check` minus the item / mobile collision phases.
- **`StepMask` / `StepChunk`** — value / storage types for the per-cell results.
- **`CacheEvictionTimer`** — periodic cap backstop (60s interval; early-returns when not over cap).
- **`Map.Sector.MultisVersion`** promoted to `public` so the cache can detect dynamic-static invalidations cheaply.

## Eviction strategy

Sampled probabilistic LRU (Redis-style). Per eviction, sample 5 random keys from a parallel `List<long>` kept in lockstep with the chunk dictionary; evict the oldest of the sample via swap-and-pop. O(1) per eviction regardless of resident count, so sustained cap pressure has no perpetual perf hit.

## Capability handling (interim)

Non-default walkers (non-GM players, creatures with `CanSwim` / `CanFly` / `CanOpenDoors` / `CanMoveOverObstacles`) route entirely through the per-cell slow path via `BitmapAStarAlgorithm.GetSuccessorsSlowPath`. The 2-pass design (cache + capability overlay + dynamic-obstacle pass) lands in the follow-up PR.
2026-05-05 21:53:43 -07:00

109 lines
4 KiB
C#

using Server.Engines.Pathing.Cache;
using Server.Mobiles;
using Server.PathAlgorithms.BitmapAStar;
using Xunit;
using Xunit.Abstractions;
namespace Server.Tests.Pathfinding;
/// <summary>
/// Smoke tests for <see cref="BitmapAStarAlgorithm"/>'s two branches: cache-direct fast
/// path (default walkers) and per-cell slow path (capability creatures and non-GM players).
/// Exercised end-to-end against the real Trammel TileMatrix to ensure neither regresses
/// to "no path found" on reachable goals.
/// </summary>
[Collection("Sequential Pathfinding Tests")]
public class BitmapAStarAlgorithmTests
{
private readonly ITestOutputHelper _output;
public BitmapAStarAlgorithmTests(ITestOutputHelper output)
{
_output = output;
}
[Theory]
// Pinned cell (1500, 1600, z=10): mask=0xC1 → N, W, NW walkable.
// Use start.Z for goal.Z so destNode lands in the same Z plane the algorithm reaches
// during expansion (GetAverageZ at the goal cell may differ from the engine's
// runtime-computed standing Z, which would break the destNode equality check).
[InlineData(1500, 1600, 1498, 1598)] // NW, 2 cells diagonal
[InlineData(1500, 1600, 1497, 1599)] // NW-ish, 3 W + 1 N
public void DefaultWalker_FindsPath_ViaCacheFastPath(int sx, int sy, int gx, int gy)
{
StepCache.Instance.Clear();
var map = Map.Maps[1];
Assert.NotNull(map);
var stub = new DefaultWalkerStub();
map.GetAverageZ(sx, sy, out _, out var startZ, out _);
var start = new Point3D(sx, sy, (sbyte)startZ);
var goal = new Point3D(gx, gy, (sbyte)startZ);
stub.MoveToWorld(start, map);
var result = BitmapAStarAlgorithm.Instance.Find(stub, map, start, goal);
stub.Delete();
Assert.NotNull(result);
Assert.NotEmpty(result);
_output.WriteLine($"default-walker ({sx},{sy})->({gx},{gy}): {result.Length} steps");
}
[Theory]
[InlineData(1500, 1600, 1498, 1598)] // NW, 2 cells diagonal
[InlineData(1500, 1600, 1497, 1599)] // NW-ish, 3 W + 1 N
public void CapabilityCreature_FindsPath_ViaInlineSlowPath(int sx, int sy, int gx, int gy)
{
StepCache.Instance.Clear();
var map = Map.Maps[1];
Assert.NotNull(map);
var stub = new SwimmingStub(World.NewMobile);
stub.DefaultMobileInit();
stub.CanSwim = true; // forces non-default-walker → inline GetSuccessorsSlowPath
map.GetAverageZ(sx, sy, out _, out var startZ, out _);
var start = new Point3D(sx, sy, (sbyte)startZ);
var goal = new Point3D(gx, gy, (sbyte)startZ);
stub.MoveToWorld(start, map);
var result = BitmapAStarAlgorithm.Instance.Find(stub, map, start, goal);
stub.Delete();
// Capability creature: assert reachability — exact length depends on terrain and
// tie-breaking, but a swimmer should always reach a goal a default walker reaches
// on dry land (CanSwim is permissive, never restrictive).
Assert.NotNull(result);
Assert.NotEmpty(result);
_output.WriteLine($"swimmer ({sx},{sy})->({gx},{gy}): {result.Length} steps");
}
/// <summary>
/// Plain Mobile — IsDefaultWalker returns true, the bitmap algorithm uses the cache
/// fast path on every expansion.
/// </summary>
private sealed class DefaultWalkerStub : Mobile
{
public DefaultWalkerStub()
{
Body = 0xC9;
}
}
/// <summary>
/// BaseCreature with CanSwim=true — IsDefaultWalker returns false, the bitmap
/// algorithm short-circuits GetSuccessors to GetSuccessorsSlowPath on every cell.
/// Use the Serial constructor (deserialization path) to bypass NPCSpeeds init,
/// which requires the npc-speeds.json table loaded — not available in tests.
/// </summary>
private sealed class SwimmingStub : BaseCreature
{
public SwimmingStub(Serial serial) : base(serial)
{
Body = 0xC9;
}
}
}