lmxopcua/tests/Server/ZB.MOM.WW.OtOpcUa.Runtime.Tests/Drivers/DriverHostActorWriteAskTests.cs

using System.Text.Json;
using Akka.Actor;
using Microsoft.EntityFrameworkCore;
using Shouldly;
using Xunit;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Deploy;
using ZB.MOM.WW.OtOpcUa.Commons.OpcUa;
using ZB.MOM.WW.OtOpcUa.Commons.Types;
using ZB.MOM.WW.OtOpcUa.Configuration;
using ZB.MOM.WW.OtOpcUa.Configuration.Entities;
using ZB.MOM.WW.OtOpcUa.Configuration.Enums;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
using ZB.MOM.WW.OtOpcUa.Runtime.Drivers;
using ZB.MOM.WW.OtOpcUa.Runtime.Tests.Harness;

namespace ZB.MOM.WW.OtOpcUa.Runtime.Tests.Drivers;

/// <summary>
///     lmxopcua#777: <see cref="DriverHostActor"/>'s write Ask is derived from the configured write timeout
///     (<see cref="DriverWriteBounds.HostAsk"/>) rather than a fixed 8 s, so a write the child is still entitled to run
///     is never reported failed by the Ask — the node would be reverted while the device could still take the value.
///     <para>
///         Drives a real apply (a real <see cref="DriverInstanceActor"/> child), then routes a held write with the
///         host and the child on ONE manual clock: the host's Ask and the child's bounds both run on it, so stepping
///         the clock shows which one answers first.
///     </para>
/// </summary>
public sealed class DriverHostActorWriteAskTests : RuntimeActorTestBase
{
    private static readonly NodeId TestNode = NodeId.Parse("driver-write-ask-test");
    private static readonly RevisionHash RevA = RevisionHash.Parse(new string('a', 64));
    private static readonly TimeSpan Timeout = TimeSpan.FromSeconds(5);
    private const string RawNodeId = "ns=2;s=Plant/Modbus/dev1/speed";

    /// <summary>The value the gated driver holds; every other value is written at once.</summary>
    private const double HeldValue = 99.0;

    /// <summary>
    ///     A write still inside a 9 s write timeout is not cut off at 8 s: the host waits for the child, which
    ///     answers with the device's real outcome. (Mutation-checked: a fixed 8 s Ask on the same clock fails this
    ///     with "write timeout" at the 8.5 s step.)
    /// </summary>
    [Fact]
    public void A_write_within_a_9s_write_timeout_is_not_failed_by_the_host_Ask()
    {
        var clock = new ManualTimeProvider();
        var driver = new GatedWritableDriver();
        var host = SpawnHostAndApply(driver, new DriverWriteBounds(TimeSpan.FromSeconds(9)), clock);

        var asker = RouteHeldWrite(host, driver, clock);
        clock.Advance(TimeSpan.FromSeconds(8.5));
        asker.ExpectNoMsg(TimeSpan.FromMilliseconds(300), TestContext.Current.CancellationToken);

        driver.Release();
        var result = asker.ExpectMsg<DriverHostActor.NodeWriteResult>(Timeout, cancellationToken: TestContext.Current.CancellationToken);
        result.Success.ShouldBeTrue(result.Reason);
    }

    /// <summary>
    ///     A token-ignoring write under a 9 s write timeout is answered by the CHILD (abandoned BadTimeout at
    ///     9 s + its grace), not by the host's Ask: the Ask covers the child's whole bound. Under the fixed 8 s Ask
    ///     this read "write timeout" from the Ask at 8 s while the device still had the write.
    /// </summary>
    [Fact]
    public void A_token_ignoring_write_is_answered_by_the_child_before_the_host_Ask()
    {
        var clock = new ManualTimeProvider();
        var driver = new GatedWritableDriver { IgnoreCancellation = true };
        var bounds = new DriverWriteBounds(TimeSpan.FromSeconds(9));
        var host = SpawnHostAndApply(driver, bounds, clock);

        var asker = RouteHeldWrite(host, driver, clock);
        clock.Advance(bounds.Timeout);
        asker.ExpectNoMsg(TimeSpan.FromMilliseconds(300), TestContext.Current.CancellationToken);
        clock.Advance(bounds.AbandonGrace);

        var result = asker.ExpectMsg<DriverHostActor.NodeWriteResult>(Timeout, cancellationToken: TestContext.Current.CancellationToken);
        result.Success.ShouldBeFalse();
        result.Reason!.ShouldContain("800A0000", customMessage: "the child's BadTimeout, not the host Ask's \"write timeout\"");
        result.Reason!.ShouldContain("within 9000 ms");
        bounds.RunBound.ShouldBeLessThan(bounds.HostAsk);
    }

    /// <summary>
    ///     The mechanism the host's Ask relies on: with an Infinite Akka timeout, the Ask's deadline is the write
    ///     clock's token alone — it does not complete a tick before <see cref="DriverWriteBounds.HostAsk"/>, and it
    ///     does complete (cancelled, which the host reports as "write timeout") at it, for a child that never answers.
    /// </summary>
    [Fact]
    public void An_Ask_with_an_Infinite_timeout_ends_exactly_at_its_clock_deadline()
    {
        var clock = new ManualTimeProvider();
        var bounds = new DriverWriteBounds(TimeSpan.FromSeconds(9));
        var silentChild = CreateTestProbe();

        using var deadline = new CancellationTokenSource(bounds.HostAsk, clock);
        var ask = silentChild.Ref.Ask<DriverInstanceActor.WriteAttributeResult>(
            new DriverInstanceActor.WriteAttribute("x", 1), System.Threading.Timeout.InfiniteTimeSpan, deadline.Token);

        clock.Advance(bounds.HostAsk - TimeSpan.FromMilliseconds(1));
        Thread.Sleep(200);
        ask.IsCompleted.ShouldBeFalse("the Ask ended before its clock deadline");

        clock.Advance(TimeSpan.FromMilliseconds(1));
        AwaitCondition(() => ask.IsCompleted, Timeout, TimeSpan.FromMilliseconds(20), cancellationToken: TestContext.Current.CancellationToken);
        ask.IsCompletedSuccessfully.ShouldBeFalse();
    }

    /// <summary>Routes the held write once the host accepts writes (a warm-up write proves the child is Connected
    /// and the reverse map is pushed), then waits until it is parked in the driver with every bound armed.</summary>
    private Akka.TestKit.TestProbe RouteHeldWrite(IActorRef host, GatedWritableDriver driver, ManualTimeProvider clock)
    {
        var lastReason = "(no reply)";
        AwaitAssert(
            () =>
            {
                var warm = CreateTestProbe();
                host.Tell(new DriverHostActor.RouteNodeWrite(RawNodeId, 1.0, AddressSpaceRealm.Raw), warm.Ref);
                var r = warm.ExpectMsg<DriverHostActor.NodeWriteResult>(TimeSpan.FromSeconds(1));
                lastReason = r.Reason ?? "(none)";
                r.Success.ShouldBeTrue($"host kept rejecting the warm-up write; last reason: {lastReason}");
            },
            duration: Timeout,
            interval: TimeSpan.FromMilliseconds(50));

        var asker = CreateTestProbe();
        host.Tell(new DriverHostActor.RouteNodeWrite(RawNodeId, HeldValue, AddressSpaceRealm.Raw), asker.Ref);
        AwaitCondition(() => driver.HeldEntered, Timeout, TimeSpan.FromMilliseconds(20));
        // The host's Ask deadline + the child's token cancel + the child's abandon bound.
        AwaitCondition(() => clock.ArmedTimers == 3, Timeout, TimeSpan.FromMilliseconds(20));
        return asker;
    }

    private IActorRef SpawnHostAndApply(GatedWritableDriver driver, DriverWriteBounds bounds, ManualTimeProvider clock)
    {
        var db = NewInMemoryDbFactory();
        var deploymentId = SeedDeployment(db);
        var coordinator = CreateTestProbe();
        var host = Sys.ActorOf(DriverHostActor.Props(
            db, TestNode, coordinator.Ref,
            driverFactory: new SingleDriverFactory(driver),
            localRoles: new HashSet<string> { "driver" },
            writeBounds: bounds,
            writeClock: clock));

        host.Tell(new DispatchDeployment(deploymentId, RevA, CorrelationId.NewId()));
        coordinator.ExpectMsg<ApplyAck>(Timeout).Outcome.ShouldBe(ApplyAckOutcome.Applied);
        return host;
    }

    /// <summary>One enabled Modbus driver "drv-1" with one ReadWrite tag, RawPath <c>Plant/Modbus/dev1/speed</c>.</summary>
    private static DeploymentId SeedDeployment(IDbContextFactory<OtOpcUaConfigDbContext> db)
    {
        var artifact = JsonSerializer.SerializeToUtf8Bytes(new
        {
            RawFolders = new[] { new { RawFolderId = "rf-plant", ParentRawFolderId = (string?)null, Name = "Plant", ClusterId = "c1" } },
            DriverInstances = new[]
            {
                new { DriverInstanceId = "drv-1", RawFolderId = "rf-plant", Name = "Modbus", DriverType = "Modbus", DriverConfig = "{}", ClusterId = "c1", Enabled = true },
            },
            Devices = new[] { new { DeviceId = "dev-1", DriverInstanceId = "drv-1", Name = "dev1", DeviceConfig = "{}" } },
            TagGroups = Array.Empty<object>(),
            Tags = new[]
            {
                new { TagId = "t-speed", DeviceId = "dev-1", TagGroupId = (string?)null, Name = "speed", DataType = "Double", AccessLevel = 1, TagConfig = "{}" },
            },
            UnsAreas = Array.Empty<object>(),
            UnsLines = Array.Empty<object>(),
            Equipment = Array.Empty<object>(),
            UnsTagReferences = Array.Empty<object>(),
        });

        var id = DeploymentId.NewId();
        using var ctx = db.CreateDbContext();
        ctx.Deployments.Add(new Deployment
        {
            DeploymentId = id.Value,
            RevisionHash = RevA.Value,
            Status = DeploymentStatus.Sealed,
            CreatedBy = "test",
            SealedAtUtc = DateTime.UtcNow,
            ArtifactBlob = artifact,
        });
        ctx.SaveChanges();
        return id;
    }

    private sealed class SingleDriverFactory(GatedWritableDriver driver) : IDriverFactory
    {
        public IDriver? TryCreate(string driverType, string driverInstanceId, string driverConfigJson) =>
            driverType == "Modbus" ? driver.Bind(driverInstanceId, driverType) : null;

        public IReadOnlyCollection<string> SupportedTypes => new[] { "Modbus" };
    }

    /// <summary>Writes <see cref="HeldValue"/> only once released (honouring its token unless
    /// <see cref="IgnoreCancellation"/>); every other value is written at once.</summary>
    private sealed class GatedWritableDriver : IDriver, IWritable
    {
        private readonly TaskCompletionSource _release = new(TaskCreationOptions.RunContinuationsAsynchronously);
        private int _heldEntered;

        public bool IgnoreCancellation { get; init; }
        public bool HeldEntered => Volatile.Read(ref _heldEntered) == 1;
        public string DriverInstanceId { get; private set; } = string.Empty;
        public string DriverType { get; private set; } = string.Empty;

        public GatedWritableDriver Bind(string id, string type)
        {
            DriverInstanceId = id;
            DriverType = type;
            return this;
        }

        public void Release() => _release.TrySetResult();

        public async Task<IReadOnlyList<WriteResult>> WriteAsync(IReadOnlyList<WriteRequest> writes, CancellationToken cancellationToken)
        {
            if (writes.Any(w => w.Value is HeldValue))
            {
                Volatile.Write(ref _heldEntered, 1);
                if (IgnoreCancellation) await _release.Task;
                else await _release.Task.WaitAsync(cancellationToken);
            }

            return writes.Select(_ => new WriteResult(OpcUaStatus.Good)).ToList();
        }

        public Task InitializeAsync(string driverConfigJson, CancellationToken cancellationToken) => Task.CompletedTask;
        public Task ReinitializeAsync(string driverConfigJson, CancellationToken cancellationToken) => Task.CompletedTask;
        public Task ShutdownAsync(CancellationToken cancellationToken)
        {
            _release.TrySetResult();
            return Task.CompletedTask;
        }

        public DriverHealth GetHealth() => new(DriverState.Healthy, DateTime.UtcNow, null);
    }
}