With the following code how do I write the custom template deduction correctly?
template<class R, class State, class... Ts> struct visitor : Ts... { using Ts::operator()...; };
template<class R, class State, class... Ts> visitor(State, Ts...)->visitor<class R, State,Ts...>;
using Event = std::variant<HeartBeat, ConfigurationRead>;
using State = std::variant<Idle, Starting, Running, Error>;
void case3()
{
Event e;
State currentState;
State newState = std::visit( visitor{
[](Idle&& state, HeartBeat event) {std::cout << "Currently stopped, heartbeat received.\n"; return Error{}; }
}, currentState, e);
}
I have looked at several examples, but I cannot find one which uses the return of the std::visit.
Your visitor does not need to deduce and encode the return type R -- one will be inferred by std::visit itself: deduced from the callable or fixed to a specified template argument (see below). That being said, it could be simplified to:
template <typename... Ts> struct visitor : Ts... { using Ts::operator()...; };
template <typename... Ts> visitor(Ts...) -> visitor<Ts...>;
In c++17 each visitor must, however, return the same type. You could avoid any type mismatch by, e.g., explicitly specifying one for each lambda expression:
State newState = std::visit( visitor{
[](Idle state, HeartBeat event) -> State { return Error{}; },
// ~~~~^
[](auto state, auto event) -> State { return state; }
// ~~~~^
}, currentState, e);
DEMO
In c++20, the return type can be specified as the first template argument of std::visit:
State newState = std::visit<State>( visitor{
// ~~~~^
[](Idle state, HeartBeat event) { return Error{}; },
[](auto state, auto event) { return state; }
}, currentState, e);
Related
I do not understand why in the snippet below commented line does not infer the state as SubState<dynamic>.
class Commander<T> {
final State<T> state;
Commander(this.state);
void doSomething() {
final state = this.state;
if (state is SubState) {
// print(state.property); // Does not work.
}
}
void doOther() {
final state = this.state;
if (state is SubState<T>) {
print(state.property); // This works.
}
}
}
abstract class State<T> {}
class SubState<T> extends State<T> {
T property;
SubState(this.property);
}
You can only promote a variable to a subtype of its declared type.
The (inferred) declared type of the local state variable is State<T>.
You then does an is test with SubState, aka. SubState<dynamic>.
Since SubState<dynamic> is not a subtype of State<T>, it doesn't promote.
Doing is SubState<T> is checking and promoting to a subtype.
(And as #jamesdlin points out, this is a known short-coming of type inference, where it would probably be better if the raw SubState type didn't automatically mean SubState<dynamic>.)
This is what I want to implement:
void fun({
bool Function(int i) predicate = (i) => false,
}) {
// do something with 'predicate(something)'
}
But I am getting the error:
The default value of an optional parameter must be constant.dart(non_constant_default_value).
I was able to get arround this error with the following:
bool falsePredicate(int i) => false;
void fun({
bool Function(int i) predicate = falsePredicate,
}) {
// do something with 'predicate(something)'
}
But now the question becomes, why can't I directly create a default function value as in the first set of code? There seems to be no difference between the first and the second cases. How is the function given in the first approach not constant?
As #Noah has pointed to the git discussion, the dart language has this missing piece of compile-time constant functions, which eventually leads to this problem.
Check this post: https://github.com/dart-lang/language/issues/1048
As the post shows, the issue has been raised in mid-2012 and now it's been 8+ years. So the hopes of this being available in the near feature is very less.
However few alternative solutions till then:
Option 1 (separate method):
class ServiceProvider {
static bool falsePredicate(int i) => false;
void fun({
bool Function(int i) predicate = falsePredicate,
}) {
// do something with 'predicate(something)'
}
}
Option 2 (Null checking while using the predicate)
class ServiceProvider {
void fun({
bool Function(int i)? predicate,
}) {
int val = 55; // for demonstration
predicate?.call(val); // Call only if the predicate is non-null
}
}
Option 3 (Only for class constructors)
class ServiceProvider {
final bool Function(int i) _predicate;
ServiceProvider ({bool Function(int i)? predicate})
: _predicate = predicate ?? ((i) => false);
void fun() {
int val = 55;
_predicate(5); // No null check is needed. The predicate is always non-null
}
}
I trying make the following code but T only can be int, double or a custom class. I couldn't find how to restrict the type in Dart or something that work like where from C#. How can I do that in Dart?
class Array3dG<T> extends ListBase<T> {
List<T> l = List<T>();
Array3dG(List<T> list) {
l = list;
}
set length(int newLength) { l.length = newLength; }
int get length => l.length;
T operator [](int index) => l[index];
void operator []=(int index, T value) { l[index] = value; }
}
There is no way to constrain the type variable at compile-time. You can only have one bound on a type variable, and the only bound satisfying both int and your custom class is Object.
As suggested by #Mattia, you can check at run-time and throw in the constructor if the type parameter is not one of the ones you supprt:
Array3dG(this.list) {
if (this is! Array3dG<int> &&
this is! Array3dG<double> &&
this is! Array3dG<MyClass>) {
throw ArgumentError('Unsupported element type $T');
}
}
This prevents creating an instance of something wrong, but doesn't catch it at compile-time.
Another option is to have factory methods instead of constructors:
class Array3dG<T> {
List<T> list;
Array3dG._(this.list);
static Array3dG<int> fromInt(List<int> list) => Array3dG<int>._(list);
static Array3dG<int> fromDouble(List<double> list) => Array3dG<double>._(list);
static Array3dG<MyClass> fromMyClass(List<MyClass> list) => Array3dG<MyClass>._(list);
...
}
which you then use as Array3dG.fromInt(listOfInt). It looks like a named constructor, but it is just a static factory method (so no using new in front).
You can check at runtime the type with the is keyword:
Array3dG(List<T> list) {
if (list is List<int>) {
//Handle int
}
else if (list is List<double>) {
//Handle double
}
else if (list is List<MyClass>) {
//Handle MyClass
}
else {
throw ArgumentError('Unsupported $T type');
}
}
Note that if you are handling int and double in the same way you can just check for num
You can check the progress of the Union types here: https://github.com/dart-lang/sdk/issues/4938
I am using Xtext 2.15 to generate a language that, among other things, processes asynchronous calls in a way they look synchronous.
For instance, the following code in my language:
int a = 1;
int b = 2;
boolean sleepSuccess = doSleep(2000); // sleep two seconds
int c = 3;
int d = 4;
would generate the following Java code:
int a = 1;
int b = 2;
doSleep(2000, new DoSleepCallback() {
public void onTrigger(boolean rc) {
boolean sleepSuccess = rc;
int c = 3;
int d = 4;
}
});
To achieve it, I defined the grammar this way:
grammar org.qedlang.qed.QED with jbase.Jbase // Jbase inherits Xbase
...
FunctionDeclaration return XExpression:
=>({FunctionDeclaration} type=JvmTypeReference name=ValidID '(')
(params+=FullJvmFormalParameter (',' params+=FullJvmFormalParameter)*)?
')' block=XBlockExpression
;
The FunctionDeclaration rule is used to define asynchronous calls. In my language library, I would have as system call:
boolean doSleep(int millis) {} // async FunctionDeclaration element stub
The underlying Java implementation would be:
public abstract class DoSleepCallback {
public abstract void onTrigger(boolean rc);
}
public void doSleep(int millis, DoSleepCallback callback) {
<perform sleep and call callback.onTrigger(<success>)>
}
So, using the inferrer, type computer and compiler, how to identify calls to FunctionDeclaration elements, add a callback parameter and process the rest of the body in an inner class?
I could, for instance, override appendFeatureCall in the language compiler, would it work? There is still a part I don't know how to do...
override appendFeatureCall(XAbstractFeatureCall call, ITreeAppendable b) {
...
val feature = call.feature
...
if (feature instanceof JvmExecutable) {
b.append('(')
val arguments = call.actualArguments
if (!arguments.isEmpty) {
...
arguments.appendArguments(b, shouldBreakFirstArgument)
// HERE IS THE PART I DON'T KNOW HOW TO DO
<IF feature IS A FunctionDeclaration>
<argument.appendArgument(NEW GENERATED CALLBACK PARAMETER)>
<INSERT REST OF XBlockExpression body INSIDE CALLBACK INSTANCE>
<ENDIF>
}
b.append(');')
}
}
So basically, how to tell if "feature" points to FunctionDeclaration? The rest, I may be able to do it...
Related to another StackOverflow entry, I had the idea of implementing FunctionDeclaration in the inferrer as a class instead of as a method:
def void inferExpressions(JvmDeclaredType it, FunctionDeclaration function) {
// now let's go over the features
for ( f : (function.block as XBlockExpression).expressions ) {
if (f instanceof FunctionDeclaration) {
members += f.toClass(f.fullyQualifiedName) [
inferVariables(f)
superTypes += typeRef(FunctionDeclarationObject)
// let's add a default constructor
members += f.toConstructor [
for (p : f.params)
parameters += p.toParameter(p.name, p.parameterType)
body = f.block
]
inferExpressions(f)
]
}
}
}
The generated class would extend FunctionDeclarationObject, so I thought there was a way to identify FunctionDeclaration as FunctionDeclarationObject subclasses. But then, I would need to extend the XFeatureCall default scoping to include classes in order to making it work...
I fully realize the question is not obvious, sorry...
Thanks,
Martin
EDIT: modified DoSleepCallback declaration from static to abstract (was erroneous)
I don't think you can generate what you need using the jvm model inferrer.
You should provide your own subclass of the XbaseCompiler (or JBaseCompiler, if any... and don't forget to register with guice in your runtime module), and override doInternalToJavaStatement(XExpression expr, ITreeAppendable it, boolean isReferenced) to manage how your FunctionDeclaration should be generated.
On Dart 1.0.0, I just tried:
class MyClass {
int x;
bool b;
MyClass(int x, [bool b = true]) {
if(?b) {
// ...
}
}
}
And am getting a compiler error on the ?b part:
The argument definition test ('?' operator) has been deprecated
So what's the "new" way of testing for whether or not an argument was supplied?
There is no way to test if an argument was provided or not. The main-reason for its removal was, that it was very complex to forward calls this way.
The generally preferred way is to use null as "not given". This doesn't always work (for example if null is a valid value), and won't catch bad arguments. If null is used, then the parameter must not have a default-value. Otherwise the parameter is not null but takes the default-value:
foo([x = true, y]) => print("$x, $y");
foo(); // prints "true, null"
So in your case you should probably do:
class MyClass {
int x;
bool b;
MyClass(int x, [bool b]) {
if(b == null) { // treat as if not given.
// ...
}
}
}
This makes new MyClass(5, null) and new MyClass(5) identical. If you really need to catch the first case, you have to work around the type-system:
class _Sentinel { const _Sentinel(); }
...
MyClass(int x, [b = const _Sentinel()]) {
if (b == const _Sentinel()) b = true;
...
}
This way you can check if an argument has been provided. In return you lose the type on b.
The argument definition test operator was deprecated because it was redundant with checking for null; an optional parameter that was omitted would get the value null, and the caller could've passed null explicitly anyway. So instead use == null:
class MyClass {
int x;
bool b;
MyClass(int x, [bool b]) {
if (b == null) {
// throw exception or assign default value for b
}
}
}