I'm currently working on a library, and ran into some strange inconsistencies in behaviour between my unit test code, and an actual app I'm basing on the code.
See, I'm trying to get a Type struct from the name of the class I want to instantiate, but I get nothing (e.g. Type is there, but name() produces null, is_a() fails, etc.)
However, if I make an instance of the type first, then try to get the Type based on name once more, it works just as I expected (which is probably why my unit tests work)
So I was wondering, when is a type registered by the type system and available via. from_name(...) ?
Does the type system only know it after it has been instantiated at run time? Is there another reason a class name wouldn't be recognized until an instance of the class has been instantiated? Should I use some other method of registration?
I'm coding in Vala if that makes any kind of a difference.
A type is first registered with the type system when its get_type() function is called. This is called in a number of places, basically whenever you need to get the GType for that type. Typically, it will first be called during class_init of the type, which happens during the first instance init.
So essentially you are right when you say “does the type system only know it after it has been instantiated at run time”, because that’s normally what happens. However, you can bring type registration forward by explicitly calling the get_type() function for a type early, and passing it to g_type_ensure(). For example, see what GLib does here:
/* Initialize types from built-in "modules" */
g_type_ensure (g_null_settings_backend_get_type ());
g_type_ensure (g_memory_settings_backend_get_type ());
g_type_ensure (g_keyfile_settings_backend_get_type ());
…
Related
My Elmish.WPF app combines the Elmish.WPF Sample projects ‘NewWindow’ + ‘NewWindow.Core’ which I am adding to a copy of ‘FileDialogsCmdMsg’ + ‘FileDialogsCmdMsg.Core’.
FileDialogsCmdMsg is the basis of this effort and provides the background processing features to which I am adding the ‘NewWindow’ popup window features.
I am stuck on two (2) type-conflicts when WpfProgram.mkProgramWithCmdMsg is called...
...errors on update and bindings calls.
The compiler reports that line has the following errors (see red underlines above)...
...saying that WpfProgram.mkProgramWithCmdMsg expects a unit -> Model but was given a Model where update and bindings are called in the larger type specification.
But here is the type information for the calling function, WpfProgram.mkProgramWithCmdMsg accepting these parameters…
... sorry that is so large.
As you can see, both ‘update’ and ‘bindings’ are producing the Model parameter types expected by WpfProgram.mdProgramWithCmdMsg. And yet we have a unit->Model expected where a Model has been provided type-conflict; why?
Here is more detail on the update type-conflict…
... see how here how here it is claimed a unit->Model is expected when previously you see that update is supposed to expect a Model? Why?
You see the same problem in this detail on the type-conflict with ‘bindings’...
..again showing unit-Model' expected when the correct Model` type parameter is passed.
Here the type information where update is defined…
...and here is the bindings type info…
QUESTION: Why is a "unit->Model expected error reported when the receiving function, WpfProgram.mkProgramWithCmdMsg, and both parameters, update and bindings both respectively provide Model parameters?
Please ask if more detail is needed to get to the bottom of this.
Thanks for your help!
I suspect the problem is with your init function, which (I'm guessing) is a function of type unit -> Model. If this is true, the lambda fun _ -> init, [] binds the concrete type unit -> Model to the 'model type variable. As result, update and bindings also expect the same concrete type, leading to the error message you're seeing.
If I'm correct, you have two choices:
Change init so it is a value of type Model, rather than a function.
Change init so it also returns an empty message list, and then pass init directly to mkProgramWithCmdMsg, rather than passing a lambda.
Note that when you hover over mkProgramWithCmdMsg, the compiler shows you that 'model is indeed bound to unit -> Model, presumably due to the init type mismatch:
I need to implement an abstract class function, which own a an specific data type. But I need inside my logic layer to make the attribute which is going to be passed as a dynamic data type. But when i Pass it to the function, i am sure that its data type will be as needed. So, i type (product.value.pickedImages) as ListOf5) . But it does an Exception.
The Abstract Class Code Is:
Future<Either<FireStoreServerFailures, List<String>>> uploadProductImages(
{required ListOf5<File> images});
The Implementation Code Is:
Future<Option<List<String>>> _uploadImagesToFirestorage() async {
return await productRepo
.uploadProductImages(
images: (product.value.pickedImages) as ListOf5<File>) // Exception
}
The Exception Is:
The argument type 'ListOf5 < dynamic>' can't be assigned to the
parameter type 'ListOf5 < File>'.
You are trying to cast the List from List<dynamic> to List<String>.
Instead, you should cast each item, using something like this:
void main() {
List<dynamic> a = ['qwerty'];
print(List<String>.from(a));
}
Not sure about the implementation of this ListOf5 though...
The cast (product.value.pickedImages) as ListOf5<File> fails.
It fails because product.value.pickedImages is-not-a ListOf5<File>, but instead of ListOf5<dynamic> (which may or may not currently contain only File objects, but that's not what's being checked).
Unlike a language like Java, Dart retains the type arguments at run-time(it doesn't do "erasure"), so a ListOf5<dynamic> which contains only File objects is really different from a ListOf5<File> at run-time.
You need to convert the ListOf5<dynamic> to a ListOf5<File>.
How to do that depends on the type ListOf5, which I don't know.
For a normal List, the two most common options are:
(product.value.pickedImages).cast<File>(). Wraps the existing list and checks on each read that you really do read a File. It throws if you ever read a non-File from the original list. Perfectly fine if you'll only read the list once.
List<File>.of(product.value.pickedImages). Creates a new List<File> containing the values of product.value.pickedImages, and throws if any of the values are not File objects. Requires more memory (because it copies the list), but fails early in case there is a problem, and for small lists, the overhead is unlikely to be significant. If you read the resulting list many times, it'll probably be more efficient overall.
If the ListOf5 class provides similar options, you can use those. If not, you might have to build a new ListOf5 manually, casting each element of the existing ListOf5<dynamic> yourself.
(If the ListOf5 class is your own, you can choose to add such functionality to the class).
I am trying to call a method of a class that I only know by name as a String. Now therefore I would need a ClassMirror of that class that allowes me to instantiate an instance. However, creating ClassMirrors seems to be only possible by entering a type using reflectClass(Type) or by passing an already existing instance of that class into reflect(dynamic). So these aren`t helping if I only have a String.
In Java you can do this pretty easily, by calling Class.forName(String). Then you would get a Constructor instance, make it accessibly and call it.
Does anyone know if this is even possible in dart? What seems weird is that once you have a ClassMirror you can access fields and methods by passing symbols, which can be created by Strings.
You can put a specific list of strings to map to a specific list of closures to create a new object with specific parameters.
But you can't get a reflection without using dart:mirrors, which is being deprecated, and also had a negative impact on tree shaking to get the payload size down.
In general, you're invited to look at the package:reflectable to achieve most of what you'd want out of dart:mirrors, using source-to-source builders.
After seeing the video here, i got confused about the use of final keyword.
Here below is an example image from the video
Here there are two classes Pet as parent class and Dog as child class, and we have function implementation of makeNoise(p) which takes Pet instance as parameter. But behind the scene compiler inserts few more lines to this method for checking class.
Since makeNoise(p) takes parameter as Pet instance, compiler should directly call the property "name" from the Pet class, as the method parameter is for this class.
Why would compiler be worried about the overriding the property in child class, because the parameter is a Pet instance and compiler knows it. isn't it?
My question may looks silly but if someone can explain it more clearly, i would appreciate it.
The parameter of makeNoise(p: Pet) must be a Pet, but it does not to be an immediate instance of Pet itself. Since Dog is a subclass of Pet, it is also a Pet. Therefore, in Swift it is perfectly valid for someone to pass a Dog instance into makeNoise, in which case the overridden version of noise will be called. Also, this dog may have chosen a different value for name.
Dynamic Type
This is called the 'dynamic type' of p. The compile-time type of p is Pet, but the compiler must account for the fact that the run-time type of p might be a Pet subclass that overrides something. This is called the 'dynamic type' of p. In Swift 1 and 2, which were in use at the time of WWDC 2015, use the .dynamicType syntax on an object to determine its runtime type. Swift 3 uses the type(of: object) global function. This is not something you'll need very often, but it demonstrates how the compiler allows subclasses to act as their parent and still override things.
final
final tells the compiler that either the class will not be subclassed, or the property or method will not be overridden. This way it won't need to check for overrides.
Changes in Swift 3
This year's WWDC 2016 session on Swift performance did not mention the final keyword once, if I remember correctly, however. While it is still available in Swift 3 and serves the same purpose, there are new ways to prevent subclassing and overriding. This is via Access Control. Swift 3 introduces the open keyword as distinct from public.
public – The object, property, method, etc. is accessible by anyone inside or outside the module
open – The class or method is not only accessible by anyone, but may also be subclassed (in the case of classes) or overridden (in the case of methods).
I am not sure if this, like final, communicates to the Swift compiler that it does not need to do its extra type checking.
I have my main form. Form_Main
It creates two instances of two classes.
Candle_Data : TCandle_Data;
Indicator_2700 : TIndicator_2700;
In order for Indicator_2700 to properly compute its values it must have access to the candle data in the obect Candle_Data from inside one of its methods. Thus how can Indicator_2700 access data inside Candle_Data? Does Form_Main have to pass it as a argument at Constructor time?
Both Class declarations are in their own unit file.
You could use any of the following (non-exhaustive) methods:
Pass the object reference as a parameter to any methods that need it. Of course you need to get hold of Candle_Data so the suitability of this approach really depends who the caller is.
Pass the Candle_Data object reference to the constructor of the other object and then store it in a private member field.
Make the object reference a public property of the single instance of the main form and access it that way.
We don't really have enough information to advise you which is best but the starting point is always to prefer parameters and local variables over global state.
TIndicator_2700 could have a property to link it to the instance of TCandle_Data that is relevant to its own instance or you should supply it as an argument to the method that needs to access the data.
You could certainly pass the TCandle_Data instance into the constructor of Indicator_2700, and store a reference within the resulting instance until you needed it.
Both class declarations are in their own unit file.
That suggests that both have nothing to do with the other. But still you want one to have knowledge about the other. It sounds like a little design mixup, but that doesn't need to be the case.
There are multiple solutions, here are three of them, each with its own purpose:
Place both classes in the same unit, only if both classes have a common theme/subject (e.g. TCar and TAirplane in the unit Transport),
Use one unit in the other unit, only if both units represent different subjects, but one may depend on the other (e.g. unit Transport uses unit Fuel: TCar needs TDiesel, but TDiesel doesn't need a TCar). This only works one-way. Delphi prevents using in both ways with a compiler error: "Circular unit reference to 'Fuel'". The only workaround is to use the second unit in the implementation section, but that usually is considered a big nono.
Declare a new base-class in a new unit, only if the base-class has a common subject, but the final descendants do not (e.g. TFuel is used by all transportation classes like TCar, TAirplane and THorse, but TFood (a descendant of TFuel) is only used by THorse and TPerson).
As for how to link both classes together, see the already given answers.