Related
This question already has answers here:
What does it mean to "program to an interface"?
(33 answers)
Closed 2 years ago.
(1) List<?> myList = new ArrayList<?>();
(2) ArrayList<?> myList = new ArrayList<?>();
I understand that with (1), implementations of the List interface can be swapped. It seems that (1) is typically used in an application regardless of need (myself I always use this).
I am wondering if anyone uses (2)?
Also, how often (and can I please get an example) does the situation actually require using (1) over (2) (i.e. where (2) wouldn't suffice..aside coding to interfaces and best practices etc.)
Almost always List is preferred over ArrayList because, for instance, List can be translated into a LinkedList without affecting the rest of the codebase.
If one used ArrayList instead of List, it's hard to change the ArrayList implementation into a LinkedList one because ArrayList specific methods have been used in the codebase that would also require restructuring.
You can read about the List implementations here.
You may start with an ArrayList, but soon after discover that another implementation is the more appropriate choice.
I am wondering if anyone uses (2)?
Yes. But rarely for a sound reason (IMO).
And people get burned because they used ArrayList when they should have used List:
Utility methods like Collections.singletonList(...) or Arrays.asList(...) don't return an ArrayList.
Methods in the List API don't guarantee to return a list of the same type.
For example of someone getting burned, in https://stackoverflow.com/a/1481123/139985 the poster had problems with "slicing" because ArrayList.sublist(...) doesn't return an ArrayList ... and he had designed his code to use ArrayList as the type of all of his list variables. He ended up "solving" the problem by copying the sublist into a new ArrayList.
The argument that you need to know how the List behaves is largely addressed by using the RandomAccess marker interface. Yes, it is a bit clunky, but the alternative is worse.
Also, how often does the situation actually require using (1) over (2) (i.e. where (2) wouldn't suffice..aside 'coding to interfaces' and best practices etc.)
The "how often" part of the question is objectively unanswerable.
(and can I please get an example)
Occasionally, the application may require that you use methods in the ArrayList API that are not in the List API. For example, ensureCapacity(int), trimToSize() or removeRange(int, int). (And the last one will only arise if you have created a subtype of ArrayList that declares the method to be public.)
That is the only sound reason for coding to the class rather than the interface, IMO.
(It is theoretically possible that you will get a slight improvement in performance ... under certain circumstances ... on some platforms ... but unless you really need that last 0.05%, it is not worth doing this. This is not a sound reason, IMO.)
You can’t write efficient code if you don’t know whether random access is efficient or not.
That is a valid point. However, Java provides better ways to deal with that; e.g.
public <T extends List & RandomAccess> void test(T list) {
// do stuff
}
If you call that with a list that does not implement RandomAccess you will get a compilation error.
You could also test dynamically ... using instanceof ... if static typing is too awkward. And you could even write your code to use different algorithms (dynamically) depending on whether or not a list supported random access.
Note that ArrayList is not the only list class that implements RandomAccess. Others include CopyOnWriteList, Stack and Vector.
I've seen people make the same argument about Serializable (because List doesn't implement it) ... but the approach above solves this problem too. (To the extent that it is solvable at all using runtime types. An ArrayList will fail serialization if any element is not serializable.)
Finally, I'm not going to say "because its is good style". That "reason" is both a circular argument ("Why is it 'good style'?") and an appeal to an unstated (and probably non-existent!) higher authority ("Who says it is 'good style'?").
(I do think it is good style to program to the interface, but I'm not going to give that as a reason. It is better for you to understand the real reasons and come to the (IMO) correct conclusions for yourself. The correct conclusion may not always be the same ... depending on the context.)
For example you might decide a LinkedList is the best choice for your application, but then later decide ArrayList might be a better choice for performance reason.
Use:
List list = new ArrayList(100); // will be better also to set the initial capacity of a collection
Instead of:
ArrayList list = new ArrayList();
For reference:
(posted mostly for Collection diagram)
It is considered good style to store a reference to a HashSet or TreeSet in a variable of type Set.
Set<String> names = new HashSet<String>();
This way, you have to change only one line if you decide to use a TreeSet instead.
Also, methods that operate on sets should specify parameters of type Set:
public static void print(Set<String> s)
Then the method can be used for all set implementations.
In theory, we should make the same recommendation for linked lists, namely to save
LinkedList references in variables of type List. However, in the Java library, the List interface is common to both the ArrayList and the LinkedList class. In particular, it has get and set methods for random access, even though these methods are very inefficient for linked lists.
You can’t write efficient code if you don’t know whether random access is efficient or not.
This is plainly a serious design error in the standard library, and I cannot recommend using
the List interface for that reason.
To see just how embarrassing that error is, have a look at
the source code for the binarySearch method of the Collections class. That method takes a
List parameter, but binary search makes no sense for a linked list. The code then clumsily
tries to discover whether the list is a linked list, and then switches to a linear search!
The Set interface and the Map interface, are well designed, and you should use them.
When you write List, you actually tell, that your object implements List interface only, but you don't specify what class your object belongs to.
When you write ArrayList, you specify that your object class is a resizable-array.
So, the first version makes your code more flexible in future.
Look at Java docs:
Class ArrayList - Resizable-array implementation of the List interface.
Interface List - An ordered collection (also known as a sequence). The user of this interface has precise control over where in the list each element is inserted.
Array - container object that holds a fixed number of values of a single type.
I use (2) if code is the "owner" of the list. This is for example true for local-only variables. There is no reason to use the abstract type List instead of ArrayList.
Another example to demonstrate ownership:
public class Test {
// This object is the owner of strings, so use the concrete type.
private final ArrayList<String> strings = new ArrayList<>();
// This object uses the argument but doesn't own it, so use abstract type.
public void addStrings(List<String> add) {
strings.addAll(add);
}
// Here we return the list but we do not give ownership away, so use abstract type. This also allows to create optionally an unmodifiable list.
public List<String> getStrings() {
return Collections.unmodifiableList(strings);
}
// Here we create a new list and give ownership to the caller. Use concrete type.
public ArrayList<String> getStringsCopy() {
return new ArrayList<>(strings);
}
}
I think the people who use (2) don't know the Liskov substitution principle or the Dependency inversion principle. Or they really have to use ArrayList.
Actually there are occasions where (2) is not only preferred but mandatory and I am very surprised, that nobody mentions this here.
Serialization!
If you have a serializable class and you want it to contain a list, then you must declare the field to be of a concrete and serializable type like ArrayList because the List interface does not extend java.io.Serializable
Obviously most people do not need serialization and forget about this.
An example:
public class ExampleData implements java.io.Serializable {
// The following also guarantees that strings is always an ArrayList.
private final ArrayList<String> strings = new ArrayList<>();
(3) Collection myCollection = new ArrayList<?>();
I am using this typically. And only if I need List methods, I will use List. Same with ArrayList. You always can switch to more "narrow" interface, but you can't switch to more "wide".
Out of the following two:
(1) List<?> myList = new ArrayList<?>();
(2) ArrayList<?> myList = new ArrayList<?>();
First is generally preferred. As you will be using methods from List interface only, it provides you the freedom to use some other implementation of List e.g. LinkedList in future. So it decouples you from specific implementation. Now there are two points worth mentioning:
We should always program to interface. More here.
You will almost always end up using ArrayList over LinkedList. More here.
I am wondering if anyone uses (2)
Yes sometimes (read rarely). When we need methods that are part of implementation of ArrayList but not part of the interface List. For example ensureCapacity.
Also, how often (and can I please get an example) does the situation
actually require using (1) over (2)
Almost always you prefer option (1). This is a classical design pattern in OOP where you always try to decouple your code from specific implementation and program to the interface.
List is an interface. It doesn't have methods. When you call a method on a List reference, it in fact calls the method of ArrayList in both cases.
And for the future you can change List obj = new ArrayList<> to List obj = new LinkList<> or other types which implement List interface.
Somebody asked this again (duplicate) which made me go a little deeper on this issue.
public static void main(String[] args) {
List<String> list = new ArrayList<String>();
list.add("a");
list.add("b");
ArrayList<String> aList = new ArrayList<String>();
aList.add("a");
aList.add("b");
}
If we use a bytecode viewer (I used http://asm.ow2.org/eclipse/index.html) weĺl see the following (only list initialization and assignment) for our list snippet:
L0
LINENUMBER 9 L0
NEW ArrayList
DUP
INVOKESPECIAL ArrayList.<init> () : void
ASTORE 1
L1
LINENUMBER 10 L1
ALOAD 1: list
LDC "a"
INVOKEINTERFACE List.add (Object) : boolean
POP
L2
LINENUMBER 11 L2
ALOAD 1: list
LDC "b"
INVOKEINTERFACE List.add (Object) : boolean
POP
and for alist:
L3
LINENUMBER 13 L3
NEW java/util/ArrayList
DUP
INVOKESPECIAL java/util/ArrayList.<init> ()V
ASTORE 2
L4
LINENUMBER 14 L4
ALOAD 2
LDC "a"
INVOKEVIRTUAL java/util/ArrayList.add (Ljava/lang/Object;)Z
POP
L5
LINENUMBER 15 L5
ALOAD 2
LDC "b"
INVOKEVIRTUAL java/util/ArrayList.add (Ljava/lang/Object;)Z
POP
The difference is list ends up calling INVOKEINTERFACE whereas aList calls INVOKEVIRTUAL. Accoding to the Bycode Outline Plugin reference,
invokeinterface is used to invoke a method declared within a Java
interface
while invokevirtual
invokes all methods except interface methods (which use
invokeinterface), static methods (which use invokestatic), and the few
special cases handled by invokespecial.
In summary, invokevirtual pops objectref off the stack while for invokeinterface
the interpreter pops 'n' items off the operand stack, where 'n' is an 8-bit unsigned
integer parameter taken from the bytecode. The first of these items is
objectref, a reference to the object whose method is being called.
If I understand this correctly, the difference is basically how each way retrieves objectref.
The only case that I am aware of where (2) can be better is when using GWT, because it reduces application footprint (not my idea, but the google web toolkit team says so). But for regular java running inside the JVM (1) is probably always better.
I would say that 1 is preferred, unless
you are depending on the implementation of optional behavior* in ArrayList, in that case explicitly using ArrayList is more clear
You will be using the ArrayList in a method call which requires ArrayList, possibly for optional behavior or performance characteristics
My guess is that in 99% of the cases you can get by with List, which is preferred.
for instance removeAll, or add(null)
List interface have several different classes - ArrayList and LinkedList. LinkedList is used to create an indexed collections and ArrayList - to create sorted lists. So, you can use any of it in your arguments, but you can allow others developers who use your code, library, etc. to use different types of lists, not only which you use, so, in this method
ArrayList<Object> myMethod (ArrayList<Object> input) {
// body
}
you can use it only with ArrayList, not LinkedList, but you can allow to use any of List classes on other places where it method is using, it's just your choise, so using an interface can allow it:
List<Object> myMethod (List<Object> input) {
// body
}
In this method arguments you can use any of List classes which you want to use:
List<Object> list = new ArrayList<Object> ();
list.add ("string");
myMethod (list);
CONCLUSION:
Use the interfaces everywhere when it possible, don't restrict you or others to use different methods which they want to use.
This question already has answers here:
What does it mean to "program to an interface"?
(33 answers)
Closed 2 years ago.
(1) List<?> myList = new ArrayList<?>();
(2) ArrayList<?> myList = new ArrayList<?>();
I understand that with (1), implementations of the List interface can be swapped. It seems that (1) is typically used in an application regardless of need (myself I always use this).
I am wondering if anyone uses (2)?
Also, how often (and can I please get an example) does the situation actually require using (1) over (2) (i.e. where (2) wouldn't suffice..aside coding to interfaces and best practices etc.)
Almost always List is preferred over ArrayList because, for instance, List can be translated into a LinkedList without affecting the rest of the codebase.
If one used ArrayList instead of List, it's hard to change the ArrayList implementation into a LinkedList one because ArrayList specific methods have been used in the codebase that would also require restructuring.
You can read about the List implementations here.
You may start with an ArrayList, but soon after discover that another implementation is the more appropriate choice.
I am wondering if anyone uses (2)?
Yes. But rarely for a sound reason (IMO).
And people get burned because they used ArrayList when they should have used List:
Utility methods like Collections.singletonList(...) or Arrays.asList(...) don't return an ArrayList.
Methods in the List API don't guarantee to return a list of the same type.
For example of someone getting burned, in https://stackoverflow.com/a/1481123/139985 the poster had problems with "slicing" because ArrayList.sublist(...) doesn't return an ArrayList ... and he had designed his code to use ArrayList as the type of all of his list variables. He ended up "solving" the problem by copying the sublist into a new ArrayList.
The argument that you need to know how the List behaves is largely addressed by using the RandomAccess marker interface. Yes, it is a bit clunky, but the alternative is worse.
Also, how often does the situation actually require using (1) over (2) (i.e. where (2) wouldn't suffice..aside 'coding to interfaces' and best practices etc.)
The "how often" part of the question is objectively unanswerable.
(and can I please get an example)
Occasionally, the application may require that you use methods in the ArrayList API that are not in the List API. For example, ensureCapacity(int), trimToSize() or removeRange(int, int). (And the last one will only arise if you have created a subtype of ArrayList that declares the method to be public.)
That is the only sound reason for coding to the class rather than the interface, IMO.
(It is theoretically possible that you will get a slight improvement in performance ... under certain circumstances ... on some platforms ... but unless you really need that last 0.05%, it is not worth doing this. This is not a sound reason, IMO.)
You can’t write efficient code if you don’t know whether random access is efficient or not.
That is a valid point. However, Java provides better ways to deal with that; e.g.
public <T extends List & RandomAccess> void test(T list) {
// do stuff
}
If you call that with a list that does not implement RandomAccess you will get a compilation error.
You could also test dynamically ... using instanceof ... if static typing is too awkward. And you could even write your code to use different algorithms (dynamically) depending on whether or not a list supported random access.
Note that ArrayList is not the only list class that implements RandomAccess. Others include CopyOnWriteList, Stack and Vector.
I've seen people make the same argument about Serializable (because List doesn't implement it) ... but the approach above solves this problem too. (To the extent that it is solvable at all using runtime types. An ArrayList will fail serialization if any element is not serializable.)
Finally, I'm not going to say "because its is good style". That "reason" is both a circular argument ("Why is it 'good style'?") and an appeal to an unstated (and probably non-existent!) higher authority ("Who says it is 'good style'?").
(I do think it is good style to program to the interface, but I'm not going to give that as a reason. It is better for you to understand the real reasons and come to the (IMO) correct conclusions for yourself. The correct conclusion may not always be the same ... depending on the context.)
For example you might decide a LinkedList is the best choice for your application, but then later decide ArrayList might be a better choice for performance reason.
Use:
List list = new ArrayList(100); // will be better also to set the initial capacity of a collection
Instead of:
ArrayList list = new ArrayList();
For reference:
(posted mostly for Collection diagram)
It is considered good style to store a reference to a HashSet or TreeSet in a variable of type Set.
Set<String> names = new HashSet<String>();
This way, you have to change only one line if you decide to use a TreeSet instead.
Also, methods that operate on sets should specify parameters of type Set:
public static void print(Set<String> s)
Then the method can be used for all set implementations.
In theory, we should make the same recommendation for linked lists, namely to save
LinkedList references in variables of type List. However, in the Java library, the List interface is common to both the ArrayList and the LinkedList class. In particular, it has get and set methods for random access, even though these methods are very inefficient for linked lists.
You can’t write efficient code if you don’t know whether random access is efficient or not.
This is plainly a serious design error in the standard library, and I cannot recommend using
the List interface for that reason.
To see just how embarrassing that error is, have a look at
the source code for the binarySearch method of the Collections class. That method takes a
List parameter, but binary search makes no sense for a linked list. The code then clumsily
tries to discover whether the list is a linked list, and then switches to a linear search!
The Set interface and the Map interface, are well designed, and you should use them.
When you write List, you actually tell, that your object implements List interface only, but you don't specify what class your object belongs to.
When you write ArrayList, you specify that your object class is a resizable-array.
So, the first version makes your code more flexible in future.
Look at Java docs:
Class ArrayList - Resizable-array implementation of the List interface.
Interface List - An ordered collection (also known as a sequence). The user of this interface has precise control over where in the list each element is inserted.
Array - container object that holds a fixed number of values of a single type.
I use (2) if code is the "owner" of the list. This is for example true for local-only variables. There is no reason to use the abstract type List instead of ArrayList.
Another example to demonstrate ownership:
public class Test {
// This object is the owner of strings, so use the concrete type.
private final ArrayList<String> strings = new ArrayList<>();
// This object uses the argument but doesn't own it, so use abstract type.
public void addStrings(List<String> add) {
strings.addAll(add);
}
// Here we return the list but we do not give ownership away, so use abstract type. This also allows to create optionally an unmodifiable list.
public List<String> getStrings() {
return Collections.unmodifiableList(strings);
}
// Here we create a new list and give ownership to the caller. Use concrete type.
public ArrayList<String> getStringsCopy() {
return new ArrayList<>(strings);
}
}
I think the people who use (2) don't know the Liskov substitution principle or the Dependency inversion principle. Or they really have to use ArrayList.
Actually there are occasions where (2) is not only preferred but mandatory and I am very surprised, that nobody mentions this here.
Serialization!
If you have a serializable class and you want it to contain a list, then you must declare the field to be of a concrete and serializable type like ArrayList because the List interface does not extend java.io.Serializable
Obviously most people do not need serialization and forget about this.
An example:
public class ExampleData implements java.io.Serializable {
// The following also guarantees that strings is always an ArrayList.
private final ArrayList<String> strings = new ArrayList<>();
(3) Collection myCollection = new ArrayList<?>();
I am using this typically. And only if I need List methods, I will use List. Same with ArrayList. You always can switch to more "narrow" interface, but you can't switch to more "wide".
Out of the following two:
(1) List<?> myList = new ArrayList<?>();
(2) ArrayList<?> myList = new ArrayList<?>();
First is generally preferred. As you will be using methods from List interface only, it provides you the freedom to use some other implementation of List e.g. LinkedList in future. So it decouples you from specific implementation. Now there are two points worth mentioning:
We should always program to interface. More here.
You will almost always end up using ArrayList over LinkedList. More here.
I am wondering if anyone uses (2)
Yes sometimes (read rarely). When we need methods that are part of implementation of ArrayList but not part of the interface List. For example ensureCapacity.
Also, how often (and can I please get an example) does the situation
actually require using (1) over (2)
Almost always you prefer option (1). This is a classical design pattern in OOP where you always try to decouple your code from specific implementation and program to the interface.
List is an interface. It doesn't have methods. When you call a method on a List reference, it in fact calls the method of ArrayList in both cases.
And for the future you can change List obj = new ArrayList<> to List obj = new LinkList<> or other types which implement List interface.
Somebody asked this again (duplicate) which made me go a little deeper on this issue.
public static void main(String[] args) {
List<String> list = new ArrayList<String>();
list.add("a");
list.add("b");
ArrayList<String> aList = new ArrayList<String>();
aList.add("a");
aList.add("b");
}
If we use a bytecode viewer (I used http://asm.ow2.org/eclipse/index.html) weĺl see the following (only list initialization and assignment) for our list snippet:
L0
LINENUMBER 9 L0
NEW ArrayList
DUP
INVOKESPECIAL ArrayList.<init> () : void
ASTORE 1
L1
LINENUMBER 10 L1
ALOAD 1: list
LDC "a"
INVOKEINTERFACE List.add (Object) : boolean
POP
L2
LINENUMBER 11 L2
ALOAD 1: list
LDC "b"
INVOKEINTERFACE List.add (Object) : boolean
POP
and for alist:
L3
LINENUMBER 13 L3
NEW java/util/ArrayList
DUP
INVOKESPECIAL java/util/ArrayList.<init> ()V
ASTORE 2
L4
LINENUMBER 14 L4
ALOAD 2
LDC "a"
INVOKEVIRTUAL java/util/ArrayList.add (Ljava/lang/Object;)Z
POP
L5
LINENUMBER 15 L5
ALOAD 2
LDC "b"
INVOKEVIRTUAL java/util/ArrayList.add (Ljava/lang/Object;)Z
POP
The difference is list ends up calling INVOKEINTERFACE whereas aList calls INVOKEVIRTUAL. Accoding to the Bycode Outline Plugin reference,
invokeinterface is used to invoke a method declared within a Java
interface
while invokevirtual
invokes all methods except interface methods (which use
invokeinterface), static methods (which use invokestatic), and the few
special cases handled by invokespecial.
In summary, invokevirtual pops objectref off the stack while for invokeinterface
the interpreter pops 'n' items off the operand stack, where 'n' is an 8-bit unsigned
integer parameter taken from the bytecode. The first of these items is
objectref, a reference to the object whose method is being called.
If I understand this correctly, the difference is basically how each way retrieves objectref.
The only case that I am aware of where (2) can be better is when using GWT, because it reduces application footprint (not my idea, but the google web toolkit team says so). But for regular java running inside the JVM (1) is probably always better.
I would say that 1 is preferred, unless
you are depending on the implementation of optional behavior* in ArrayList, in that case explicitly using ArrayList is more clear
You will be using the ArrayList in a method call which requires ArrayList, possibly for optional behavior or performance characteristics
My guess is that in 99% of the cases you can get by with List, which is preferred.
for instance removeAll, or add(null)
List interface have several different classes - ArrayList and LinkedList. LinkedList is used to create an indexed collections and ArrayList - to create sorted lists. So, you can use any of it in your arguments, but you can allow others developers who use your code, library, etc. to use different types of lists, not only which you use, so, in this method
ArrayList<Object> myMethod (ArrayList<Object> input) {
// body
}
you can use it only with ArrayList, not LinkedList, but you can allow to use any of List classes on other places where it method is using, it's just your choise, so using an interface can allow it:
List<Object> myMethod (List<Object> input) {
// body
}
In this method arguments you can use any of List classes which you want to use:
List<Object> list = new ArrayList<Object> ();
list.add ("string");
myMethod (list);
CONCLUSION:
Use the interfaces everywhere when it possible, don't restrict you or others to use different methods which they want to use.
Occasionally I see somebody create an arraylist like this, why?
List numbers = new ArrayList( );
Instead of:
ArrayList<something> numbers = new ArrayList<something>();
If you asking about using interface instead of concrete object, than it is a good practice. Imagine, you will switch to LinkedList tomorrow. In first case you won't need to fix variable declaration.
If the question was about non-using generics, then it is bad. Generics are always good as they give type safety.
What's good:
1. List is a general case for many implementations.
List trololo = new ListImpl();
Hides real implementation for the user:
public List giveMeTheList(){
List trololo = new SomeCoolListImpl();
return trololo;
}
By design it's good: user shouldn't pay attention to the realization. He just gets interface access for the implementation. Implementation should already has all neccessary properties: be fast for appending, be fast for inserting or be unmodifiable, e.t.c.
What's bad:
I've read that all raw types will be restricted in future Java versions, so such code better write this way:
List<?> trololo = new ListImpl<?>();
In general wildcard has the same meaning: you don't know fo sure will your collection be heterogenous or homogeneous?
Someday you could do:
List<something> numbers = new LinkedList<something>();without changing client code which calls numbers.
Declaring interface instead of implementation is indeed the rather good and widespread practice, but it is not always the best way. Use it everytime except for all of the following conditions are true:
You are completely sure, that chosen implementation will satisfy your needs.
You need some implementation-specific feauture, that is not available through interface, e.g. ArrayList.trimToSize()
Of course, you may use casting, but then using interface makes no sense at all.
The first line is old style Java, we had to do it before Java 1.5 introduced generics. But a lot of brilliant software engineers are still forced to use Java 1.4 (or less), because their companies fear risk and effort to upgrade the applications...
OK, that was off the records. A lot of legacy code has been produced with java 1.4 or less and has not been refactored.
The second line includes generics (so it's clearly 1.5+) and the variable is declared as an ArrayList. There's actually no big problem. Sure, always better to code against interfaces, so to my (and others) opinion, don't declare a variable as ArrayList unless you really need the special ArrayList methods.
Most of the time, when you don't care about the implementation, it's better to program to interface. So, something like:
List<something> numbers = new ArrayList<something>();
would be preferred than:
ArrayList<something> numbers = new ArrayList<something>();
The reason is you can tweak your program later for performance reason.
But, you have to be careful not to just choose the most generic interface available. For example, if you want to have a sorted set, instead of to Set, you should program to SortedSet, like this:
SortedSet<something> s = new TreeSet<something>();
If you just blatantly use interface like this:
Set<something> s = new TreeSet<something>();
Someone can modify the implementation to HashSet and your program will be broken.
Lastly, this program to interface will even be much more useful when you define a public API.
Two differences are that numbers in the first line is of type List, not ArrayList. This is possible because ArrayList is a descendant of List; that is, it has everything that List has, so can fill in for a List object. (This doesn't work the other way around.)
The second line's ArrayList is typed. This means that the second numbers list can only hold type something objects.
Ok so I know that Set, List and Map are interfaces but what makes the first line of code any better than the second line?
List myArr = new ArrayList();
ArrayList myArr = new ArrayList();
If you use the first form, you are saying all you are ever going to use is the functionality of the List interface - nothing else, especially nothing extra added by any implementation of it. This means you can easily change the implementation used (e.g. just substitute LinkedList for ArrayList in the instantiation), and not worry about it breaking the rest of the code because you might have used something specific to ArrayList.
A useful general principle about types in programming (sometime referred to as the robustness principle) is as follows:
Be liberal about what you accept
Be conservative about what you emit
List is more liberal than ArrayList, since List can be any kind of List implementation e.g. an ArrayList, a LinkedList or FrancosSpecialList. Hence it is a good idea to be liberal and accept any kind of list since you may want to change the implementation later.
The main reason to use ArrayList explicitly as a type (your second case) is if you need to use methods that are specific to ArrayList that are not available through the List interface. In this case a generic List won't work (unless you want to do lots of ugly and confusing casting), so you might as well be explicit and use an ArrayList directly. This has the added bonus of hinting to a reader that specific features of ArrayList are needed.
As you can see from the source of ArrayList here, most of the methods implemented are annotated as #override because all of them that are defined through List interface so, if you are gonna use just basic functionalities (that is what you are gonna do most of the time) the difference won't be any practical one.
The difference will come if someday you will think that the features of the ArrayList are not suitable anymore for your kind of problem and you will need something different (a LinkedList for example). If you declared everything as List but instantiated as ArrayList you will easily switch to new implementation by changing the instantiations to new LinkedList() while in other case you will have to change also all variable declarations.
Using List list = new ArrayList() is more OOP style since you declare that you don't care about the specific implementation of the list, and that you want to discard the static information about the type since you will rely on the interface provided by this kind of collection abstracting from its implementation.
So for starters lets say that I have a LinkedList<String>,
I can easily convert it to an array via toArray(). i.e.
LinkedList<String> strList = new LinkedList<String>();
String[] strArray = strList.toArray(new String[0]);
But Lets say I have a LinkedList<T>
Then I the following code:
LinkedList<T> tList = new LinkedList<T>();
T[] strArray = tList.toArray(new T[0]);
I get the Cannot create a generic array of T error message.
How can I get around this?
Specifically in my class I have LinkedList<AbstractNode<T>> nodes, and I am trying to implement a getAll() method that returns all the nodes as an Array.
Thanks!
Note Péter Török's answer provides the correct answer to my problem, but for me simply returning an ArrayList instead of [] as Bar mentioned, ended up smoothing my code out a lot.
Note2 after looking at my code a bit more i'm not even sure if any conversion was necessary to begin with, LinkedList was fine for what I was trying to do... :-/
A workaround which is used in the class library would be (using an unchecked cast)
public <T> T[] toArray(List<T> list, T[] a) {
if (a.length < list.size()) {
a = (T[])java.lang.reflect.Array.newInstance(
a.getClass().getComponentType(), list.size()));
}
return list.toArray(a);
}
or, with a Class parameter instead of an array:
public <T> T[] toArray(List<T> list, Class<T> k) {
return list.toArray(
(T[])java.lang.reflect.Array.newInstance(k, list.size()));
}
From Java Generics and Collections, chapters 6.4-5.
Inability to create generic arrays is one of the most serious restrictions in Java. Because it is so annoying, it is worth reiterating the reason it occurs: generic arrays are problematic because generics are implemented via erasure, but erasure is beneficial because it eases evolution.
The best workaround is to use ArrayList or some other class from the Collections Framework in preference to an array. We discussed the tradeoffs between collection classes and arrays in Section 2.5, and we noted that in many cases collections are preferable to arrays: because they catch more errors at compile time, because they provide more operations, and because they offer more flexibility in representation. By far, the best solution to the problems offered by arrays is to "just say no": use collections in preference to arrays.
Sometimes this won't work, because you need an array for reasons of compatibility or efficiency. Examples of this occur in the Collections Framework: for compatibility, the method toArray converts a collection to an array [...]
[...] a naïve method to convert a collection to an array will not work. The first fix we might try is to add an unchecked cast, but we will see shortly that this leads to even more perplexing problems. The correct fix will require us to resort to reflection.
Since Java generic is really replaceing your T with type object and casting to concrete type resolved at compilation, there is possible to create List<T>, but not T[]. (First one will be list of objects, second one is unknown)
Well reflection allows you to do many ugly hacks. You can use them. Personally I use them only if there is no other ways to do something. I don't like runtime errors.
But my question is: Do you REALLY needs T[], why don't you use ArrayList<T> instead?