How the .NET Runtime Loads a Type
mattwarren.org
mattwarren.org
Eg. An API to a IoT cloud that lets user define their own type. The payload that i fetch is <T> so users can define their own type.
Another one is an api where i take a generic approach, so my only code that i write ( for every "class" / " model" ) is:
//every endpoint does a complete crud + jsonpatch + batch, also, all methods can be overridden to customize code. This is the code i need to write every time :p https://pastebin.com/fp8MtpBu
Note to myselve: extend the implementation to support expressions ( odata ) in the url ( similar to sql on top of an API through changing url parameters) and return the dto as response. :)
public Result WriteValue<T>(T value) where T : INumeric/Numeric { }
https://stackoverflow.com/questions/32664/is-there-a-constra...:(
So instead the interface has an identical method for ints, doubles, floats, decimals, strings...
http://referencesource.microsoft.com/#mscorlib/system/double...
public interface Num<A>
{
A Add(A x, A y);
A Subtract(A x, A y);
A Multiply(A x, A y);
A Divide(A x, A y);
A FromInteger(int value);
}
public struct NumInt : Num<int>
{
public int Add(int x, int y) => x + y;
public int Subtract(int x, int y) => x - y;
public int Multiply(int x, int y) => x * y;
public int Divide(int x, int y) => x / y;
public int FromInteger(int value) => value;
}
public struct NumDouble : Num<double>
{
public double Add(double x, double y) => x + y;
public double Subtract(double x, double y) => x - y;
public double Multiply(double x, double y) => x * y;
public double Divide(double x, double y) => x / y;
public double FromInteger(int value) => (double)value;
}
public struct NumBigInt : Num<BigInteger>
{
public BigInteger Add(BigInteger x, BigInteger y) => x + y;
public BigInteger Subtract(BigInteger x, BigInteger y) => x - y;
public BigInteger Multiply(BigInteger x, BigInteger y) => x * y;
public BigInteger Divide(BigInteger x, BigInteger y) => x / y;
public BigInteger FromInteger(int value) => (BigInteger)value;
}
public static class TestGenericNums
{
public static void Test()
{
var a = DoubleIt<NumInt, int>(10); // 20
var b = SquareIt<NumInt, int>(10); // 100
var c = NegateIt<NumInt, int>(10); // -10
var t = DoubleIt<NumDouble, double>(10); // 20
var u = SquareIt<NumDouble, double>(10); // 100
var v = NegateIt<NumDouble, double>(10); // -10
var x = DoubleIt<NumBigInt, BigInteger>(10); // 20
var y = SquareIt<NumBigInt, BigInteger>(10); // 100
var z = NegateIt<NumBigInt, BigInteger>(10); // -10
}
public static A DoubleIt<NumA, A>(A value) where NumA : struct, Num<A> =>
default(NumA).Add(value, value);
public static A SquareIt<NumA, A>(A value) where NumA : struct, Num<A> =>
default(NumA).Multiply(value, value);
public static A NegateIt<NumA, A>(A value) where NumA : struct, Num<A> =>
default(NumA).Subtract(default(NumA).FromInteger(0), value);
} var x = DoubleIt<NumInt, int>(Int32.MaxValue); // 20
Performs exactly the same as this: var a = Int32.MaxValue;
var b = Int32.MaxValue;
var c = a + b;
DoubleIt calls NumInt.Add which invokes the same code, why wouldn't it work in exactly the same way?Please don't. I've ended up maintaining far too many abominations where developers got excited about generics and linq rather than solving real problems. You just end up with apps that should be simple and straightforward being a nightmare to maintain.
Generics, linq and inheritance all have their uses but if they aren't used sparingly they create more spaghetti.
Edit - at the very least, please be careful and use generics as sparingly as possible.
Its all about stepping back and going: "am I trying too hard?". Like that last 10% of making the code base 100% elegant is what ruins it.
It's usually a good idea to be provide both an generic and a non-generic interface to an object.
A good technique I've found is to have the base object be non-generic and then have a generic version that overrides the properties (with new) that you want to use generic types on. (using as against an underlying field for conversion)
interface IValidator<T> {
bool Validate(T thingToValidate);
}
class ThingValidator : IValidator<Thing> {}
class ThingValidator : IValidator<OtherThing> {}
var validators = GetAllValidators();
What type is `validators`?So it's whatever type of T you pass to GetAllValidators.
And then you can do for example:
var validators = GetAllValidators<EntityBase>();
Where validators might be IEnumerable<IValidator<EntityBase>>. Compile-time type safety is preserved and each individual validator's generic type can be any type that is derived from EntityBase.:)
An example is:
For some reason or another you've written a thread dispatcher that will dynamically spin up threads and route the outputs of these threads into new threads. (e.g. https://github.com/bilus/pipes )
Obviously this library will need to do run-time casting of types already as C# makes it too difficult to orchestrate these threads without some shared type. (Here is an example of Microsoft following this pattern. https://msdn.microsoft.com/en-us/library/dd321424(v=vs.110).... )
Now you need to hook validation into this existing dispatching library by assosciating validators with the outputs of particular threads. Without a shared type you cannot invoke the validator, but it sure is convenient to write the validator using generics.
Another example is:
You have two interfaces, "ISerializable" and "IEquatable", and two validators, "Validator<ISerizable>" and Validator<IEquatable>", and these validators apply to multiple classes. To run all of these validators against a given class you have three (un)reasonable choices:
* Create a boilerplate wrapper class for each of these validators for each class that they apply to, and store them against that class somehow. The disadvantage of this is that you lose track of which validators are shared between classes which may be important for some optimisation.
* Create a non-genic type "IValidator" or "Validator" that all validators share and put these in a collection. (i.e. as mentioned by the grandparent of this post).
* Bite the bullet: The code that calls these validators has to be duplicated. (This is the incorrect choice though because some day you may need to add a spin-up/teardown before running the validators and now you don't have a centralized location to do this)
Another example is:
You are writing validators for types from a 3rd party library. You may not modify the source code of this library, because it's WinForms. You've been tasked with a set of arbitrary validation tasks for your multi-million line codebase. Maybe you need to add a spell checker for all user-editable text? Maybe any control that is bound to a particular property needs an orange background.
Suddenly you have a case where you really do need to do casting and reflection to figure out which validators to run, as you perform the unfortunate evil of a recursive search through the entire control tree to hook up your validation, rather than re-writing a few thousand files.
No I'm not thinking that, I'm saying that implementing a non-generic base interface like
public interface IValidator {
bool Validate(object obj);
}
is pointless because you'll have to do the same type checking/casting in Validate you would have to do in a wrapper which implemented IValidator<object>. Requiring a base interface is worse because you'll have to implement the same boilerplate implementation of Validate(object) in every implementing class - see IEnumerable/IEnumerable<T> as an example.> Obviously this library will need to do run-time casting of types
You've linked to a clojure library which is dynamically typed so I don't see how this is relevant. It's not obvious to me why a C# version would require casting.
Task/Task<T> does follow the pattern but they could have simply introduced a Unit type in the BCL and replaced the non-generic Task type with Task<Unit> instead.
> To run all of these validators against a given class you have three (un)reasonable choices:
You only need to write one wrapper class which encapsulates the cast you'd have to do anyway in a non-generic IValidator interface:
public class Wrapper<T, TBase> : IValidator<TBase>
{
private readonly IValidator<T> inner;
public Wrapper(IValidator<T> v)
{
this.inner = v;
}
public bool Validate(TBase b)
{
if(b is T)
{
return this.inner.Validate((T)(object)b);
}
else return false;
}
}Microsoft could have added some unified Unit type but they did not, which makes this irrelevant.
Just because there is an obvious negative consequence to choosing a particular type schema (e.g. IEnumerable<t> extending IEnumerable) does not mean that the pattern is pointless. Sometimes biting the bullet is necessary since C# has a bullshit type system that makes dynamic orchestration a pain in the arse.
It's not irrelevant - you're using Task/Task<T> as an example of the non-generic base type/generic subtype pattern being necessary, but it isn't.
IEnumerable<T> extending IEnumerable is a consequence of C#1 not supporting generics at all, and the non-generic version would not be necessary if it had. I was just using it as a common example of the boilerplate the pattern incurs.
Honestly if you can that clojure library I linked a few posts ago without using this technique I will be extremeley impressed, but otherwise I disbelieve you. (Note: I have done this a few years ago.)
But it could be `IValidator<object>`, in the worst case. That being said, why not make a `class ThingValidator<T> : IValidator<T> {}`, for all the lack of detail we have?
There aren't enough details here to demonstrate the weaknesses of generics.
Thats the point. You need an `IValidator` interface.
interface IValidator
{
bool Validate(object item);
}
interface IValidator<T> : IValidator
{
bool Validate(T item);
}
abstract class Validator<T>
: IValidator<T> where T : class
{
public abstract bool Validate(T item);
bool IValidator.Validate(object item)
{
var typedItem = item as T;
if (typedItem == null)
throw new InvalidOperationException();
return Validate(typedItem);
}
}But it's not as bad as it sounds and the only way (AFAIK) to be able to work with both polymorphic objects and have the convenience (and type checking) of a generic counterpart. Limiting yourself to just a generic object/interface unfortunately breaks a lot of design patterns. It's also safe afaik since the only thing the generic version does is some casting and it's generic parameter enforces that casting.
but if you have a better way I'm genuinely interested to hear it
> Limiting yourself to just a generic object/interface unfortunately breaks a lot of design patterns.
Which ones and why?
> It's also safe afaik since the only thing the generic version does is some casting and it's generic parameter enforces that casting.
If you don't have a known constraint that is enforced by the type system then it can never be 'safe'.
>If you don't have a known constraint that is enforced by the type system then it can never be 'safe'.
I don't quite know what you mean by this. If you cast to your "T" then casting is ensured. I guess there's some margin for error in that you must enter your casting code to be correct, but it's pretty hard to mess that part up, granted not as good as enforced genericy. Note that in many instances you only initialize generic classes, but you needed to be extract a non-generic interface from then to pass to other classes.
I'd like to see an example of what you mean. You can absolutely pass generic values to other classes. If the type you're passing the generic value to is not generic then the method should be. If you are casting then 99 times out of 100 you have a problem with your code because the types aren't compatible. i.e.
static void DoSomethingGeneric<A>(A value) => ...
static void DoSomethingSpecific(string value) => ...
IEnumerable<int> array = new [] { 1, 2, 3, 4, 5 };
DoSomethingGeneric(array.First()); // Works
DoSomethingSpecific(array.First()); // Can only ever work if array is a string[]
> If you cast to your "T" then casting is ensured.If you 'get out' of the generic situation by casting to an object or dynamic, then at some point you're going to have to cast back to a concrete type. That is the point where your code will blow up. You're carrying ticking time bombs as values.
The point is that in some cases you want to be able to pass a more general interface while getting some Type-checking and convenience of generics.
In this case we want to hold on to PopProtocolSender and perhaps use it as a variable somewhere. When we use it we need to acccess it's specific IPopProtocol Protocol Property. But we also want to be able to treat it as a general IProtocolSender that has a IProtocol Property, so it can be sent to other classes that takes that interface.
These kinds of situations mainly arise when you have a more complicated architecture. If there's a better way to have your cake and eat it too I'd be glad to know it :)
public interface IPopProtocol : IProtocol
{
void SomeUniqueMetod();
}
public interface IProtocol
{
}
public interface IProtocolManager
{
void SendMessage(IProtocol protocol, string message);
}
public interface IProtocolSender
{
IProtocol Protocol
{
get;
}
}
public class PopProtocol : IPopProtocol
{
public void SomeUniqueMetod()
{
}
}
public class Program
{
public void DoStuff()
{
var sender = new PopProtocolSender ();
var manager = new ProtocolManager();
manager.SendMessage(sender.Protocol, "Testing");
sender.Protocol.SomeUniqueMethod();
}
}
public class ProtocolManager : IProtocolManager
{
public void SendMessage(IProtocol protocol, string message)
{
}
}
public class ProtocolSender : IProtocolSender
{
protected readonly IProtocol _protocol;
public ProtocolSender(IProtocol protocol)
{
_protocol = protocol;
}
public IProtocol Protool => _protocol;
}
public class PopProtocolSender : ProtocolSender<IPopProtocol>
{
public PopProtocolSender() : base(new PopProtocol)
{
}
}
public class ProtocolSender<T> : ProtocolSender where T : IProtocol
{
public ProtocolSender(T protocol) : base(protocol)
{
}
public new T Protocol => (T)_protocol;
} public interface IPopProtocol : IProtocol
{
void SomeUniqueMethod();
}
public interface IProtocol
{
}
public interface IProtocolManager
{
void SendMessage(IProtocol protocol, string message);
}
public interface IProtocolSender<out T> where T : IProtocol
{
T Protocol
{
get;
}
}
public class PopProtocol : IPopProtocol
{
public void SomeUniqueMethod()
{
Console.WriteLine($"Hi, I'm {nameof(PopProtocol)}");
}
}
void Main()
{
var sender = new ProtocolSender<IPopProtocol>(new PopProtocol());
var manager = new ProtocolManager();
manager.SendMessage(sender.Protocol, "Testing");
sender.Protocol.SomeUniqueMethod();
}
public class ProtocolManager : IProtocolManager
{
public void SendMessage(IProtocol protocol, string message)
{
}
}
public class ProtocolSender<T> : IProtocolSender<T> where T : IProtocol
{
private T _protocol;
public T Protocol => _protocol;
public ProtocolSender(T protocol)
{
_protocol = protocol;
}
}
Not sure what the fuss is: you can use SendMessage as usual, and also have access to SomeUniqueMethod.EDIT:
Also please note that the signature for the generic IProtocolSender uses the out keyword, making it covariant. This will allow you to cast IProtocolSender<IPopProtocol> to IProtocolSender<IProtocol> if you need to, or to pass an IProtocolSender<IPopProtocol> object to a method that expects an IProtocolSender<IProtocol>.
The idea that the propagation of type safety is an issue is something I find bewildering, sorry.
Say that we have some kind of datastorage where everything is an IEntity, and we need to be able to treat them as such. Them we have a specific entity IApple : IEntity. Every IEntity has IVersion as a Version property, that is passed in it's implementation's construction. IApple has a more specific IAppleVersion as it's Version Property. When we're working with an IApple instance we want to be able to have access to the version as IAppleVersion. But other classes handles only IEntities with IVersions. In this case it makes sense to have the Version as a generic parameter to the class:
Apple : Entity<IAppleVersion>
which makes sure its constructor is:
Apple(IAppleVersion version).
We need to put the stuff that's shared between all entities in the non-generic Entity class, add the Typing of version in a generic version of Entity so Apple can expose the type provided. The Apple class that implements IApple has specific methods for its type. Both entity classes are abstract so they will never be initiated.
I'm not sure that's clear but I hope you can see the problem that I'm trying to solve. I think it's the disconnect with wanting to be able to access both more specific and more general interfaces (and their properties) that causes trouble.
On the flip side, if we have a legitimate need for both a more specific and a more general version of an class/interface, can you tell me a specific reason the way I showed would be bad. It's easy to get hung up "new" is bad because it's 95% true, but as with everything there are exceptions.
But if you never have the need for it, don't use it. I stand by my opinion though that you should be careful and in some situations be able to provide non-generic interfaces rather than requiring each method and class in a library to be generic, but that's only my opinion man :)
No it really doesn't. It makes sense to use inheritance properly:
public interface IVersion
{
string GetVersion();
}
public class Version : IVersion
{
string version;
public Version(string version)
{
this.version = version;
}
public string GetVersion() =>
version;
}
public class AppleVersion : IVersion
{
string version;
public AppleVersion(string version)
{
this.version = version;
}
public string GetVersion() =>
"apple-" + version;
}
public abstract class Entity
{
public IVersion Version { get; private set; }
public Entity(IVersion version) =>
Version = version;
}
public abstract class Apple : Entity
{
public readonly AppleVersion AppleVersion;
public Apple(AppleVersion version) : base(version)
{
AppleVersion = version;
}
}
That allows for "When we're working with an IApple instance we want to be able to have access to the version as IAppleVersion", as well as general access to the IVersion from an Entity.If you don't like the storage overhead, you could just downcast knowing that the Version is an AppleVersion:
public abstract class Apple : Entity
{
public AppleVersion AppleVersion => Version as AppleVersion;
public Apple(AppleVersion version) : base(version)
{
}
}
> I'm not sure that's clear but I hope you can see the problem that I'm trying to solve. I think it's the disconnect with wanting to be able to access both more specific and more general interfaces (and their properties) that causes trouble.You're describing issues with inheritance, not generics.
> I stand by my opinion though that you should be careful and in some situations be able to provide non-generic interfaces rather than requiring each method and class in a library to be generic, but that's only my opinion man :)
That's fine, you're entitled to your opinion, and if it works for you that's great. But your original comment was very 'matter of fact'. I've not seen anything that has communicated to me an issue with generics; and every example so far should have been implemented differently to the way you state (in my opinion).
a) You now have an AppleVersion Property as well as Version Property, that feels potentially confusing for any user of that class. I now have to find the differently named Version property every time I use an entity?
b) You have to manually write that exact boilerplate code on each specific case of Entity, and there might be a lot more if it isn't a simple example. In my case you get it for free through the generics.
The automatic casting doesn't strike me as a problem, eliminates boilerplate code and the where statement ensures type safety. I've yet to see a single concrete argument against my approach.
I didn't mean it as my way or the highway, I merely wanted to issue a warning on creeping dependencies that generics might cause if you don't have non-generic interfaces, the rest I guess is a matter of taste.
It's a specialised type, so it has specialised behaviour. That's the point of inheritance. You can't put the specialised behaviour in the base because then it's not a specialised type and doesn't understand the context. This is an argument against inheritance.
> b) You have to manually write that exact boilerplate code on each specific case of Entity, and there might be a lot more if it isn't a simple example. In my case you get it for free through the generics.
Yes, every type that has specialised behaviour needs to implement it. And no, you don't get it for free in your version, because you can't have a collection of entities.
IEnumerable<Entity<Version>>
Or, IEnumerable<Entity<AppleVersion>>
All of your complaints about generics appear to be because you're using them to add specialisation to non-specialised types. That will always cause headaches, and that's why, I assume, you have the opinion you do on generics.And? Why is this a bad thing? Are you trying to maintain compatibility with .NET 1.0 for some reason?
> And some polymorphism/patterns simply won't be possible. Trust me, I've encountered this problem several times
Which patterns? Please demonstrate the problem.
is the horrible bit. You should almost NEVER override with "new".
So you're saying that unless you load types step-by-step, you can't have fully reified generics types, because of circular references/dependencies, is that the reason?
> Having implemented a large subset of this stuff for my own .NET runtime
Interesting, which runtime was this? Is it the source available anywhere, I'd love to have a look if it was?