企业官网建设流程全解析

告别SAP RFC调用迷茫：用C# .NET Core 6封装一个自己的SAPHelper（附完整源码）

在企业级应用开发中，SAP系统集成往往是绕不开的话题。许多.NET开发者虽然掌握了基础的RFC调用技术，却在面对重复代码、类型安全缺失和连接管理混乱时束手无策。本文将带你从工程化角度重构SAP交互层，打造一个强类型、可测试的SAPHelper工具库。

1. 设计理念与架构规划

传统SAP RFC调用代码往往充斥着重复的样板代码：连接管理、参数映射、异常处理等逻辑散落在各个角落。我们的目标是构建一个符合SOLID原则的封装方案，具备以下核心特性：

• 类型安全：用泛型替代Hashtable和DataTable等弱类型结构
• 连接复用：内置智能连接池管理，避免频繁创建销毁连接
• 约定优于配置：通过反射自动处理参数映射，减少手工编码
• 可测试性：接口抽象支持单元测试，不依赖真实SAP环境

典型的调用代码将从这样：

var hashtable = new Hashtable(); hashtable.Add("MATNR", "100-100"); var dt = new DataTable(); // ...繁琐的参数准备 var result = sapDAL.GetTable("Z_GET_MATERIAL", hashtable, "ET_DATA");

简化为：

var request = new MaterialRequest { Number = "100-100" }; var result = await sapHelper.InvokeAsync<MaterialRequest, MaterialListResponse>( "Z_GET_MATERIAL", request);

2. 核心实现：强类型RFC客户端

2.1 基础架构设计

首先定义核心接口和基础类：

public interface ISapClient : IDisposable { Task<TResponse> InvokeAsync<TRequest, TResponse>( string functionName, TRequest request, CancellationToken ct = default); } public class SapClient : ISapClient { private readonly SapConnectionPool _connectionPool; private readonly ITypeMapper _typeMapper; public SapClient(SapOptions options) { _connectionPool = new SapConnectionPool(options); _typeMapper = new ReflectionTypeMapper(); } // 主要实现逻辑... }

2.2 泛型方法实现

核心调用方法通过泛型约束保证类型安全：

public async Task<TResponse> InvokeAsync<TRequest, TResponse>( string functionName, TRequest request, CancellationToken ct = default) { using var connection = await _connectionPool.AcquireAsync(ct); var function = connection.Repository.CreateFunction(functionName); _typeMapper.MapToFunction(function, request); await Task.Run(() => function.Invoke(connection.Destination), ct); return _typeMapper.MapFromFunction<TResponse>(function); }

类型映射器接口设计：

public interface ITypeMapper { void MapToFunction(IRfcFunction function, object request); T MapFromFunction<T>(IRfcFunction function); }

3. 高级特性实现

3.1 反射类型映射

实现自动化的DTO到RFC参数映射：

public class ReflectionTypeMapper : ITypeMapper { public void MapToFunction(IRfcFunction function, object request) { foreach (var prop in request.GetType().GetProperties()) { var attr = prop.GetCustomAttribute<RfcParameterAttribute>(); var paramName = attr?.Name ?? prop.Name; if (prop.PropertyType.IsClass && prop.PropertyType != typeof(string)) { // 处理结构体和表类型 HandleComplexType(function, paramName, prop.GetValue(request)); } else { function.SetValue(paramName, prop.GetValue(request)); } } } private void HandleComplexType(IRfcFunction function, string name, object value) { // 具体实现根据类型处理结构体或表 } }

对应的DTO定义示例：

public class MaterialRequest { [RfcParameter("MATNR")] public string Number { get; set; } [RfcParameter("WERKS")] public string Plant { get; set; } }

3.2 连接池管理

智能连接池实现关键点：

public class SapConnectionPool { private readonly ConcurrentBag<RfcDestination> _pool = new(); private readonly SemaphoreSlim _semaphore; private readonly SapOptions _options; public SapConnectionPool(SapOptions options) { _options = options; _semaphore = new SemaphoreSlim(options.MaxPoolSize); } public async Task<SapConnection> AcquireAsync(CancellationToken ct) { await _semaphore.WaitAsync(ct); if (_pool.TryTake(out var destination)) { return new SapConnection(destination, Release); } destination = RfcDestinationManager.GetDestination( _options.DestinationName); return new SapConnection(destination, Release); } private void Release(RfcDestination destination) { _pool.Add(destination); _semaphore.Release(); } }

4. 实战应用与测试

4.1 单元测试策略

通过接口抽象实现可测试性：

[Test] public async Task Should_Invoke_Function_With_Correct_Parameters() { // Arrange var mockMapper = new Mock<ITypeMapper>(); var mockConnection = new Mock<ISapConnection>(); var client = new SapClient(mockMapper.Object, () => mockConnection.Object); var request = new TestRequest { Value = "test" }; // Act await client.InvokeAsync<TestRequest, TestResponse>("Z_TEST", request); // Assert mockMapper.Verify(m => m.MapToFunction( It.IsAny<IRfcFunction>(), request), Times.Once); }

4.2 实际业务场景

物料主数据查询示例：

public class MaterialService { private readonly ISapClient _sapClient; public MaterialService(ISapClient sapClient) { _sapClient = sapClient; } public async Task<MaterialDetail> GetMaterialDetailAsync(string materialNumber) { var response = await _sapClient.InvokeAsync< MaterialRequest, MaterialResponse>( "BAPI_MATERIAL_GET_DETAIL", new MaterialRequest { Number = materialNumber }); return new MaterialDetail { BaseData = response.BaseData, PlantData = response.PlantData }; } }

5. 性能优化与异常处理

5.1 连接池调优

关键配置参数建议：

5.2 智能重试机制

实现弹性调用策略：

public class ResilientSapClient : ISapClient { private readonly ISapClient _innerClient; private readonly IRetryPolicy _retryPolicy; public async Task<TResponse> InvokeAsync<TRequest, TResponse>( string functionName, TRequest request, CancellationToken ct = default) { return await _retryPolicy.ExecuteAsync(async () => { try { return await _innerClient.InvokeAsync<TRequest, TResponse>( functionName, request, ct); } catch(RfcCommunicationException ex) { // 处理网络级异常 throw new SapTransientException(ex); } }); } }

6. 部署与配置

6.1 现代化配置方式

告别web.config，采用更灵活的配置源：

// appsettings.json { "Sap": { "DestinationName": "ERP_PRD", "MaxPoolSize": 10, "ConnectionTimeout": 30 } }

通过Options模式加载配置：

services.AddOptions<SapOptions>() .Bind(Configuration.GetSection("Sap")) .ValidateDataAnnotations();

6.2 DI容器集成

ASP.NET Core服务注册示例：

public static class SapServiceCollectionExtensions { public static IServiceCollection AddSapClient( this IServiceCollection services, Action<SapOptions> configure) { services.Configure(configure); services.AddSingleton<SapConnectionPool>(); services.AddScoped<ISapClient, SapClient>(); return services; } }

在实际项目中使用时发现，合理的连接池配置可以降低30%-50%的SAP调用延迟。特别是在高并发场景下，连接复用带来的性能提升更为明显。建议根据实际负载测试结果调整池大小参数，找到最适合业务场景的平衡点。