Changes

24,907 bytes added ,  08:27, 6 May 2021
Rewrite Capability documentation to be better in documenting stuff
Line 1: Line 1: −
Capabilities allow exposing features in a dynamic and flexible way, without having to resort to directly implementing many interfaces.
+
'''Capabilities''' are a Forge system that allows cross-mod interactions by allowing capability ''providers'' to
 +
dynamically respect contracts and provide specialized behavior without requiring the implementation of many interfaces
 +
or hard dependencies on mods.
   −
In general terms, each capability provides a feature in the form of an interface, alongside with a default implementation which can be requested, and a storage handler for at least this default implementation. The storage handler can support other implementations, but this is up to the capability implementor, so look it up in their documentation before trying to use the default storage with non-default implementations.
+
== History ==
 +
In an ideal world, all that would be needed for a mod to provide the equivalent of a capability would be implementing an
 +
interface. This is in fact how cross-mod interaction used to work prior to the introduction of capabilities.
   −
Forge adds capability support to <code>TileEntities</code>, <code>Entities</code>, <code>ItemStack</code>s, <code>World</code>s and <code>Chunk</code>s, which can be exposed either by attaching them through an event or by overriding the capability methods in your own implementations of the objects. This will be explained in more detail in the following sections.
+
The real world, though, is often much more complicated: users wanted to be free to combine mods the way they wanted and
 +
saw fit, and developers wanted to be able to declare ''soft'' dependencies on other mods, thus reducing the need of
 +
having a huge mod pack just for testing.
   −
== Forge-provided Capabilities ==
+
The first approach used by Forge was conditional stripping of interfaces and methods, but this proved to be problematic.
Forge provides three capabilities: <code>IItemHandler</code>, <code>IFluidHandler</code> and <code>IEnergyStorage</code>.
+
While the idea works well in theory, in practice the ASM editing of classes relied on complex mechanics and could lead
 +
to hard to spot bugs.
   −
<code>IItemHandler</code> exposes an interface for handling inventory slots. It can be applied to <code>TileEntities</code> (chests, machines, etc.), <code>Entities</code> (extra player slots, mob/creature inventories/bags), or <code>ItemStacks</code> (portable backpacks and such). It replaces the old <code>IInventory</code> and <code>ISidedInventory</code> with an automation-friendly system.
+
For this reason, the entire system was redesigned and the concept of '''capabilities''' was born.
   −
<code>IFluidHandler</code> exposes an interface for handling fluid inventories. It can also be applied to <code>TileEntities</code> <code>Entities</code>, or <code>ItemStacks</code>. It replaces the old <code>IFluidHandler</code> with a more consistent and automation-friendly system.
+
== The Concept ==
 +
A capability allows any capability provider to conditionally expose a certain ability to do something, e.g. accepting
 +
power or handling items. A capability provider, moreover, can decide to expose a capability only on certain sides,
 +
allowing for easy interactions with hoppers, cables, etc.
   −
<code>IEnergyStorage</code> exposes an interface for handling energy containers. It can be applied to <code>TileEntities</code>, <code>Entities</code> or <code>ItemStacks</code>. It is based on the RedstoneFlux API by TeamCoFH.
+
Capabilities may also be added and removed dynamically both from the "owner" of the capability provider and other mods,
 +
allowing even easier cross-mod interaction. For example, a mod that isn't compatible with Forge Energy could be
 +
converted into one by dynamically attaching the Forge Energy capability and handling the conversion to a third-party
 +
energy system without having to alter the original mod.
   −
== Using an Existing Capability ==
+
== Terminology ==
As mentioned earlier, <code>TileEntities</code>, <code>Entities</code>, and <code>ItemStacks</code> implement the capability provider feature, through the <code>ICapabilityProvider</code> interface. This interface adds the method <code>getCapability</code>, which can be used to query the capabilities present in the objects.
+
The high flexibility of the system comes with a cost, though, which is terminology. The following section wants to be a
 +
dictionary of sorts, defining all the terms that you may come across when dealing with capabilities.
 +
 
 +
In the rest of this article, we will refer to these terms frequently, so make sure you are familiar with them.
 +
 
 +
; Capability
 +
: the ability to perform something. In-code this is represented by the <code>Capability</code> class.
 +
; Capability Provider
 +
: something that is able to support capabilities and provides a mean of accessing them. In-code they are represented by implementations of <code>ICapabilityProvider</code>. There are multiple kinds of capability providers:
 +
:; Volatile Provider
 +
:: a provider that doesn't persist data to disk; once the provider ceases to exist for any number of reasons, all capability data gets deleted.
 +
:; Persistent Provider
 +
:: a provider that requires all capabilities to serialize data to disk, in order to persist data even across game restarts. They implement the <code>INBTSerializable</code> interface.
 +
:; Agnostic Provider
 +
:: a provider that isn't neither volatile nor persistent, rather delegates the decision either to the capability directly or to sub-implementations. They also implement the <code>INBTSerializable</code> interface.
 +
; Capability Interface
 +
: the interface that defines the contract of the capability, so what operations the capability exposes.
 +
; Capability Implementation
 +
: one of the possibly many implementations of the capability interface, that actually carries out the work; one of the various implementations may also be considered the '''default capability implementation'''.
 +
; Capability Storage
 +
: the manager that handles loading and storing persistent capabilities data from and to disk, guaranteeing preservation of information; in-code this is represented by an implementation of the <code>Capability.IStorage</code> interface.
 +
 
 +
The wary reader may note that both ''persistent'' and ''agnostic'' providers are represented the same way in code. In
 +
fact, the only difference between them comes from pure semantics in how their serialization methods are designed. This
 +
will be further discussed in their respective sections.
 +
 
 +
Moreover, it is also common to refer to the capability interface as simply the ''capability''. While not strictly
 +
correct, due to common usage we will also use this convention. So, to refer to the capability interface
 +
<code>MyCapability</code>, we will usually talk about the "<code>MyCapability</code> capability".
 +
 
 +
== Forge-provided Capabilities and Providers ==
 +
In order to ensure mods can work together seamlessly, Forge provides a set of default capabilities and capability
 +
providers.
 +
 
 +
The default capability providers in a Forge environment are: <code>TileEntity</code>, <code>Entity</code>,
 +
<code>ItemStack</code>, <code>World</code>, and <code>Chunk</code>. These are all agnostic providers, since they don't
 +
mandate any sort of capability persistency requirements. Rather, it is the job of whoever subclasses these providers to
 +
deal with either volatile or non-volatile capabilities.
 +
 
 +
The default capabilities that forge provides are represented by the interfaces <code>IItemHandler</code>,
 +
<code>IFluidHandler</code>, <code>IFluidHandlerItem</code>, <code>IEnergyStorage</code>, and
 +
<code>IAnimationStateMachine</code>. Each one of these capabilities will be discussed in the corresponding section.
 +
 
 +
=== <tt>IItemHandler</tt> ===
 +
 
 +
The <code>IItemHandler</code> capability refers to the ability for any capability provider to have some sort of internal
 +
'''inventory''' with a certain number of slots, from which items can be inserted and extracted. It is also possible,
 +
though, to expose this capability even if no such inventory is present as long as the capability provider can emulate
 +
its presence (e.g. tools that allow accessing remote inventories).
 +
 
 +
This effectively '''replaces''' the vanilla interfaces <code>IInventory</code> and <code>ISidedInventory</code>. These
 +
interfaces are in fact retained only to allow vanilla code to compile and should not be used in mod code. This extends
 +
to anything that implements those vanilla interfaces, such as <code>LockableLootTileEntity</code>.
 +
 
 +
A default reference implementation for this capability interface is provided in <code>ItemStackHandler</code>.
 +
 
 +
=== <tt>IFluidHandler</tt> ===
 +
 
 +
The <code>IFluidHandler</code> capability refers to the ability for any capability provider to handle and store fluids
 +
in one or multiple fluid tanks. It is effectively the equivalent in terms of fluids of the <code>IItemHandler</code>
 +
capability.
 +
 
 +
A default reference implementation for this capability interface is provided in <code>TileFluidHandler</code>.
 +
 
 +
=== <tt>IFluidHandlerItem</tt> ===
 +
 
 +
The <code>IFluidHandlerItem</code> capability referes to the ability for an <code>ItemStack</code> capability provider
 +
to handle and store fluids in one or multiple fluid tanks. It is basically a specialized version of the
 +
<code>IFluidHandler</code> capability that allows <code>ItemStack</code>s to define a custom container.
 +
 
 +
=== <tt>IEnergyStorage</tt> ===
 +
 
 +
The <code>IEnergyStorage</code> capability refers to the ability for any capability provider to store, consume, and
 +
produce energy. This capability is the base capability for what's commonly known in the modded world as Forge Energy (or
 +
FE), i.e. the energy system most mods use. Its internal design is heavily based on the (now defunct) Redstone Flux
 +
Energy API, supporting both a push and pull system.
 +
 
 +
A default reference implementation for this capability interface is provided in <code>EnergyStorage</code>.
 +
 
 +
=== <tt>IAnimationStateMachine</tt> ===
 +
 
 +
The <code>IAnimationStateMachine</code> capability refers to the ability for any capability provider to leverage the
 +
Forge Animation State Machine API for animations.
 +
 
 +
== Working with Capabilities ==
 +
 
 +
Both capability providers and users need to be able to provide and access capabilities through a common framework,
 +
otherwise the ideal of dynamic and mod-agnostic would not really exist. For this reason, both capability providers and
 +
capability ''accessors'' (which we define as everything that wants to access a capability), also known as '''clients''' or '''users''',
 +
need to work together and with Forge to ensure that the common interface is used sensibly and correctly by all parties.
 +
 
 +
=== Obtaining a Capability ===
 +
 
 +
Before being able to work with a capability, it is necessary to obtain an instance of the <code>Capability</code> object
 +
itself. Since these objects are created by Forge and there is only '''one''' unique instance for each capability that
 +
may exist, this instance cannot be obtained by "common" means. Forge provides two different methods of obtaining such
 +
instances: '''injecting''' into a field, or a '''callback method'''.
 +
 
 +
==== Injecting into a Field ====
 +
 
 +
A <code>Capability</code> can be injected automatically into a field as soon as it gets created by Forge, following the
 +
principle commonly known as '''dependency injection'''. This provides less flexibility, since it doesn't notify the user
 +
that the capability has been injected nor runs arbitrary code. Nevertheless, it is '''suggested''' to use this method
 +
instead of the callback approach.
 +
 
 +
To inject the <code>Capability</code> into a field, all that's needed is to declare a <code>static</code> field of type
 +
<code>Capability&lt;T&gt;</code>, where <code>T</code> represents the capability interface, and annotate it with
 +
<code>@CapabilityInject(T.class)</code>.
 +
 
 +
For a more practical example, consider the following snippet:
   −
In order to obtain a capability, you will need to refer it by its unique instance. In the case of the <code>IItemHandler</code>, this capability is primarily stored in <code><nowiki>CapabilityItemHandler#ITEM_HANDLER_CAPABILITY</nowiki></code>, but it is possible to get other instance references by using the <code>@CapabilityInject</code> annotation.
   
<syntaxhighlight lang="java">
 
<syntaxhighlight lang="java">
 
@CapabilityInject(IItemHandler.class)
 
@CapabilityInject(IItemHandler.class)
static Capability<IItemHandler> ITEM_HANDLER_CAPABILITY = null;
+
public static Capability<IItemHandler> ITEM_HANDLER_CAPABILITY = null;
 
</syntaxhighlight>
 
</syntaxhighlight>
This annotation can be applied to fields and methods. When applied to a field, it will assign the instance of the capability (the same one gets assigned to all fields) upon registration of the capability, and left to the existing value (<code>null</code>), if the capability was never registered. Because local static field accesses are fast, it is a good idea to keep your own local copy of the reference for objects that work with capabilities. This annotation can also be used on a method, in order to get notified when a capability is registered, so that certain features can be enabled conditionally.
     −
Both the <code>getCapability</code> methods have a second parameter, of type <code>Direction</code>, which can be used in the to request the specific instance for that one face. If passed <code>null</code>, it can be assumed that the request comes either from within the block, or from some place where the side has no meaning, such as a different dimension. In this case a general capability instance that does not care about sides will be requested instead. The return type of <code>getCapability</code> will correspond to the type declared in the capability passed to the method. For the item handler capability, this is indeed <code>IItemHandler</code>.
+
The above code will let Forge know that the field <code>ITEM_HANDLER_CAPABILITY</code> should be injected with the
 +
unique instance of the <code>IItemHandler</code> capability. Assigning the field to <code>null</code> allows us to
 +
provide a reasonable fallback in case the capability we want hasn't been registered yet.
 +
 
 +
This injection is, for obvious reasons, redundant, since that capability is also available through
 +
<code>CapabilityItemHandler</code>.
   −
==Exposing a Capability==
+
==== Declaring a Callback ====
In order to expose a capability, you will first need an instance of the underlying capability type. Note that you should assign a separate instance to each object that keeps the capability, since the capability will most probably be tied to the containing object.
     −
There’s two ways to obtain such an instance, through the <code>Capability</code> itself, or by explicitly instantiating an implementation of it. The first method is designed to use a default implementation via <code>Capability#getDefaultInstance</code>, if those default values are useful for you. In the case of the item handler capability, the default implementation will expose a single slot inventory, which is most probably not what you want.
+
Another option is to declare a callback method, meaning a method that will be called with the value of the desired
 +
<code>Capability</code> once the instance is available. This gives more flexibility since the method may perform a
 +
number of arbitrary actions with the received instance prior to storing it in a field, or may even discard the
 +
capability entirely if wanted. Nevertheless, the usage of a field instead of a method is encouraged as a matter of
 +
style.
   −
The second method can be used to provide custom implementations. In the case of <code>IItemHandler</code>, the default implementation uses the <code>ItemStackHandler</code> class, which has an optional argument in the constructor, to specify a number of slots. However, relying on the existence of these default implementations should be avoided, as the purpose of the capability system is to prevent loading errors in contexts where the capability is not present, so instantiation should be protected behind a check testing if the capability has been registered (see the remarks about <code>@CapabilityInject</code> in the previous section).
+
To use a method as a callback, the method must be declared as <code>static</code> and accepting a single parameter of
 +
type <code>Capability&lt;T&gt;</code>, where <code>T</code> represents the capability interface. The method should also
 +
be annotated with <code>@CapabilityInject(T.class)</code>.
   −
Once you have your own instance of the capability interface, you will want to notify users of the capability system that you expose this capability and provide a holder of the instance. This is done by overriding the <code>getCapability</code> method, and comparing the instance with the capability you are exposing. If your machine has different slots based on which side is being queried, you can test this with the <code>side</code> parameter. For <code>Entities</code> and <code>ItemStack</code>s, this parameter can be ignored, but it is still possible to have side as a context, such as different armor slots on a player (top side => head slot?), or about the surrounding blocks in the inventory (west => slot on the left?). Don’t forget to fall back to <code>super</code>, otherwise the attached capabilities will stop working. Make sure to invalidate the holder of the instance at the end of the provider's lifecycle.
+
For a more practical example, consider the following snippet:
    
<syntaxhighlight lang="java">
 
<syntaxhighlight lang="java">
// Somewhere in your TileEntity subclass
+
public static Capability<IEnergyStorage> ENERGY = null;
LazyOptional<IItemHandler> inventoryHandlerLazyOptional;
+
 
 +
@CapabilityInject(IEnergyStorage.class)
 +
private static void onEnergyStorageInit(Capability<IEnergyStorage> capability) {
 +
    LOGGER.info("Received IEnergyStorage capability '{}': enabling Forge Energy support", capability);
 +
    ENERGY = capability;
 +
}
 +
</syntaxhighlight>
 +
 
 +
The above code declares a callback method that will be invoked when a <code>Capability</code> instance for
 +
<code>IEnergyStorage</code> is available. The callback then prints a log message and stores the capability into a
 +
<code>public</code> field for accessibility. The field is initialized to <code>null</code> to provide a reasonable
 +
fallback in case the capability does not exist.
   −
// After initializing inventoryHandler
+
This callback is, for obvious reasons, redundant, since that capability is also available through
inventoryHandlerLazyOptional = LazyOptional.of(() -> inventoryHandler);
+
<code>CapabilityEnergy</code>.
   −
public <T> LazyOptional<T> getCapability(Capability<T> cap, Direction side) {
+
=== Exposing a Capability ===
  if (cap == CapabilityItemHandler.ITEM_HANDLER_CAPABILITY) {
+
 
    return inventoryHandlerLazyOptional.cast();
+
Exposing a capability is a voluntary act by a capability provider that allows the capability to be discovered and
  }
+
accessed by users.
  return super.getCapability(cap, side);
+
 
 +
To do so, a capability provider needs to juggle a couple more moving pieces to ensure that the capability state remains
 +
consistent and that the lookup remains fast. It is in fact possible for a capability provider to be asked to provide
 +
many capabilities many times in the same tick. For this reason, a provider is asked to do the following:
 +
 
 +
* the <code>LazyOptional</code>s that get returned '''must be cached''';
 +
* if a capability changes exposure state (more on this later), all listeners '''must be notified''';
 +
* if a capability gets invalidated (more on this later), all listeners '''must be notified'''
 +
* the lookup inside <code>getCapability</code> must be performed with '''an <code>if</code>-<code>else</code> chain''';
 +
* all unexposed but still present capabilities '''should be available''' if the provider is queried with a <code>null</code> direction (see ''Accessing a Capability'' for more information);
 +
* if no capability of a given type is available or accessible, the provider '''must call <code>super</code> as long as it is possible to do so'''.
 +
 
 +
Capability providers must also reflect changes in the ''exposure state'' of a capability, meaning that if the
 +
accessibility of a capability from a certain <code>Direction</code> changes (refer to
 +
[[#Accessing a Capability|Accessing a Capability]] for more information), it is the provider's responsibility to trigger
 +
a state response by invalidating the returned <code>LazyOptional</code> and caching a new one. This should also be
 +
performed when a capability gets ''invalidated'', such as when a capability provider gets removed.
 +
 
 +
With all of the above in mind, part of a capability provider implementation may be similar to the following snippet:
 +
 
 +
<syntaxhighlight lang="java">
 +
// suppose the presence of a field 'inventory' of type 'IItemHandler'
 +
 
 +
private final LazyOptional<IItemhandler> inventoryOptional = LazyOptional.of(() -> this.inventory);
 +
 
 +
@Override
 +
public <T> LazyOptional<T> getCapability(Capability<T> capability, @Nullable Direction direction) {
 +
    if (capability == CapabilityItemHandler.ITEM_HANDLER_CAPABILITY
 +
            && (direction == null || direction == Direction.UP || direction == Direction.DOWN)) {
 +
        return this.inventoryOptional.cast();
 +
    }
 +
    return super.getCapability(capability, direction); // See note after snippet
 
}
 
}
    
@Override
 
@Override
 
protected void invalidateCaps() {
 
protected void invalidateCaps() {
  super.invalidateCaps();
+
    super.invalidateCaps();
  inventoryHandlerLazyOptional.invalidate();
+
    this.inventoryOptional.invalidate();
 +
}
 +
</syntaxhighlight>
 +
 
 +
This possible implementation of a capability provider exposes an <code>IItemHandler</code> capability and restricts
 +
access only to the <code>UP</code> and <code>DOWN</code> directions. If we assume this capability provider is a
 +
<code>TileEntity</code>, then we may also say that the inventory is only accessible from the top and the bottom of the
 +
block.
 +
 
 +
Moreover, the capability gets automatically invalidated when the provider gets invalidated. Assuming this is a
 +
<code>TileEntity</code>, this usually happens when the block gets removed from the world or unloaded due to distance.
 +
 
 +
The <code>super</code> call at the end of the <code>getCapability</code> method is extremely important, since it's what
 +
allows Attaching external Capabilities to capability providers. Nevertheless, it is not always possible to invoke
 +
<code>super</code>: in those cases, an empty <code>LazyOptional</code> should be returned.
 +
 
 +
=== Attaching a Capability ===
 +
 
 +
Attaching a Capability is a process by which external agents "modify" a Capability Provider, making it expose additional
 +
capabilities other than the already available ones.
 +
 
 +
To do so, the '''attaching agent''' (which means the thing that wants to attach a capability to another provider) must
 +
listen to the <code>AttachCapabilitiesEvent&lt;T&gt;</code>. The <code>T</code> in this case represents the capability
 +
provider you want to attach the capability to. Note that the type of <code>T</code> '''must''' be the base type of the
 +
capability provider, not a subclass. As an example, if you want to attach a capability to a <code>MyTileEntity</code>,
 +
which extends <code>TileEntity</code>, you'll have to listen to <code>AttachCapabilitiesEvent&lt;TileEntity&gt;</code>,
 +
'''NOT''' to <code>AttachCapabilitiesEvent&lt;MyTileEntity&gt;</code>, since the latter will never fire.
 +
 
 +
The attaching agent can use the provided methods <code>getObject</code>, <code>addCapability</code>, and
 +
<code>addListener</code> to check whether the capability should be attached to the current object and perform the
 +
desired action.
 +
 
 +
When attaching a capability, the attaching agent should also provide a name in the form of a
 +
<code>ResourceLocation</code>. The name '''must''' be under the attaching agent's namespace, but no restrictions are
 +
placed on the actual name, as long as it is unique inside the given namespace.
 +
 
 +
Maybe a little counter-intuitively, the process of attaching does not attach a capability nor a capability interface
 +
directly. Rather, the attaching agent should create its own implementation of a '''Capability Provider''' and attach it
 +
via the event. This is done so that the attaching agent can have control over when, how, and where its capabilities are
 +
exposed, instead of relying on the game itself deciding these parameters. For this reason, all considerations given in
 +
the [[#Exposing a Capability|Exposing a Capability]] section on how to correctly create a Capability Provider.
 +
 
 +
With the above in mind, part of an attaching agent may be similar to the following snippet of code:
 +
 
 +
<syntaxhighlight lang="java">
 +
@SubscribeEvent
 +
public void onAttachingCapabilities(final AttachCapabilitiesEvent<TileEntity> event) {
 +
    if (!(event.getObject() instanceof EnergyBasedTileEntity)) return;
 +
 
 +
    EnergyStorage backend = new EnergyStorage(((EnergyBasedTileEntity) event.getObject()).capacity);
 +
    LazyOptional<IEnergyStorage> optionalStorage = LazyOptional.of(() -> backend);
 +
 
 +
    ICapabilityProvider provider = new ICapabilityProvider() {
 +
        @Override
 +
        public <T> LazyOptional<T> getCapability(Capability<T> cap, @Nullable Direction direction) {
 +
            if (cap == CapabilityEnergy.ENERGY) {
 +
                return optionalStorage.cast();
 +
            }
 +
            return LazyOptional.empty();
 +
        }
 +
    };
 +
 
 +
    event.addCapability(new ResourceLocation("examplemod", "fe_compatibility"), provider);
 +
    event.addListener(optionalStorage::invalidate);
 
}
 
}
 
</syntaxhighlight>
 
</syntaxhighlight>
   −
<code>Item</code>s are a special case since their capability providers are stored on an <code>ItemStack</code>. Instead, a provider should be attached through <code>Item#initCapabilities</code> when applicable. This should hold your capabilities for the lifecycle of the stack.
+
This example implementation of an attaching agent attaches a <code>IEnergyStorage</code> capability to all
 +
<code>TileEntity</code> instance that are a subclass of <code>EnergyBasedTileEntity</code>. It also sets up the
 +
<code>LazyOptional</code> for invalidation if the parent capability provider gets invalidated.
 +
 
 +
Note also the call of <code>LazyOptional.empty()</code> rather than a <code>super</code>. This is needed because when
 +
attaching a capability, the parent capability provider isn't known. For this reason, it is necessary to return an empty
 +
<code>LazyOptional</code>. The game will then handle automatic merging of the various different providers into a single
 +
one.
 +
 
 +
The above example is one of a '''Volatile''' Capability Provider. On the other hand, mods may want to persist their
 +
data across sessions. In this case, they should attach a '''Persistent''' Capability Provider: this can be done either
 +
by implementing the <code>INBTSerializable</code> interface along with <code>ICapabilityProvider</code> or by
 +
implementing the <code>ICapabilitySerializable</code> interface.
 +
 
 +
The previous example reworked to use a Persistent Capability Provider may be similar to the following snippet:
 +
 
 +
<syntaxhighlight lang="java">
 +
@SubscribeEvent
 +
public void onAttachingCapabilities(final AttachCapabilitiesEvent<TileEntity> event) {
 +
    if (!(event.getObject() instanceof EnergyBasedTileEntity)) return;
 +
 
 +
    EnergyStorage backend = new EnergyStorage(((EnergyBasedTileEntity) event.getObject()).capacity);
 +
    LazyOptional<IEnergyStorage> optionalStorage = LazyOptional.of(() -> backend);
 +
    Capability<IEnergyStorage> capability = CapabilityEnergy.ENERGY;
 +
 
 +
    ICapabilityProvider provider = new ICapabilitySerializable<IntNBT>() {
 +
        @Override
 +
        public <T> LazyOptional<T> getCapability(Capability<T> cap, @Nullable Direction direction) {
 +
            if (cap == capability) {
 +
                return optionalStorage.cast();
 +
            }
 +
            return LazyOptional.empty();
 +
        }
 +
 
 +
        @Override
 +
        public IntNBT serializeNBT() {
 +
            return capability.getStorage().writeNbt(capability, backend, null);
 +
        }
 +
 
 +
        @Override
 +
        public void deserializeNBT(IntNBT nbt) {
 +
            capability.getStorage().readNBT(capability, backend, null, nbt);
 +
        }
 +
    };
 +
 
 +
    event.addCapabilities(new ResourceLocation("examplemod", "fe_compatibility"), provider);
 +
    event.addListener(optionalStorage::invalidate);
 +
}
 +
</syntaxhighlight>
 +
 
 +
 
 +
=== Accessing a Capability ===
 +
 
 +
Accessing a Capability is the process by which a user is able to '''query''' a Capability Provider for a specific
 +
instance of a capability.
 +
 
 +
This is perhaps the second most important part of the entire capability system, since it is what allows cross-mod
 +
interaction. To obtain an instance of a Capability, the user must first get a hold of the Capability Provider that
 +
should be queried. This can be done in a variety of ways and is outside the scope of this article. The user should
 +
then invoke the <code>getCapability</code> method passing the unique instance of the capability that should be queried
 +
(see [[#Obtaining a Capability|Obtaining a Capability]] for more information) and the querying <code>Direction</code>,
 +
if applicable.
 +
 
 +
The returned object is a <code>LazyOptional</code> wrapping the queried Capability, if the capability provider exposes
 +
it, otherwise it will be empty. The <code>LazyOptional</code> can be either unwrapped via an <code>orElse</code> or
 +
used directly via <code>ifPresent</code>.
 +
 
 +
It is '''highly suggested''' to cache the returned <code>LazyOptional</code> to avoid querying the same provider every
 +
time, in order to improve performance. The user should thus register itself to the invalidation listener via the
 +
<code>addListener</code> method. This ensures that the user will be able to react to the invalidation of the
 +
<code>LazyOptional</code> and remove it from the cache.
 +
 
 +
With the above in mind, part of an user may be similar to the following snippet of code:
 +
 
 +
<syntaxhighlight lang="java">
 +
private final Map<Direction, LazyOptional<IEnergyStorage>> cache = new HashMap<>();
 +
 
 +
private void sendPowerTo(int power, Direction direction) {
 +
    LazyOptional<IEnergyStorage> targetCapability = cache.get(direction);
 +
 
 +
    if (targetCapability == null) {
 +
        ICapabilityProvider provider = world.getTileEntity(pos.offset(direction));
 +
        targetCapability = provider.getCapability(CapabilityEnergy.ENERGY, direction.getOpposite());
 +
        cache.put(direction, targetCapability);
 +
        targetCapability.addListener(self -> cache.put(direction, null));
 +
    }
 +
 
 +
    targetCapability.ifPresent(storage -> storage.receiveEnergy(power, false));
 +
}
 +
</syntaxhighlight>
 +
 
 +
This example implementation of an user is querying via a <code>TileEntity</code> the neighboring capability provider
 +
for an <code>IEnergyStorage</code> capability. Before obtaining the provider, the code performs a cache lookup for the
 +
targeted capability. If the check succeeds, then no lookup is performed; if the check fails, the targeted Capability
 +
Provider is obtained and queried for the Capability. The obtained <code>LazyOptional</code> is then cached and a
 +
listener is attached to it so that the cache would be emptied on invalidation. The code then continues with the
 +
interaction with the capability, which is outside the scope of this article.
 +
 
 +
== Creating Custom Capabilities ==
 +
 
 +
While the various capabilities provided by Forge may satisfy the most common use cases, there is always the chance that
 +
a mod may require a custom solution. For this reason, Forge provides a way to define a custom Capability.
 +
 
 +
Defining a custom Capability requires the user to provide three main components: the Capability Interface, at least one
 +
Capability Implementation, and the Capability Storage. Optionally, a Capability Provider can also be created. In this
 +
case, the provider will be used as described in [[#Attaching a Capability|Attaching a Capability]]. The various details
 +
for all these components are described in the respective sections of this article.
 +
 
 +
In this section, we will refer to the implementation of a <code>MyCapability</code> capability, that can be used to
 +
store a single mutable <code>String</code>.
 +
 
 +
Refer also to [[#Code Examples|Code Examples]] for an example on how the various components may be implemented in a
 +
real-world scenario.
 +
 
 +
=== The Capability Interface and the Capability Implementation ===
 +
 
 +
The Capability Interface is one of the most important parts of a Capability: without it, the Capability effectively does
 +
not exist. Designing a Capability Interface is exactly like designing any Java interface, so the particular details will
 +
be glossed over in this section.
 +
 
 +
The Capability Implementation, on the other hand, is the implementation of the previously defined Capability Interface.
 +
Usual rules for interface implementations follow. There can be more than one Capability Implementation for each
 +
capability, but no less than one.
 +
 
 +
Note that a '''well-formed''' capability implementation should '''not store''' the Capability Provider inside of it: we
 +
call the capability implementation ''provider-agnostic''. This is not a hard-requirement, though, rather it should act
 +
more as a guideline. There are in fact certain situations where this cannot be avoided (e.g. attaching a client-synced
 +
capability to an <code>ItemStack</code>).
 +
 
 +
One of the various Capability Implementation should also act as the default implementation. Other mods can ask the
 +
capability to create an instance of the default implementation without ever referring to such an implementation
 +
themselves. This guarantees separation of API code from implementation code, which is also one of the goals of the
 +
capability system.
 +
 
 +
Given all of the above information, this may be an example implementation of both a Capability Interface and a
 +
Capability Implementation:
 +
 
 +
<syntaxhighlight lang="java">
 +
public interface MyCapability {
 +
    String getValue();
 +
    void setValue(String value);
 +
}
 +
 
 +
public class MyCapabilityImplementation implements MyCapability {
 +
    private String value;
 +
 
 +
    @Override
 +
    public String getValue() {
 +
        return this.value;
 +
    }
 +
 
 +
    @Override
 +
    public void setValue(String value) {
 +
        this.value = value;
 +
    }
 +
}
 +
</syntaxhighlight>
 +
 
 +
Note that in this case, only a single implementation is provided, which will also act as the default implementation for
 +
the <code>MyCapability</code> capability.
 +
 
 +
=== The Capability Storage ===
 +
 
 +
The Capability Storage is that component of the Capability System that is responsible for serializing and deserializing
 +
a capability. All capabilities must provide one, since certain providers may or may not require that their capabilities
 +
are serializable.
 +
 
 +
The Capability Storage implements the <code>Capability.IStorage&lt;T&gt;</code> interface, where <code>T</code> is the
 +
generic type of the Capability Interface the storage refers to. Each capability must have '''one and exactly one'''
 +
Capability Storage.
 +
 
 +
The Storage is usually called by the Capability Provider when serialization or deserialization needs to happen. The
 +
Storage is then responsible of reading the data from the given capability instance and convert that into an NBT-based
 +
structure that can be serialized. A Storage may also return <code>null</code> to indicate that no serialization is
 +
necessary, although some providers may require an empty tag to be supplied instead. At the same time, the Storage is
 +
also responsible for restoring the original state of the capability when deserialization happens. In this case, the
 +
given NBT structure is guaranteed not to be <code>null</code>.
 +
 
 +
In all cases, a <code>Direction</code> is provided for context, if available.
 +
 
 +
Although discouraged as a matter of code cleanliness, it is legal for a Capability Storage to require a specific
 +
Capability Implementation for the serialization and deserialization to be successful. If this is the case, this
 +
requirement '''must''' be documented, though the code should be refactored wherever possible to remove this requirement.
 +
 
 +
Given the above information, an example of an implementation of a Capability Storage may be the following:
 +
 
 +
<syntaxhighlight lang="java">
 +
public class MyCapabilityStorage implements Capability.IStorage<MyCapability> {
 +
    @Override
 +
    @Nullable
 +
    INBT writeNBT(Capability<MyCapability> capability, MyCapability instance, Direction direction) {
 +
        return StringNBT.valueOf(instance.getValue());
 +
    }
 +
 
 +
    @Override
 +
    public void readNBT(Capability<MyCapability> capability, MyCapability instance, Direction direction, INBT nbtData) {
 +
        if (!(nbtData instanceof StringNBT)) {
 +
            throw new IllegalArgumentException("Unable to deserialize 'MyCapability' from a non-String NBT structure");
 +
        }
 +
        instance.setValue(((StringNBT) nbtData).getString());
 +
    }
 +
}
 +
</syntaxhighlight>
 +
 
 +
Note the <code>instanceof</code> check needed to ensure that the given <code>INBT</code> instance is valid for
 +
deserialization. A Capability Storage should always perform this check prior to the cast in order to provide a
 +
meaningful error message, rather than a cryptic <code>ClassCastException</code>.
 +
 
 +
=== The Capability Provider ===
 +
 
 +
The Capability Provider is an optional component of the capability that allows the Capability to be attached to a
 +
component. The details on how a Capability Provider should behave have already been discussed in the two previous
 +
sections [[#Exposing a Capability|Exposing a Capability]] and [[#Attaching a Capability|Attaching a Capability]]: refer
 +
to those for more information.
 +
 
 +
=== Tying it All Together ===
 +
 
 +
Once all components of a Capability have been created, they must be registered so that the game is aware of the
 +
capability's presence. The registration requires specifying the Capability Interface, the Capability Storage, and a
 +
factory for the default capability implementation.
 +
 
 +
The registration can be performed by calling the <code>register</code> method on the <code>CapabilityManager</code>.
 +
This '''needs''' to happen when the <code>FMLCommonSetupEvent</code> is fired on the <code>MOD</code> event bus. The
 +
registration will also automatically inject the created Capability into all relevant fields and methods: refer to
 +
[[#Obtaining a Capability|Obtaining a Capability]] for more information.
 +
 
 +
An example of registration can be found in the snippet that follows:
 +
 
 +
<syntaxhighlight lang="java">
 +
public void onCommonSetup(FMLCommonSetupEvent event) {
 +
    CapabilityManager.INSTANCE.register(MyCapability.class, new MyCapabilityStorage(), MyCapabilityImplementation::new);
 +
}
 +
</syntaxhighlight>
 +
 
 +
== Custom Capability Providers ==
 +
 
 +
Much like custom Capabilities, Forge also allows the creation of custom Capability Providers. The main advantage of this
 +
is allowing mods to create custom providers for their custom objects, in order to promote not only cross-mod
 +
compatibility, but also uniformity in the way users may interact with different mod APIs.
 +
 
 +
This section will only give the basic outline of what is necessary to implement a custom Capability Provider: for more
 +
in-depth explanation, people are referred to the game code.
 +
 
 +
By definition, a custom Capability Provider is everything that implements the <code>ICapabilityProvider</code>
 +
interface. In this section, though, we will only cover people that may want to replicate the functionality of one of
 +
the default providers, such as <code>TileEntity</code> or <code>Chunk</code>.
 +
 
 +
The easiest way of doing this is extending the <code>CapabilityProvider</code> class provided by Forge. This will
 +
automatically set up an ''agnostic'' Capability Provider. To fully initialize the capability provider, the subclass
 +
should then invoke the <code>gatherCapabilities</code> method as the last instruction in its constructor, to ensure that
 +
the game is able to recollect and attach all capabilities that other mods may want to attach to the capability provider.
 +
 
 +
== Code Examples ==
   −
It is strongly suggested that direct checks in code are used to test for capabilities instead of attempting to rely on maps or other data structures, since capability tests can be done by many objects every tick, and they need to be as fast as possible in order to avoid slowing down the game.
+
* [https://gist.github.com/TheSilkMiner/5cc92ba573e7bdd871dfdbffdd5c2806 A Gist showing a quick and dirty example on how to implement a Capability effectively]