为什么 Assembly.GetExecutingAssembly() 在 NuGet 包中返回不同的结果？

Question

我是创建 NuGet 包的新手，我在各种环境中运行了以下代码片段：

/// <summary>
/// Tries to find type by name in the executing assembly and after that
/// in referenced assemblies.
/// </summary>
/// <param name="typeName">Name of the type to find (can be full or assembly qualified name as well).</param>
/// <returns>Type found using the given name (or null if not found).</returns>
public static Type FindType(string typeName)
{
    if (typeName == null) throw new ArgumentNullException(nameof(typeName));

    // Helper method for finding the type in an assembly
    Type Finder(Assembly ass) => ass?.GetTypes().FirstOrDefault(type =>
        typeName.In(type.Name, type.FullName, type.AssemblyQualifiedName)
    );

    // Get the current assembly
    var executingAssembly = Assembly.GetExecutingAssembly();

    // Check if the type is inside the current assembly
    var targetType = Finder(executingAssembly);

    // Go through all of the referenced assemblies
    foreach (var assName in executingAssembly.GetReferencedAssemblies())
    {
        // If the type was found, return it
        if (targetType != null)
            return targetType;

        // Check if the type is inside the assembly
        targetType = Finder(Assembly.Load(assName));
    }

    return null; // Type wasn't found, return null
}

如果我将其作为本地函数或通过引用的项目运行，它可以正常工作，但是当我创建 NuGet 包并使用 NuGet 包内的方法的实现来调用该方法时，它会返回 null。

方法

Assembly.GetExecutingAssembly

声称它返回

The assembly that contains the code that is currently executing

但从 NuGet 包运行它时我得到了不同的结果。

如果将方法打包到 NuGet 包中，我该怎么做才能从方法中获得正确的输出？

Answer 1

不要依赖于

GetCallingAssembly()

，行为可能会有所不同 取决于 JIT 编译器

要解决您的问题，您可以使用：

Assembly.GetEntryAssembly()