Fmacias.TplQueue.Cache.MemCache 0.1.0-preview.1

Fmacias.TplQueue.Cache.MemCache

In-memory cache provider built on top of Fmacias.TplQueue.Cache.Abstract.

See also:

• TplQueue.Adapter root README
• TplQueue.Core cache section
• TplQueue.Usage QueueObserverSignalRDashboard sample
• Fmacias.TplQueue README
• Fmacias.TplQueue.Cache.Abstract README

Repository-wide packaging and strong-name signing rules are documented in the TplQueue.Adapter root README.

Use this package when you want a lightweight in-process cache provider for dehydrating payload-aware jobs, hydrating them later, and validating cache-backed flows without introducing an external storage dependency.

Install

dotnet add package Fmacias.TplQueue.Cache.MemCache --version 0.1.0-preview.1

Contents

• MemCacheFactory creation entry point.
• MemCache in-memory implementation of IMemCache.
• Internal in-memory repository model for cache entries.

Repository build notes

Run from TplQueue.Adapter root:

1. Build module:

dotnet build .\src\Fmacias.TplQueue.Cache.MemCache\Fmacias.TplQueue.Cache.MemCache.csproj

2. Run module tests:

dotnet test .\test\Fmacias.TplQueue.Cache.MemCache.Test\Fmacias.TplQueue.Cache.MemCache.Test.csproj

3. Pack through repo pipeline:

powershell -NoProfile -ExecutionPolicy Bypass -File .\pack-local.ps1

Usage outline

Create with:

• IUniversalDataSerializer
• IDataJobFactory
• ITypeResolver
• IPayloadHandlers
• IRetryPolicyAbstractFactory

Then use cache dehydration/hydration APIs from IMemCache.

Default facade-owned resolver example:

using Fmacias.TplQueue;
using Fmacias.TplQueue.Cache.MemCache;
using Fmacias.TplQueue.Contracts;

IUniversalDataSerializer jsonSerializer = api.SystemTextSerializerFactory().Serializer();
IUniversalDataSerializer xmlSerializer = api.XmlSerializerFactory().Serializer();

IMemCache cache = api.Cache<IMemCache>(
    MemCacheFactory.Create(),
    jsonSerializer);

If payload types must be resolved from a dedicated AppDomain, provide a custom ITypeResolver implementation and pass it through the explicit facade overload. MemCache depends only on the abstraction, not on the concrete runtime resolver.

using Fmacias.TplQueue.Cache.Abstract.Factories;

ITypeResolver typeResolver = RuntimeNodeTypeResolverFactory.Create().Resolver();

IMemCache cache = api.Cache<IMemCache>(
    MemCacheFactory.Create(),
    jsonSerializer,
    typeResolver);

Use the same cache object to dehydrate a payload root, hydrate it back, and dispatch it through a queue:

cache.Dehydrate(payloadRoot, isFifo: false);

ILogger<IParallelQ> queueLogger = loggerFactory.CreateLogger<IParallelQ>();

if (cache.TryHydrateNextJob(out IDataJobRoot hydratedRoot, out ICacheEntry lease))
{
    IParallelQ queue = api.QFactory.Parallel("main", queueLogger);

    queue.Enqueue(hydratedRoot, CancellationToken.None);
    queue.ResumePolling();

    await hydratedRoot.WaitUntilFinishedAsync();
}

Use xmlSerializer instead of jsonSerializer in api.Cache<IMemCache>(...) when XML payload storage is required.

Compatibility note

Some public persisted members still expose JSON-oriented names such as PayloadJson. Those names are retained for compatibility and should be read as serializer-specific payload content. MemCache can store payload content produced by either the JSON or XML IUniversalDataSerializer implementation selected during cache creation.

Runtime type resolution status

Current state:

• MemCache depends only on ITypeResolver, so it can work with the current AppDomain-based runtime resolver

Deferred work:

• if plugin payload types start being loaded from dedicated runtime boundaries in modern .NET, the preferred evolution is an AssemblyLoadContext-aware resolver behind ITypeResolver