What's so great about Func<> delegate? - delegates

Sorry if this is basic but I was trying to pick up on .Net 3.5.
Question: Is there anything great about Func<> and it's 5 overloads? From the looks of it, I can still create a similar delgate on my own say, MyFunc<> with the exact 5 overloads and even more.
eg: public delegate TResult MyFunc<TResult>() and a combo of various overloads...
The thought came up as I was trying to understand Func<> delegates and hit upon the following scenario:
Func<int,int> myDelegate = (y) => IsComposite(10);
This implies a delegate with one parameter of type int and a return type of type int. There are five variations (if you look at the overloads through intellisense). So I am guessing that we can have a delegate with no return type?
So am I justified in saying that Func<> is nothing great and just an example in the .Net framework that we can use and if needed, create custom "func<>" delegates to suit our own needs?
Thanks,

The greatness lies in establishing shared language for better communication.
Instead of defining your own delegate types for the same thing (delegate explosion), use the ones provided by the framework. Anyone reading your code instantly grasps what you are trying to accomplish.. minimizes the time to 'what is this piece of code actually doing?'
So as soon as I see a
Action = some method that just does something and returns no output
Comparison = some method that compares two objects of the same type and returns an int to indicate order
Converter = transforms Obj A into equivalent Obj B
EventHandler = response/handler to an event raised by some object given some input in the form of an event argument
Func = some method that takes some parameters, computes something and returns a result
Predicate = evaluate input object against some criteria and return pass/fail status as bool
I don't have to dig deeper than that unless it is my immediate area of concern. So if you feel the delegate you need fits one of these needs, use them before rolling your own.
Disclaimer: Personally I like this move by the language designers.
Counter-argument : Sometimes defining your delegate may help communicate intent better. e.g. System.Threading.ThreadStart over System.Action. So it’s a judgment call in the end.

The Func family of delegates (and their return-type-less cousins, Action) are not any greater than anything else you'd find in the .NET framework. They're just there for re-use so you don't have to redefine them. They have type parameters to keep things generic. E.g., a Func<T0,bool> is the same as a System.Predicate<T> delegate. They were originally designed for LINQ.
You should be able to just use the built-in Func delegate for any value-returning method that accepts up to 4 arguments instead of defining your own delegate for such a purpose unless you want the name to reflect your intention, which is cool.
Cases where you would absolutely need to define your delegate types include methods that accept more than 4 arguments, methods with out, ref, or params parameters, or recursive method signatures (e.g., delegate Foo Foo(Foo f)).

In addition to Marxidad's correct answer:
It's worth being aware of Func's related family, the Action delegates. Again, these are types overloaded by the number of type parameters, but declared to return void.
If you want to use Func/Action in a .NET 2.0 project but with a simple route to upgrading later on, you can cut and paste the declarations from my version comparison page. If you declare them in the System namespace then you'll be able to upgrade just by removing the declarations later - but then you won't be able to (easily) build the same code in .NET 3.5 without removing the declarations.

Decoupling dependencies and unholy tie-ups is one singular thing that makes it great. Everything else one can debate and claim to be doable in some home-grown way.
I've been refactoring slightly more complex system with an old and heavy lib and got blocked on not being able to break compile time dependency - because of the named delegate lurking on "the other side". All assembly loading and reflection didn't help - compiler would refuse to just cast a delegate() {...} to object and whatever you do to pacify it would fail on the other side.
Delegate type comparison which is structural at compile time turns nominal after that (loading, invoking). That may seem OK while you are thinking in terms of "my darling lib is going to be used forever and by everyone" but it doesn't scale to even slightly more complex systems. Fun<> templates bring a degree of structural equivalence back into the world of nominal typing . That's the aspect you can't achieve by rolling out your own.
Example - converting:
class Session (
public delegate string CleanBody(); // tying you up and you don't see it :-)
public static void Execute(string name, string q, CleanBody body) ...
to:
public static void Execute(string name, string q, Func<string> body)
Allows completely independent code to do reflection invocation like:
Type type = Type.GetType("Bla.Session, FooSessionDll", true);
MethodInfo methodInfo = type.GetMethod("Execute");
Func<string> d = delegate() { .....} // see Ma - no tie-ups :-)
Object [] params = { "foo", "bar", d};
methodInfo.Invoke("Trial Execution :-)", params);
Existing code doesn't notice the difference, new code doesn't get dependence - peace on Earth :-)

One thing I like about delegates is that they let me declare methods within methods like so, this is handy when you want to reuse a piece of code but you only need it within that method. Since the purpose here is to limit the scope as much as possible Func<> comes in handy.
For example:
string FormatName(string pFirstName, string pLastName) {
Func<string, string> MakeFirstUpper = (pText) => {
return pText.Substring(0,1).ToUpper() + pText.Substring(1);
};
return MakeFirstUpper(pFirstName) + " " + MakeFirstUpper(pLastName);
}
It's even easier and more handy when you can use inference, which you can if you create a helper function like so:
Func<T, TReturn> Lambda<T, TReturn>(Func<T, TReturn> pFunc) {
return pFunc;
}
Now I can rewrite my function without the Func<>:
string FormatName(string pFirstName, string pLastName) {
var MakeFirstUpper = Lambda((string pText) => {
return pText.Substring(0,1).ToUpper() + pText.Substring(1);
});
return MakeFirstUpper(pFirstName) + " " + MakeFirstUpper(pLastName);
}
Here's the code to test the method:
Console.WriteLine(FormatName("luis", "perez"));

Though it is an old thread I had to add that func<> and action<> also help us use covariance and contra variance.
http://msdn.microsoft.com/en-us/library/dd465122.aspx

Related

Removing a std::function<()> from a vector c++

I'm building a publish-subscribe class (called SystermInterface), which is responsible to receive updates from its instances, and publish them to subscribers.
Adding a subscriber callback function is trivial and has no issues, but removing it yields an error, because std::function<()> is not comparable in C++.
std::vector<std::function<void()> subs;
void subscribe(std::function<void()> f)
{
subs.push_back(f);
}
void unsubscribe(std::function<void()> f)
{
std::remove(subs.begin(), subs.end(), f); // Error
}
I've came down to five solutions to this error:
Registering the function using a weak_ptr, where the subscriber must keep the returned shared_ptr alive.
Solution example at this link.
Instead of registering at a vector, map the callback function by a custom key, unique per callback function.
Solution example at this link
Using vector of function pointers. Example
Make the callback function comparable by utilizing the address.
Use an interface class (parent class) to call a virtual function.
In my design, all intended classes inherits a parent class called
ServiceCore, So instead of registering a callback function, just
register ServiceCore reference in the vector.
Given that the SystemInterface class has a field attribute per instance (ID) (Which is managed by ServiceCore, and supplied to SystemInterface by constructing a ServiceCore child instance).
To my perspective, the first solution is neat and would work, but it requires handling at subscribers, which is something I don't really prefer.
The second solution would make my implementation more complex, where my implementation looks as:
using namespace std;
enum INFO_SUB_IMPORTANCE : uint8_t
{
INFO_SUB_PRIMARY, // Only gets the important updates.
INFO_SUB_COMPLEMENTARY, // Gets more.
INFO_SUB_ALL // Gets all updates
};
using CBF = function<void(string,string)>;
using INFO_SUBTREE = map<INFO_SUB_IMPORTANCE, vector<CBF>>;
using REQINF_SUBS = map<string, INFO_SUBTREE>; // It's keyed by an iterator, explaining it goes out of the question scope.
using INFSRC_SUBS = map<string, INFO_SUBTREE>;
using WILD_SUBS = INFO_SUBTREE;
REQINF_SUBS infoSubrs;
INFSRC_SUBS sourceSubrs;
WILD_SUBS wildSubrs;
void subscribeInfo(string info, INFO_SUB_IMPORTANCE imp, CBF f) {
infoSubrs[info][imp].push_back(f);
}
void subscribeSource(string source, INFO_SUB_IMPORTANCE imp, CBF f) {
sourceSubrs[source][imp].push_back(f);
}
void subscribeWild(INFO_SUB_IMPORTANCE imp, CBF f) {
wildSubrs[imp].push_back(f);
}
The second solution would require INFO_SUBTREE to be an extended map, but can be keyed by an ID:
using KEY_T = uint32_t; // or string...
using INFO_SUBTREE = map<INFO_SUB_IMPORTANCE, map<KEY_T,CBF>>;
For the third solution, I'm not aware of the limitations given by using function pointers, and the consequences of the fourth solution.
The Fifth solution would eliminate the purpose of dealing with CBFs, but it'll be more complex at subscriber-side, where a subscriber is required to override the virtual function and so receives all updates at one place, in which further requires filteration of the message id and so direct the payload to the intended routines using multiple if/else blocks, which will increase by increasing subscriptions.
What I'm looking for is an advice for the best available option.
Regarding your proposed solutions:
That would work. It can be made easy for the caller: have subscribe() create the shared_ptr and corresponding weak_ptr objects, and let it return the shared_ptr.
Then the caller must not lose the key. In a way this is similar to the above.
This of course is less generic, and then you can no longer have (the equivalent of) captures.
You can't: there is no way to get the address of the function stored inside a std::function. You can do &f inside subscribe() but that will only give you the address of the local variable f, which will go out of scope as soon as you return.
That works, and is in a way similar to 1 and 2, although now the "key" is provided by the caller.
Options 1, 2 and 5 are similar in that there is some other data stored in subs that refers to the actual std::function: either a std::shared_ptr, a key or a pointer to a base class. I'll present option 6 here, which is kind of similar in spirit but avoids storing any extra data:
Store a std::function<void()> directly, and return the index in the vector where it was stored. When removing an item, don't std::remove() it, but just set it to std::nullptr. Next time subscribe() is called, it checks if there is an empty element in the vector and reuses it:
std::vector<std::function<void()> subs;
std::size_t subscribe(std::function<void()> f) {
if (auto it = std::find(subs.begin(), subs.end(), std::nullptr); it != subs.end()) {
*it = f;
return std::distance(subs.begin(), it);
} else {
subs.push_back(f);
return subs.size() - 1;
}
}
void unsubscribe(std::size_t index) {
subs[index] = std::nullptr;
}
The code that actually calls the functions stored in subs must now of course first check against std::nullptrs. The above works because std::nullptr is treated as the "empty" function, and there is an operator==() overload that can check a std::function against std::nullptr, thus making std::find() work.
One drawback of option 6 as shown above is that a std::size_t is a rather generic type. To make it safer, you might wrap it in a class SubscriptionHandle or something like that.
As for the best solution: option 1 is quite heavy-weight. Options 2 and 5 are very reasonable, but 6 is, I think, the most efficient.

Why does using Linq's .Select() return IEnumerable<dynamic> even though the type is clearly defined?

I'm using Dapper to return dynamic objects and sometimes mapping them manually. Everything's working fine, but I was wondering what the laws of casting were and why the following examples hold true.
(for these examples I used 'StringBuilder' as my known type, though it is usually something like 'Product')
Example1: Why does this return an IEnumerable<dynamic> even though 'makeStringBuilder' clearly returns a StringBuilder object?
Example2: Why does this build, but 'Example1' wouldn't if it was IEnumerable<StringBuilder>?
Example3: Same question as Example2?
private void test()
{
List<dynamic> dynamicObjects = {Some list of dynamic objects};
IEnumerable<dynamic> example1 = dynamicObjects.Select(s => makeStringBuilder(s));
IEnumerable<StringBuilder> example2 = dynamicObjects.Select(s => (StringBuilder)makeStringBuilder(s));
IEnumerable<StringBuilder> example3 = dynamicObjects.Select(s => makeStringBuilder(s)).Cast<StringBuilder>();
}
private StringBuilder makeStringBuilder(dynamic s)
{
return new StringBuilder(s);
}
With the above examples, is there a recommended way of handling this? and does casting like this hurt performance? Thanks!
When you use dynamic, even as a parameter, the entire expression is handled via dynamic binding and will result in being "dynamic" at compile time (since it's based on its run-time type). This is covered in 7.2.2 of the C# spec:
However, if an expression is a dynamic expression (i.e. has the type dynamic) this indicates that any binding that it participates in should be based on its run-time type (i.e. the actual type of the object it denotes at run-time) rather than the type it has at compile-time. The binding of such an operation is therefore deferred until the time where the operation is to be executed during the running of the program. This is referred to as dynamic binding.
In your case, using the cast will safely convert this to an IEnumerable<StringBuilder>, and should have very little impact on performance. The example2 version is very slightly more efficient than the example3 version, but both have very little overhead when used in this way.
While I can't speak very well to the "why", I think you should be able to write example1 as:
IEnumerable<StringBuilder> example1 = dynamicObjects.Select<dynamic, StringBuilder>(s => makeStringBuilder(s));
You need to tell the compiler what type the projection should take, though I'm sure someone else can clarify why it can't infer the correct type. But I believe by specifying the projection type, you can avoid having to actually cast, which should yield some performance benefit.

Does this method call violate the Law Of Demeter?

Say you have something like the following (sadly, I'm not allowed to post the original code):
public void foo() {
MyObject obj = getMyObject();
bar(obj);
}
public void bar(MyObject obj) {
Type type = new Type(obj.getOtherObject());
}
foo calls bar, passes in obj. But instead of using obj,it calls a getter on it to retrieve the needed information. Does this violate the Law Of Demeter?
Would it be better to write something like this:
public void foo() {
MyObject obj = getMyObject();
bar(obj.getOtherObject());
}
public void bar(MyOtherObject otherObj) {
Type type = new Type(otherObj);
}
Indeed according to the wiki on the Law of Demeter:
The fundamental notion is that a given object should assume as little
as possible about the structure or properties of anything else...
Your bar assumes that a given MyObject (a concrete type so strongly coupled, again against LoD) has a method called getOtherObject, so your proposed solution sorts the assumption and moves the code closer to adhering to LoD. You can go even further and instead provide the type that bar wants:
bar(new Type(obj.getOtherObject());
Depending on your language, can you not pass an interface/contract instead of a solid type? This would turn the strong coupling into a looser coupling.
Of course, if this is all internal to a given object then perhaps it isn't breaking LoD because it's a "close friend":
Each unit should have only limited knowledge about other units: only units "closely" related to the current unit.
Each unit should only talk to its friends; don't talk to strangers.
Only talk to your immediate friends.
In OO I think your original code is breaking LoD based on this argument:
...an object A can request a service (call a method) of an object
instance B, but object A cannot "reach through" object B to access yet
another object, C, to request its services. Doing so would mean that
object A implicitly requires greater knowledge of object B's internal
structure.
To me it seems that you are using obj in order to call getOtherObj. Your proposed code is a potential solution.

JNI - Converting jobject representing Basic Java Objects (Boolean) to native basic types (bool)

I think I managed to fit most of the question in to the title on this one!
I'm pulling back an Object from Java in my native C++ code:
jobject valueObject = env->CallObjectMethod(hashMapObject, hashMapGetMID, keyObject);
It's possible for me to check wether the return object is one of the native types using something like:
jclass boolClass = env->FindClass("java/lang/Boolean");
if(env->IsInstanceOf(valueObject, boolClass) == JNI_TRUE) { }
So, I now have a jobject which I know is a Boolean (note the upper case B) - The question is, what is the most efficient way (considering I already have the jobject in my native code) to convert this to a bool. Typecasting doesn't work which makes sense.
Although the above example is a Boolean I also want to convert Character->char, Short->short, Integer->int, Float->float, Double->double.
(Once i've implemented it I will post an answer to this which does Boolean.booleanValue())
You have two choices.
Option #1 is what you wrote in your self-answer: use the public method defined for each class to extract the primitive value.
Option #2 is faster but not strictly legal: access the internal field directly. For Boolean, that would be Boolean.value. For each primitive box class you have a fieldID for the "value" field, and you just read the field directly. (JNI cheerfully ignores the fact that it's declared private. You can also write to "final" fields and do other stuff that falls into the "really bad idea" category.)
The name of the "value" field is unlikely to change since that would break serialization. So officially this is not recommended, but in practice you can get away with it if you need to.
Either way, you should be caching the jmethodID / jfieldID values, not looking them up every time (the lookups are relatively expensive).
You could also use the less expensive IsSameObject function rather than IsInstanceof, because the box classes are "final". That requires making an extra GetObjectClass call to get valueObject's class, but you only have to do that once before your various comparisons.
BTW, be careful with your use of "char". In your example above you're casting the result of CallCharMethod (a 16-bit UTF-16 value) to a char (an 8-bit value). Remember, char != jchar (unless you're somehow configured for wide chars), long != jlong (unless you're compiling with 64-bit longs).
This is the solution I'm going to use if I get no more input. Hopefully it isn't this difficult but knowing JNI i'm thinking it might be:
if (env->IsInstanceOf(valueObject, boolClass) == JNI_TRUE)
{
jmethodID booleanValueMID = env->GetMethodID(boolClass, "booleanValue", "()Z");
bool booleanValue = (bool) env->CallBooleanMethod(valueObject, booleanValueMID);
addBoolean(key, booleanValue);
}
else if(env->IsInstanceOf(valueObject, charClass) == JNI_TRUE)
{
jmethodID characterValueMID = env->GetMethodID(charClass, "charValue", "()C");
char characterValue = (char) env->CallCharMethod(valueObject, characterValueMID);
addChar (key, characterValue);
}
In general, I write jni for the better performance.
How to gain the better performance ? Using asm, primitive types and few method call.
I suggest that design your method return type can use in c/c++, such as
jint, jlong, jboolean, jbyte and jchar etc.
The redundant function call and convert will make inefficient and unmaintainable implementation.

What's the best way to refactor a method that has too many (6+) parameters?

Occasionally I come across methods with an uncomfortable number of parameters. More often than not, they seem to be constructors. It seems like there ought to be a better way, but I can't see what it is.
return new Shniz(foo, bar, baz, quux, fred, wilma, barney, dino, donkey)
I've thought of using structs to represent the list of parameters, but that just seems to shift the problem from one place to another, and create another type in the process.
ShnizArgs args = new ShnizArgs(foo, bar, baz, quux, fred, wilma, barney, dino, donkey)
return new Shniz(args);
So that doesn't seem like an improvement. So what is the best approach?
I'm going to assume you mean C#. Some of these things apply to other languages, too.
You have several options:
switch from constructor to property setters. This can make code more readable, because it's obvious to the reader which value corresponds to which parameters. Object Initializer syntax makes this look nice. It's also simple to implement, since you can just use auto-generated properties and skip writing the constructors.
class C
{
public string S { get; set; }
public int I { get; set; }
}
new C { S = "hi", I = 3 };
However, you lose immutability, and you lose the ability to ensure that the required values are set before using the object at compile time.
Builder Pattern.
Think about the relationship between string and StringBuilder. You can get this for your own classes. I like to implement it as a nested class, so class C has related class C.Builder. I also like a fluent interface on the builder. Done right, you can get syntax like this:
C c = new C.Builder()
.SetX(4) // SetX is the fluent equivalent to a property setter
.SetY("hello")
.ToC(); // ToC is the builder pattern analog to ToString()
// Modify without breaking immutability
c = c.ToBuilder().SetX(2).ToC();
// Still useful to have a traditional ctor:
c = new C(1, "...");
// And object initializer syntax is still available:
c = new C.Builder { X = 4, Y = "boing" }.ToC();
I have a PowerShell script that lets me generate the builder code to do all this, where the input looks like:
class C {
field I X
field string Y
}
So I can generate at compile time. partial classes let me extend both the main class and the builder without modifying the generated code.
"Introduce Parameter Object" refactoring. See the Refactoring Catalog. The idea is that you take some of the parameters you're passing and put them in to a new type, and then pass an instance of that type instead. If you do this without thinking, you will end up back where you started:
new C(a, b, c, d);
becomes
new C(new D(a, b, c, d));
However, this approach has the greatest potential to make a positive impact on your code. So, continue by following these steps:
Look for subsets of parameters that make sense together. Just mindlessly grouping all parameters of a function together doesn't get you much; the goal is to have groupings that make sense. You'll know you got it right when the name of the new type is obvious.
Look for other places where these values are used together, and use the new type there, too. Chances are, when you've found a good new type for a set of values that you already use all over the place, that new type will make sense in all those places, too.
Look for functionality that is in the existing code, but belongs on the new type.
For example, maybe you see some code that looks like:
bool SpeedIsAcceptable(int minSpeed, int maxSpeed, int currentSpeed)
{
return currentSpeed >= minSpeed & currentSpeed < maxSpeed;
}
You could take the minSpeed and maxSpeed parameters and put them in a new type:
class SpeedRange
{
public int Min;
public int Max;
}
bool SpeedIsAcceptable(SpeedRange sr, int currentSpeed)
{
return currentSpeed >= sr.Min & currentSpeed < sr.Max;
}
This is better, but to really take advantage of the new type, move the comparisons into the new type:
class SpeedRange
{
public int Min;
public int Max;
bool Contains(int speed)
{
return speed >= min & speed < Max;
}
}
bool SpeedIsAcceptable(SpeedRange sr, int currentSpeed)
{
return sr.Contains(currentSpeed);
}
And now we're getting somewhere: the implementation of SpeedIsAcceptable() now says what you mean, and you have a useful, reusable class. (The next obvious step is to make SpeedRange in to Range<Speed>.)
As you can see, Introduce Parameter Object was a good start, but its real value was that it helped us discover a useful type that has been missing from our model.
The best way would be to find ways to group the arguments together. This assumes, and really only works if, you would end up with multiple "groupings" of arguments.
For instance, if you are passing the specification for a rectangle, you can pass x, y, width, and height or you could just pass a rectangle object that contains x, y, width, and height.
Look for things like this when refactoring to clean it up somewhat. If the arguments really can't be combined, start looking at whether you have a violation of the Single Responsibility Principle.
If it's a constructor, particularly if there are multiple overloaded variants, you should look at the Builder pattern:
Foo foo = new Foo()
.configBar(anything)
.configBaz(something, somethingElse)
// and so on
If it's a normal method, you should think about the relationships between the values being passed, and perhaps create a Transfer Object.
The classic answer to this is to use a class to encapsulate some, or all, of the parameters. In theory that sounds great, but I'm the kind of guy who creates classes for concepts that have meaning in the domain, so it's not always easy to apply this advice.
E.g. instead of:
driver.connect(host, user, pass)
You could use
config = new Configuration()
config.setHost(host)
config.setUser(user)
config.setPass(pass)
driver.connect(config)
YMMV
When I see long parameter lists, my first question is whether this function or object is doing too much. Consider:
EverythingInTheWorld earth=new EverythingInTheWorld(firstCustomerId,
lastCustomerId,
orderNumber, productCode, lastFileUpdateDate,
employeeOfTheMonthWinnerForLastMarch,
yearMyHometownWasIncorporated, greatGrandmothersBloodType,
planetName, planetSize, percentWater, ... etc ...);
Of course this example is deliberately ridiculous, but I've seen plenty of real programs with examples only slightly less ridiculous, where one class is used to hold many barely related or unrelated things, apparently just because the same calling program needs both or because the programmer happened to think of both at the same time. Sometimes the easy solution is to just break the class into multiple pieces each of which does its own thing.
Just slightly more complicated is when a class really does need to deal with multiple logical things, like both a customer order and general information about the customer. In these cases, crate a class for customer and a class for order, and let them talk to each other as necessary. So instead of:
Order order=new Order(customerName, customerAddress, customerCity,
customerState, customerZip,
orderNumber, orderType, orderDate, deliveryDate);
We could have:
Customer customer=new Customer(customerName, customerAddress,
customerCity, customerState, customerZip);
Order order=new Order(customer, orderNumber, orderType, orderDate, deliveryDate);
While of course I prefer functions that take just 1 or 2 or 3 parameters, sometimes we have to accept that, realistically, this function takes a bunch, and that the number of itself does not really create complexity. For example:
Employee employee=new Employee(employeeId, firstName, lastName,
socialSecurityNumber,
address, city, state, zip);
Yeah, it's a bunch of fields, but probably all we're going to do with them is save them to a database record or throw them on a screen or some such. There's not really a lot of processing here.
When my parameter lists do get long, I much prefer if I can give the fields different data types. Like when I see a function like:
void updateCustomer(String type, String status,
int lastOrderNumber, int pastDue, int deliveryCode, int birthYear,
int addressCode,
boolean newCustomer, boolean taxExempt, boolean creditWatch,
boolean foo, boolean bar);
And then I see it called with:
updateCustomer("A", "M", 42, 3, 1492, 1969, -7, true, false, false, true, false);
I get concerned. Looking at the call, it's not at all clear what all these cryptic numbers, codes, and flags mean. This is just asking for errors. A programmer might easily get confused about the order of the parameters and accidentally switch two, and if they're the same data type, the compiler would just accept it. I'd much rather have a signature where all these things are enums, so a call passes in things like Type.ACTIVE instead of "A" and CreditWatch.NO instead of "false", etc.
This is quoted from Fowler and Beck book: "Refactoring"
Long Parameter List
In our early programming days we were taught to pass in as parameters everything needed by
a routine. This was understandable because the alternative was global data, and global data is
evil and usually painful. Objects change this situation because if you don't have something
you need, you can always ask another object to get it for you. Thus with objects you don't
pass in everything the method needs; instead you pass enough so that the method can get to
everything it needs. A lot of what a method needs is available on the method's host class. In
object-oriented programs parameter lists tend to be much smaller than in traditional
programs.
This is good because long parameter lists are hard to understand, because they become
inconsistent and difficult to use, and because you are forever changing them as you need
more data. Most changes are removed by passing objects because you are much more likely
to need to make only a couple of requests to get at a new piece of data.
Use Replace Parameter with Method when you can get the data in one parameter by making
a request of an object you already know about. This object might be a field or it might be
another parameter. Use Preserve Whole Object to take a bunch of data gleaned from an
object and replace it with the object itself. If you have several data items with no logical
object, use Introduce Parameter Object.
There is one important exception to making these changes. This is when you explicitly do
not want to create a dependency from the called object to the larger object. In those cases
unpacking data and sending it along as parameters is reasonable, but pay attention to the pain
involved. If the parameter list is too long or changes too often, you need to rethink your
dependency structure.
I don't want to sound like a wise-crack, but you should also check to make sure the data you are passing around really should be passed around: Passing stuff to a constructor (or method for that matter) smells a bit like to little emphasis on the behavior of an object.
Don't get me wrong: Methods and constructors will have a lot of parameters sometimes. But when encountered, do try to consider encapsulating data with behavior instead.
This kind of smell (since we are talking about refactoring, this horrible word seems appropriate...) might also be detected for objects that have a lot (read: any) properties or getters/setters.
If some of the constructor parameters are optional it makes sense to use a builder, which would get the required parameters in the constructor, and have methods for the optional ones, returning the builder, to be used like this:
return new Shniz.Builder(foo, bar).baz(baz).quux(quux).build();
The details of this are described in Effective Java, 2nd Ed., p. 11. For method arguments, the same book (p. 189) describes three approaches for shortening parameter lists:
Break the method into multiple methods that take fewer arguments
Create static helper member classes to represent groups of parameters, i.e. pass a DinoDonkey instead of dino and donkey
If parameters are optional, the builder above can be adopted for methods, defining an object for all parameters, setting the required ones and then calling some execute method on it
You can try to group your parameter into multiples meaningful struct/class (if possible).
I would generally lean towards the structs approach - presumably the majority of these parameters are related in some way and represent the state of some element that is relevant to your method.
If the set of parameters can't be made into a meaningful object, that's probably a sign that Shniz is doing too much, and the refactoring should involve breaking the method down into separate concerns.
I would use the default constructor and property settors. C# 3.0 has some nice syntax to do this automagically.
return new Shniz { Foo = foo,
Bar = bar,
Baz = baz,
Quuz = quux,
Fred = fred,
Wilma = wilma,
Barney = barney,
Dino = dino,
Donkey = donkey
};
The code improvement comes in simplifying the constructor and not having to support multiple methods to support various combinations. The "calling" syntax is still a little "wordy", but not really any worse than calling the property settors manually.
You haven't provided enough information to warrant a good answer. A long parameter list isn't inherently bad.
Shniz(foo, bar, baz, quux, fred, wilma, barney, dino, donkey)
could be interpreted as:
void Shniz(int foo, int bar, int baz, int quux, int fred,
int wilma, int barney, int dino, int donkey) { ...
In this case you're far better off to create a class to encapsulate the parameters because you give meaning to the different parameters in a way that the compiler can check as well as visually making the code easier to read. It also makes it easier to read and refactor later.
// old way
Shniz(1,2,3,2,3,2,1,2);
Shniz(1,2,2,3,3,2,1,2);
//versus
ShnizParam p = new ShnizParam { Foo = 1, Bar = 2, Baz = 3 };
Shniz(p);
Alternatively if you had:
void Shniz(Foo foo, Bar bar, Baz baz, Quux quux, Fred fred,
Wilma wilma, Barney barney, Dino dino, Donkey donkey) { ...
This is a far different case because all the objects are different (and aren't likely to be muddled up). Agreed that if all objects are necessary, and they're all different, it makes little sense to create a parameter class.
Additionally, are some parameters optional? Are there method override's (same method name, but different method signatures?) These sorts of details all matter as to what the best answer is.
* A property bag can be useful as well, but not specifically better given that there is no background given.
As you can see, there is more than 1 correct answer to this question. Take your pick.
If you have that many parameters, chances are that the method is doing too much, so address this first by splitting the method into several smaller methods. If you still have too many parameters after this try grouping the arguments or turning some of the parameters into instance members.
Prefer small classes/methods over large. Remember the single responsibility principle.
You can trade complexity for source code lines. If the method itself does too much (Swiss knife) try to halve its tasks by creating another method. If the method is simple only it needs too many parameters then the so called parameter objects are the way to go.
If your language supports it, use named parameters and make as many optional (with reasonable defaults) as possible.
I think the method you described is the way to go. When I find a method with a lot of parameters and/or one that is likely to need more in the future, I usually create a ShnizParams object to pass through, like you describe.
How about not setting it in all at once at the constructors but doing it via properties/setters? I have seen some .NET classes that utilize this approach such as Process class:
Process p = new Process();
p.StartInfo.UseShellExecute = false;
p.StartInfo.CreateNoWindow = true;
p.StartInfo.RedirectStandardOutput = true;
p.StartInfo.RedirectStandardError = true;
p.StartInfo.FileName = "cmd";
p.StartInfo.Arguments = "/c dir";
p.Start();
I concur with the approach of moving the parameters into a parameter object (struct). Rather than just sticking them all in one object though, review if other functions use similar groups of parameters. A paramater object is more valuable if its used with multiple functions where you expect that set of parameters to change consistently across those functions. It may be that you only put some of the parameters into the new parameter object.
Named arguments are a good option (presuming a language which supports them) for disambiguating long (or even short!) parameter lists while also allowing (in the case of constructors) the class's properties to be immutable without imposing a requirement for allowing it to exist in a partially-constructed state.
The other option I would look for in doing this sort of refactor would be groups of related parameters which might be better handled as an independent object. Using the Rectangle class from an earlier answer as an example, the constructor which takes parameters for x, y, height, and width could factor x and y out into a Point object, allowing you to pass three parameters to the Rectangle's constructor. Or go a little further and make it two parameters (UpperLeftPoint, LowerRightPoint), but that would be a more radical refactoring.
It depends on what kind of arguments you have, but if they are a lot of boolean values/options maybe you could use a Flag Enum?
I think that problem is deeply tied to the domain of the problem you're trying to solve with the class.
In some cases, a 7-parameter constructor may indicate a bad class hierarchy: in that case, the helper struct/class suggested above is usually a good approach, but then you also tend to end up with loads of structs which are just property bags and don't do anything useful.
The 8-argument constructor might also indicate that your class is too generic / too all-purpose so it needs a lot of options to be really useful. In that case you can either refactor the class or implement static constructors that hide the real complex constructors: eg. Shniz.NewBaz (foo, bar) could actually call the real constructor passing the right parameters.
One consideration is which of the values would be read-only once the object is created?
Publicly writable properties could perhaps be assigned after construction.
Where ultimately do the values come from? Perhaps some values are truely external where as others are really from some configuration or global data that is maintained by the library.
In this case you could conceal the constructor from external use and provide a Create function for it. The create function takes the truely external values and constructs the object, then uses accessors only avaiable to the library to complete the creation of the object.
It would be really strange to have an object that requires 7 or more parameters to give the object a complete state and all truely being external in nature.
When a clas has a constructor that takes too many arguments, it is usually a sign that it has too many responsibilities. It can probably be broken into separate classes that cooperate to give the same functionalities.
In case you really need that many arguments to a constructor, the Builder pattern can help you. The goal is to still pass all the arguments to the constructor, so its state is initialized from the start and you can still make the class immutable if needed.
See below :
public class Toto {
private final String state0;
private final String state1;
private final String state2;
private final String state3;
public Toto(String arg0, String arg1, String arg2, String arg3) {
this.state0 = arg0;
this.state1 = arg1;
this.state2 = arg2;
this.state3 = arg3;
}
public static class TotoBuilder {
private String arg0;
private String arg1;
private String arg2;
private String arg3;
public TotoBuilder addArg0(String arg) {
this.arg0 = arg;
return this;
}
public TotoBuilder addArg1(String arg) {
this.arg1 = arg;
return this;
}
public TotoBuilder addArg2(String arg) {
this.arg2 = arg;
return this;
}
public TotoBuilder addArg3(String arg) {
this.arg3 = arg;
return this;
}
public Toto newInstance() {
// maybe add some validation ...
return new Toto(this.arg0, this.arg1, this.arg2, this.arg3);
}
}
public static void main(String[] args) {
Toto toto = new TotoBuilder()
.addArg0("0")
.addArg1("1")
.addArg2("2")
.addArg3("3")
.newInstance();
}
}
The short answer is that:
You need to group the related parameters or redesigning our model
Below example, the constructor takes 8 parameters
public Rectangle(
int point1X,
int point1Y,
int point2X,
int point2Y,
int point3X,
int point3Y,
int point4X,
int point4Y) {
this.point1X = point1X;
this.point1Y = point1Y;
this.point2X = point2X;
this.point2Y = point2Y;
this.point3X = point3X;
this.point3Y = point3Y;
this.point4X = point4X;
this.point4Y = point4Y;
}
After grouping the related parameters,
Then, the constructor will take ONLY 4 parameters
public Rectangle(
Point point1,
Point point2,
Point point3,
Point point4) {
this.point1 = point1;
this.point2 = point2;
this.point3 = point3;
this.point4 = point4;
}
public Point(int x, int y) {
this.x = x;
this.y= y;
}
Or even make the constructor smarter,
After redesigning our model
Then, the constructor will take ONLY 2 parameters
public Rectangle(
Point leftLowerPoint,
Point rightUpperPoint) {
this.leftLowerPoint = leftLowerPoint;
this.rightUpperPoint = rightUpperPoint;
}

Resources