Google Git Google Git
Sign in
third-party-mirror/openjdk/jdk25u/HEAD/./src/java.base/share/classes/java/lang/invoke/MethodHandles.java
blob: b0f7a53ea713a1834e00b4b4d97e4bb33d67ef14 [file]
/*
* Copyright (c) 2008, 2025, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License version 2 only, as
* published by the Free Software Foundation. Oracle designates this
* particular file as subject to the "Classpath" exception as provided
* by Oracle in the LICENSE file that accompanied this code.
*
* This code is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
* version 2 for more details (a copy is included in the LICENSE file that
* accompanied this code).
*
* You should have received a copy of the GNU General Public License version
* 2 along with this work; if not, write to the Free Software Foundation,
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
*
* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
* or visit www.oracle.com if you need additional information or have any
* questions.
*/
package java.lang.invoke;
import jdk.internal.access.SharedSecrets;
import jdk.internal.misc.Unsafe;
import jdk.internal.misc.VM;
import jdk.internal.reflect.CallerSensitive;
import jdk.internal.reflect.CallerSensitiveAdapter;
import jdk.internal.reflect.Reflection;
import jdk.internal.util.ClassFileDumper;
import jdk.internal.vm.annotation.ForceInline;
import jdk.internal.vm.annotation.Stable;
import sun.invoke.util.ValueConversions;
import sun.invoke.util.VerifyAccess;
import sun.invoke.util.Wrapper;
import java.lang.classfile.ClassFile;
import java.lang.classfile.ClassModel;
import java.lang.constant.ClassDesc;
import java.lang.constant.ConstantDescs;
import java.lang.invoke.LambdaForm.BasicType;
import java.lang.invoke.MethodHandleImpl.Intrinsic;
import java.lang.reflect.Constructor;
import java.lang.reflect.Field;
import java.lang.reflect.Member;
import java.lang.reflect.Method;
import java.lang.reflect.Modifier;
import java.nio.ByteOrder;
import java.security.ProtectionDomain;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.BitSet;
import java.util.Comparator;
import java.util.Iterator;
import java.util.List;
import java.util.Objects;
import java.util.Set;
import java.util.concurrent.ConcurrentHashMap;
import java.util.stream.Stream;
import static java.lang.classfile.ClassFile.*;
import static java.lang.invoke.LambdaForm.BasicType.V_TYPE;
import static java.lang.invoke.MethodHandleNatives.Constants.*;
import static java.lang.invoke.MethodHandleStatics.*;
import static java.lang.invoke.MethodType.methodType;
/**
* This class consists exclusively of static methods that operate on or return
* method handles. They fall into several categories:
* <ul>
* <li>Lookup methods which help create method handles for methods and fields.
* <li>Combinator methods, which combine or transform pre-existing method handles into new ones.
* <li>Other factory methods to create method handles that emulate other common JVM operations or control flow patterns.
* </ul>
* A lookup, combinator, or factory method will fail and throw an
* {@code IllegalArgumentException} if the created method handle's type
* would have <a href="MethodHandle.html#maxarity">too many parameters</a>.
*
* @author John Rose, JSR 292 EG
* @since 1.7
*/
public final class MethodHandles {
private MethodHandles() { } // do not instantiate
static final MemberName.Factory IMPL_NAMES = MemberName.getFactory();
// See IMPL_LOOKUP below.
//--- Method handle creation from ordinary methods.
/**
* Returns a {@link Lookup lookup object} with
* full capabilities to emulate all supported bytecode behaviors of the caller.
* These capabilities include {@linkplain Lookup#hasFullPrivilegeAccess() full privilege access} to the caller.
* Factory methods on the lookup object can create
* <a href="MethodHandleInfo.html#directmh">direct method handles</a>
* for any member that the caller has access to via bytecodes,
* including protected and private fields and methods.
* This lookup object is created by the original lookup class
* and has the {@link Lookup#ORIGINAL ORIGINAL} bit set.
* This lookup object is a <em>capability</em> which may be delegated to trusted agents.
* Do not store it in place where untrusted code can access it.
* <p>
* This method is caller sensitive, which means that it may return different
* values to different callers.
* In cases where {@code MethodHandles.lookup} is called from a context where
* there is no caller frame on the stack (e.g. when called directly
* from a JNI attached thread), {@code IllegalCallerException} is thrown.
* To obtain a {@link Lookup lookup object} in such a context, use an auxiliary class that will
* implicitly be identified as the caller, or use {@link MethodHandles#publicLookup()}
* to obtain a low-privileged lookup instead.
* @return a lookup object for the caller of this method, with
* {@linkplain Lookup#ORIGINAL original} and
* {@linkplain Lookup#hasFullPrivilegeAccess() full privilege access}.
* @throws IllegalCallerException if there is no caller frame on the stack.
*/
@CallerSensitive
@ForceInline // to ensure Reflection.getCallerClass optimization
public static Lookup lookup() {
final Class<?> c = Reflection.getCallerClass();
if (c == null) {
throw new IllegalCallerException("no caller frame");
}
return new Lookup(c);
}
/**
* This lookup method is the alternate implementation of
* the lookup method with a leading caller class argument which is
* non-caller-sensitive. This method is only invoked by reflection
* and method handle.
*/
@CallerSensitiveAdapter
private static Lookup lookup(Class<?> caller) {
if (caller.getClassLoader() == null) {
throw newInternalError("calling lookup() reflectively is not supported: "+caller);
}
return new Lookup(caller);
}
/**
* Returns a {@link Lookup lookup object} which is trusted minimally.
* The lookup has the {@code UNCONDITIONAL} mode.
* It can only be used to create method handles to public members of
* public classes in packages that are exported unconditionally.
* <p>
* As a matter of pure convention, the {@linkplain Lookup#lookupClass() lookup class}
* of this lookup object will be {@link java.lang.Object}.
*
* @apiNote The use of Object is conventional, and because the lookup modes are
* limited, there is no special access provided to the internals of Object, its package
* or its module. This public lookup object or other lookup object with
* {@code UNCONDITIONAL} mode assumes readability. Consequently, the lookup class
* is not used to determine the lookup context.
*
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* The lookup class can be changed to any other class {@code C} using an expression of the form
* {@link Lookup#in publicLookup().in(C.class)}.
* Also, it cannot access
* <a href="MethodHandles.Lookup.html#callsens">caller sensitive methods</a>.
* @return a lookup object which is trusted minimally
*/
public static Lookup publicLookup() {
return Lookup.PUBLIC_LOOKUP;
}
/**
* Returns a {@link Lookup lookup} object on a target class to emulate all supported
* bytecode behaviors, including <a href="MethodHandles.Lookup.html#privacc">private access</a>.
* The returned lookup object can provide access to classes in modules and packages,
* and members of those classes, outside the normal rules of Java access control,
* instead conforming to the more permissive rules for modular <em>deep reflection</em>.
* <p>
* A caller, specified as a {@code Lookup} object, in module {@code M1} is
* allowed to do deep reflection on module {@code M2} and package of the target class
* if and only if all of the following conditions are {@code true}:
* <ul>
* <li>The caller lookup object must have {@linkplain Lookup#hasFullPrivilegeAccess()
* full privilege access}. Specifically:
* <ul>
* <li>The caller lookup object must have the {@link Lookup#MODULE MODULE} lookup mode.
* (This is because otherwise there would be no way to ensure the original lookup
* creator was a member of any particular module, and so any subsequent checks
* for readability and qualified exports would become ineffective.)
* <li>The caller lookup object must have {@link Lookup#PRIVATE PRIVATE} access.
* (This is because an application intending to share intra-module access
* using {@link Lookup#MODULE MODULE} alone will inadvertently also share
* deep reflection to its own module.)
* </ul>
* <li>The target class must be a proper class, not a primitive or array class.
* (Thus, {@code M2} is well-defined.)
* <li>If the caller module {@code M1} differs from
* the target module {@code M2} then both of the following must be true:
* <ul>
* <li>{@code M1} {@link Module#canRead reads} {@code M2}.</li>
* <li>{@code M2} {@link Module#isOpen(String,Module) opens} the package
* containing the target class to at least {@code M1}.</li>
* </ul>
* </ul>
* <p>
* If any of the above checks is violated, this method fails with an
* exception.
* <p>
* Otherwise, if {@code M1} and {@code M2} are the same module, this method
* returns a {@code Lookup} on {@code targetClass} with
* {@linkplain Lookup#hasFullPrivilegeAccess() full privilege access}
* with {@code null} previous lookup class.
* <p>
* Otherwise, {@code M1} and {@code M2} are two different modules. This method
* returns a {@code Lookup} on {@code targetClass} that records
* the lookup class of the caller as the new previous lookup class with
* {@code PRIVATE} access but no {@code MODULE} access.
* <p>
* The resulting {@code Lookup} object has no {@code ORIGINAL} access.
*
* @apiNote The {@code Lookup} object returned by this method is allowed to
* {@linkplain Lookup#defineClass(byte[]) define classes} in the runtime package
* of {@code targetClass}. Extreme caution should be taken when opening a package
* to another module as such defined classes have the same full privilege
* access as other members in {@code targetClass}'s module.
*
* @param targetClass the target class
* @param caller the caller lookup object
* @return a lookup object for the target class, with private access
* @throws IllegalArgumentException if {@code targetClass} is a primitive type or void or array class
* @throws NullPointerException if {@code targetClass} or {@code caller} is {@code null}
* @throws IllegalAccessException if any of the other access checks specified above fails
* @since 9
* @see Lookup#dropLookupMode
* @see <a href="MethodHandles.Lookup.html#cross-module-lookup">Cross-module lookups</a>
*/
public static Lookup privateLookupIn(Class<?> targetClass, Lookup caller) throws IllegalAccessException {
if (caller.allowedModes == Lookup.TRUSTED) {
return new Lookup(targetClass);
}
if (targetClass.isPrimitive())
throw new IllegalArgumentException(targetClass + " is a primitive class");
if (targetClass.isArray())
throw new IllegalArgumentException(targetClass + " is an array class");
// Ensure that we can reason accurately about private and module access.
int requireAccess = Lookup.PRIVATE|Lookup.MODULE;
if ((caller.lookupModes() & requireAccess) != requireAccess)
throw new IllegalAccessException("caller does not have PRIVATE and MODULE lookup mode");
// previous lookup class is never set if it has MODULE access
assert caller.previousLookupClass() == null;
Class<?> callerClass = caller.lookupClass();
Module callerModule = callerClass.getModule(); // M1
Module targetModule = targetClass.getModule(); // M2
Class<?> newPreviousClass = null;
int newModes = Lookup.FULL_POWER_MODES & ~Lookup.ORIGINAL;
if (targetModule != callerModule) {
if (!callerModule.canRead(targetModule))
throw new IllegalAccessException(callerModule + " does not read " + targetModule);
if (targetModule.isNamed()) {
String pn = targetClass.getPackageName();
assert !pn.isEmpty() : "unnamed package cannot be in named module";
if (!targetModule.isOpen(pn, callerModule))
throw new IllegalAccessException(targetModule + " does not open " + pn + " to " + callerModule);
}
// M2 != M1, set previous lookup class to M1 and drop MODULE access
newPreviousClass = callerClass;
newModes &= ~Lookup.MODULE;
}
return Lookup.newLookup(targetClass, newPreviousClass, newModes);
}
/**
* Returns the <em>class data</em> associated with the lookup class
* of the given {@code caller} lookup object, or {@code null}.
*
* <p> A hidden class with class data can be created by calling
* {@link Lookup#defineHiddenClassWithClassData(byte[], Object, boolean, Lookup.ClassOption...)
* Lookup::defineHiddenClassWithClassData}.
* This method will cause the static class initializer of the lookup
* class of the given {@code caller} lookup object be executed if
* it has not been initialized.
*
* <p> A hidden class created by {@link Lookup#defineHiddenClass(byte[], boolean, Lookup.ClassOption...)
* Lookup::defineHiddenClass} and non-hidden classes have no class data.
* {@code null} is returned if this method is called on the lookup object
* on these classes.
*
* <p> The {@linkplain Lookup#lookupModes() lookup modes} for this lookup
* must have {@linkplain Lookup#ORIGINAL original access}
* in order to retrieve the class data.
*
* @apiNote
* This method can be called as a bootstrap method for a dynamically computed
* constant. A framework can create a hidden class with class data, for
* example that can be {@code Class} or {@code MethodHandle} object.
* The class data is accessible only to the lookup object
* created by the original caller but inaccessible to other members
* in the same nest. If a framework passes security sensitive objects
* to a hidden class via class data, it is recommended to load the value
* of class data as a dynamically computed constant instead of storing
* the class data in private static field(s) which are accessible to
* other nestmates.
*
* @param <T> the type to cast the class data object to
* @param caller the lookup context describing the class performing the
* operation (normally stacked by the JVM)
* @param name must be {@link ConstantDescs#DEFAULT_NAME}
* ({@code "_"})
* @param type the type of the class data
* @return the value of the class data if present in the lookup class;
* otherwise {@code null}
* @throws IllegalArgumentException if name is not {@code "_"}
* @throws IllegalAccessException if the lookup context does not have
* {@linkplain Lookup#ORIGINAL original} access
* @throws ClassCastException if the class data cannot be converted to
* the given {@code type}
* @throws NullPointerException if {@code caller} or {@code type} argument
* is {@code null}
* @see Lookup#defineHiddenClassWithClassData(byte[], Object, boolean, Lookup.ClassOption...)
* @see MethodHandles#classDataAt(Lookup, String, Class, int)
* @since 16
* @jvms 5.5 Initialization
*/
public static <T> T classData(Lookup caller, String name, Class<T> type) throws IllegalAccessException {
Objects.requireNonNull(caller);
Objects.requireNonNull(type);
if (!ConstantDescs.DEFAULT_NAME.equals(name)) {
throw new IllegalArgumentException("name must be \"_\": " + name);
}
if ((caller.lookupModes() & Lookup.ORIGINAL) != Lookup.ORIGINAL) {
throw new IllegalAccessException(caller + " does not have ORIGINAL access");
}
Object classdata = classData(caller.lookupClass());
if (classdata == null) return null;
try {
return BootstrapMethodInvoker.widenAndCast(classdata, type);
} catch (RuntimeException|Error e) {
throw e; // let CCE and other runtime exceptions through
} catch (Throwable e) {
throw new InternalError(e);
}
}
/*
* Returns the class data set by the VM in the Class::classData field.
*
* This is also invoked by LambdaForms as it cannot use condy via
* MethodHandles::classData due to bootstrapping issue.
*/
static Object classData(Class<?> c) {
UNSAFE.ensureClassInitialized(c);
return SharedSecrets.getJavaLangAccess().classData(c);
}
/**
* Returns the element at the specified index in the
* {@linkplain #classData(Lookup, String, Class) class data},
* if the class data associated with the lookup class
* of the given {@code caller} lookup object is a {@code List}.
* If the class data is not present in this lookup class, this method
* returns {@code null}.
*
* <p> A hidden class with class data can be created by calling
* {@link Lookup#defineHiddenClassWithClassData(byte[], Object, boolean, Lookup.ClassOption...)
* Lookup::defineHiddenClassWithClassData}.
* This method will cause the static class initializer of the lookup
* class of the given {@code caller} lookup object be executed if
* it has not been initialized.
*
* <p> A hidden class created by {@link Lookup#defineHiddenClass(byte[], boolean, Lookup.ClassOption...)
* Lookup::defineHiddenClass} and non-hidden classes have no class data.
* {@code null} is returned if this method is called on the lookup object
* on these classes.
*
* <p> The {@linkplain Lookup#lookupModes() lookup modes} for this lookup
* must have {@linkplain Lookup#ORIGINAL original access}
* in order to retrieve the class data.
*
* @apiNote
* This method can be called as a bootstrap method for a dynamically computed
* constant. A framework can create a hidden class with class data, for
* example that can be {@code List.of(o1, o2, o3....)} containing more than
* one object and use this method to load one element at a specific index.
* The class data is accessible only to the lookup object
* created by the original caller but inaccessible to other members
* in the same nest. If a framework passes security sensitive objects
* to a hidden class via class data, it is recommended to load the value
* of class data as a dynamically computed constant instead of storing
* the class data in private static field(s) which are accessible to other
* nestmates.
*
* @param <T> the type to cast the result object to
* @param caller the lookup context describing the class performing the
* operation (normally stacked by the JVM)
* @param name must be {@link java.lang.constant.ConstantDescs#DEFAULT_NAME}
* ({@code "_"})
* @param type the type of the element at the given index in the class data
* @param index index of the element in the class data
* @return the element at the given index in the class data
* if the class data is present; otherwise {@code null}
* @throws IllegalArgumentException if name is not {@code "_"}
* @throws IllegalAccessException if the lookup context does not have
* {@linkplain Lookup#ORIGINAL original} access
* @throws ClassCastException if the class data cannot be converted to {@code List}
* or the element at the specified index cannot be converted to the given type
* @throws IndexOutOfBoundsException if the index is out of range
* @throws NullPointerException if {@code caller} or {@code type} argument is
* {@code null}; or if unboxing operation fails because
* the element at the given index is {@code null}
*
* @since 16
* @see #classData(Lookup, String, Class)
* @see Lookup#defineHiddenClassWithClassData(byte[], Object, boolean, Lookup.ClassOption...)
*/
public static <T> T classDataAt(Lookup caller, String name, Class<T> type, int index)
throws IllegalAccessException
{
@SuppressWarnings("unchecked")
List<Object> classdata = (List<Object>)classData(caller, name, List.class);
if (classdata == null) return null;
try {
Object element = classdata.get(index);
return BootstrapMethodInvoker.widenAndCast(element, type);
} catch (RuntimeException|Error e) {
throw e; // let specified exceptions and other runtime exceptions/errors through
} catch (Throwable e) {
throw new InternalError(e);
}
}
/**
* Performs an unchecked "crack" of a
* <a href="MethodHandleInfo.html#directmh">direct method handle</a>.
* The result is as if the user had obtained a lookup object capable enough
* to crack the target method handle, called
* {@link java.lang.invoke.MethodHandles.Lookup#revealDirect Lookup.revealDirect}
* on the target to obtain its symbolic reference, and then called
* {@link java.lang.invoke.MethodHandleInfo#reflectAs MethodHandleInfo.reflectAs}
* to resolve the symbolic reference to a member.
* @param <T> the desired type of the result, either {@link Member} or a subtype
* @param expected a class object representing the desired result type {@code T}
* @param target a direct method handle to crack into symbolic reference components
* @return a reference to the method, constructor, or field object
* @throws NullPointerException if either argument is {@code null}
* @throws IllegalArgumentException if the target is not a direct method handle
* @throws ClassCastException if the member is not of the expected type
* @since 1.8
*/
public static <T extends Member> T reflectAs(Class<T> expected, MethodHandle target) {
Lookup lookup = Lookup.IMPL_LOOKUP; // use maximally privileged lookup
return lookup.revealDirect(target).reflectAs(expected, lookup);
}
/**
* A <em>lookup object</em> is a factory for creating method handles,
* when the creation requires access checking.
* Method handles do not perform
* access checks when they are called, but rather when they are created.
* Therefore, method handle access
* restrictions must be enforced when a method handle is created.
* The caller class against which those restrictions are enforced
* is known as the {@linkplain #lookupClass() lookup class}.
* <p>
* A lookup class which needs to create method handles will call
* {@link MethodHandles#lookup() MethodHandles.lookup} to create a factory for itself.
* When the {@code Lookup} factory object is created, the identity of the lookup class is
* determined, and securely stored in the {@code Lookup} object.
* The lookup class (or its delegates) may then use factory methods
* on the {@code Lookup} object to create method handles for access-checked members.
* This includes all methods, constructors, and fields which are allowed to the lookup class,
* even private ones.
*
* <h2><a id="lookups"></a>Lookup Factory Methods</h2>
* The factory methods on a {@code Lookup} object correspond to all major
* use cases for methods, constructors, and fields.
* Each method handle created by a factory method is the functional
* equivalent of a particular <em>bytecode behavior</em>.
* (Bytecode behaviors are described in section {@jvms 5.4.3.5} of
* the Java Virtual Machine Specification.)
* Here is a summary of the correspondence between these factory methods and
* the behavior of the resulting method handles:
* <table class="striped">
* <caption style="display:none">lookup method behaviors</caption>
* <thead>
* <tr>
* <th scope="col"><a id="equiv"></a>lookup expression</th>
* <th scope="col">member</th>
* <th scope="col">bytecode behavior</th>
* </tr>
* </thead>
* <tbody>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findGetter lookup.findGetter(C.class,"f",FT.class)}</th>
* <td>{@code FT f;}</td><td>{@code (FT) this.f;}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findStaticGetter lookup.findStaticGetter(C.class,"f",FT.class)}</th>
* <td>{@code static}<br>{@code FT f;}</td><td>{@code (FT) C.f;}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findSetter lookup.findSetter(C.class,"f",FT.class)}</th>
* <td>{@code FT f;}</td><td>{@code this.f = x;}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findStaticSetter lookup.findStaticSetter(C.class,"f",FT.class)}</th>
* <td>{@code static}<br>{@code FT f;}</td><td>{@code C.f = arg;}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findVirtual lookup.findVirtual(C.class,"m",MT)}</th>
* <td>{@code T m(A*);}</td><td>{@code (T) this.m(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findStatic lookup.findStatic(C.class,"m",MT)}</th>
* <td>{@code static}<br>{@code T m(A*);}</td><td>{@code (T) C.m(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findSpecial lookup.findSpecial(C.class,"m",MT,this.class)}</th>
* <td>{@code T m(A*);}</td><td>{@code (T) super.m(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findConstructor lookup.findConstructor(C.class,MT)}</th>
* <td>{@code C(A*);}</td><td>{@code new C(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#unreflectGetter lookup.unreflectGetter(aField)}</th>
* <td>({@code static})?<br>{@code FT f;}</td><td>{@code (FT) aField.get(thisOrNull);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#unreflectSetter lookup.unreflectSetter(aField)}</th>
* <td>({@code static})?<br>{@code FT f;}</td><td>{@code aField.set(thisOrNull, arg);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#unreflect lookup.unreflect(aMethod)}</th>
* <td>({@code static})?<br>{@code T m(A*);}</td><td>{@code (T) aMethod.invoke(thisOrNull, arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#unreflectConstructor lookup.unreflectConstructor(aConstructor)}</th>
* <td>{@code C(A*);}</td><td>{@code (C) aConstructor.newInstance(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#unreflectSpecial lookup.unreflectSpecial(aMethod,this.class)}</th>
* <td>{@code T m(A*);}</td><td>{@code (T) super.m(arg*);}</td>
* </tr>
* <tr>
* <th scope="row">{@link java.lang.invoke.MethodHandles.Lookup#findClass lookup.findClass("C")}</th>
* <td>{@code class C { ... }}</td><td>{@code C.class;}</td>
* </tr>
* </tbody>
* </table>
*
* Here, the type {@code C} is the class or interface being searched for a member,
* documented as a parameter named {@code refc} in the lookup methods.
* The method type {@code MT} is composed from the return type {@code T}
* and the sequence of argument types {@code A*}.
* The constructor also has a sequence of argument types {@code A*} and
* is deemed to return the newly-created object of type {@code C}.
* Both {@code MT} and the field type {@code FT} are documented as a parameter named {@code type}.
* The formal parameter {@code this} stands for the self-reference of type {@code C};
* if it is present, it is always the leading argument to the method handle invocation.
* (In the case of some {@code protected} members, {@code this} may be
* restricted in type to the lookup class; see below.)
* The name {@code arg} stands for all the other method handle arguments.
* In the code examples for the Core Reflection API, the name {@code thisOrNull}
* stands for a null reference if the accessed method or field is static,
* and {@code this} otherwise.
* The names {@code aMethod}, {@code aField}, and {@code aConstructor} stand
* for reflective objects corresponding to the given members declared in type {@code C}.
* <p>
* The bytecode behavior for a {@code findClass} operation is a load of a constant class,
* as if by {@code ldc CONSTANT_Class}.
* The behavior is represented, not as a method handle, but directly as a {@code Class} constant.
* <p>
* In cases where the given member is of variable arity (i.e., a method or constructor)
* the returned method handle will also be of {@linkplain MethodHandle#asVarargsCollector variable arity}.
* In all other cases, the returned method handle will be of fixed arity.
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* The equivalence between looked-up method handles and underlying
* class members and bytecode behaviors
* can break down in a few ways:
* <ul style="font-size:smaller;">
* <li>If {@code C} is not symbolically accessible from the lookup class's loader,
* the lookup can still succeed, even when there is no equivalent
* Java expression or bytecoded constant.
* <li>Likewise, if {@code T} or {@code MT}
* is not symbolically accessible from the lookup class's loader,
* the lookup can still succeed.
* For example, lookups for {@code MethodHandle.invokeExact} and
* {@code MethodHandle.invoke} will always succeed, regardless of requested type.
* <li>If the looked-up method has a
* <a href="MethodHandle.html#maxarity">very large arity</a>,
* the method handle creation may fail with an
* {@code IllegalArgumentException}, due to the method handle type having
* <a href="MethodHandle.html#maxarity">too many parameters.</a>
* </ul>
*
* <h2><a id="access"></a>Access checking</h2>
* Access checks are applied in the factory methods of {@code Lookup},
* when a method handle is created.
* This is a key difference from the Core Reflection API, since
* {@link java.lang.reflect.Method#invoke java.lang.reflect.Method.invoke}
* performs access checking against every caller, on every call.
* <p>
* All access checks start from a {@code Lookup} object, which
* compares its recorded lookup class against all requests to
* create method handles.
* A single {@code Lookup} object can be used to create any number
* of access-checked method handles, all checked against a single
* lookup class.
* <p>
* A {@code Lookup} object can be shared with other trusted code,
* such as a metaobject protocol.
* A shared {@code Lookup} object delegates the capability
* to create method handles on private members of the lookup class.
* Even if privileged code uses the {@code Lookup} object,
* the access checking is confined to the privileges of the
* original lookup class.
* <p>
* A lookup can fail, because
* the containing class is not accessible to the lookup class, or
* because the desired class member is missing, or because the
* desired class member is not accessible to the lookup class, or
* because the lookup object is not trusted enough to access the member.
* In the case of a field setter function on a {@code final} field,
* finality enforcement is treated as a kind of access control,
* and the lookup will fail, except in special cases of
* {@link Lookup#unreflectSetter Lookup.unreflectSetter}.
* In any of these cases, a {@code ReflectiveOperationException} will be
* thrown from the attempted lookup. The exact class will be one of
* the following:
* <ul>
* <li>NoSuchMethodException &mdash; if a method is requested but does not exist
* <li>NoSuchFieldException &mdash; if a field is requested but does not exist
* <li>IllegalAccessException &mdash; if the member exists but an access check fails
* </ul>
* <p>
* In general, the conditions under which a method handle may be
* looked up for a method {@code M} are no more restrictive than the conditions
* under which the lookup class could have compiled, verified, and resolved a call to {@code M}.
* Where the JVM would raise exceptions like {@code NoSuchMethodError},
* a method handle lookup will generally raise a corresponding
* checked exception, such as {@code NoSuchMethodException}.
* And the effect of invoking the method handle resulting from the lookup
* is <a href="MethodHandles.Lookup.html#equiv">exactly equivalent</a>
* to executing the compiled, verified, and resolved call to {@code M}.
* The same point is true of fields and constructors.
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* Access checks only apply to named and reflected methods,
* constructors, and fields.
* Other method handle creation methods, such as
* {@link MethodHandle#asType MethodHandle.asType},
* do not require any access checks, and are used
* independently of any {@code Lookup} object.
* <p>
* If the desired member is {@code protected}, the usual JVM rules apply,
* including the requirement that the lookup class must either be in the
* same package as the desired member, or must inherit that member.
* (See the Java Virtual Machine Specification, sections {@jvms
* 4.9.2}, {@jvms 5.4.3.5}, and {@jvms 6.4}.)
* In addition, if the desired member is a non-static field or method
* in a different package, the resulting method handle may only be applied
* to objects of the lookup class or one of its subclasses.
* This requirement is enforced by narrowing the type of the leading
* {@code this} parameter from {@code C}
* (which will necessarily be a superclass of the lookup class)
* to the lookup class itself.
* <p>
* The JVM imposes a similar requirement on {@code invokespecial} instruction,
* that the receiver argument must match both the resolved method <em>and</em>
* the current class. Again, this requirement is enforced by narrowing the
* type of the leading parameter to the resulting method handle.
* (See the Java Virtual Machine Specification, section {@jvms 4.10.1.9}.)
* <p>
* The JVM represents constructors and static initializer blocks as internal methods
* with special names ({@value ConstantDescs#INIT_NAME} and {@value
* ConstantDescs#CLASS_INIT_NAME}).
* The internal syntax of invocation instructions allows them to refer to such internal
* methods as if they were normal methods, but the JVM bytecode verifier rejects them.
* A lookup of such an internal method will produce a {@code NoSuchMethodException}.
* <p>
* If the relationship between nested types is expressed directly through the
* {@code NestHost} and {@code NestMembers} attributes
* (see the Java Virtual Machine Specification, sections {@jvms
* 4.7.28} and {@jvms 4.7.29}),
* then the associated {@code Lookup} object provides direct access to
* the lookup class and all of its nestmates
* (see {@link java.lang.Class#getNestHost Class.getNestHost}).
* Otherwise, access between nested classes is obtained by the Java compiler creating
* a wrapper method to access a private method of another class in the same nest.
* For example, a nested class {@code C.D}
* can access private members within other related classes such as
* {@code C}, {@code C.D.E}, or {@code C.B},
* but the Java compiler may need to generate wrapper methods in
* those related classes. In such cases, a {@code Lookup} object on
* {@code C.E} would be unable to access those private members.
* A workaround for this limitation is the {@link Lookup#in Lookup.in} method,
* which can transform a lookup on {@code C.E} into one on any of those other
* classes, without special elevation of privilege.
* <p>
* The accesses permitted to a given lookup object may be limited,
* according to its set of {@link #lookupModes lookupModes},
* to a subset of members normally accessible to the lookup class.
* For example, the {@link MethodHandles#publicLookup publicLookup}
* method produces a lookup object which is only allowed to access
* public members in public classes of exported packages.
* The caller sensitive method {@link MethodHandles#lookup lookup}
* produces a lookup object with full capabilities relative to
* its caller class, to emulate all supported bytecode behaviors.
* Also, the {@link Lookup#in Lookup.in} method may produce a lookup object
* with fewer access modes than the original lookup object.
*
* <p style="font-size:smaller;">
* <a id="privacc"></a>
* <em>Discussion of private and module access:</em>
* We say that a lookup has <em>private access</em>
* if its {@linkplain #lookupModes lookup modes}
* include the possibility of accessing {@code private} members
* (which includes the private members of nestmates).
* As documented in the relevant methods elsewhere,
* only lookups with private access possess the following capabilities:
* <ul style="font-size:smaller;">
* <li>access private fields, methods, and constructors of the lookup class and its nestmates
* <li>create method handles which {@link Lookup#findSpecial emulate invokespecial} instructions
* <li>create {@link Lookup#in delegated lookup objects} which have private access to other classes
* within the same package member
* </ul>
* <p style="font-size:smaller;">
* Similarly, a lookup with module access ensures that the original lookup creator was
* a member in the same module as the lookup class.
* <p style="font-size:smaller;">
* Private and module access are independently determined modes; a lookup may have
* either or both or neither. A lookup which possesses both access modes is said to
* possess {@linkplain #hasFullPrivilegeAccess() full privilege access}.
* <p style="font-size:smaller;">
* A lookup with <em>original access</em> ensures that this lookup is created by
* the original lookup class and the bootstrap method invoked by the VM.
* Such a lookup with original access also has private and module access
* which has the following additional capability:
* <ul style="font-size:smaller;">
* <li>create method handles which invoke <a href="MethodHandles.Lookup.html#callsens">caller sensitive</a> methods,
* such as {@code Class.forName}
* <li>obtain the {@linkplain MethodHandles#classData(Lookup, String, Class)
* class data} associated with the lookup class</li>
* </ul>
* <p style="font-size:smaller;">
* Each of these permissions is a consequence of the fact that a lookup object
* with private access can be securely traced back to an originating class,
* whose <a href="MethodHandles.Lookup.html#equiv">bytecode behaviors</a> and Java language access permissions
* can be reliably determined and emulated by method handles.
*
* <h2><a id="cross-module-lookup"></a>Cross-module lookups</h2>
* When a lookup class in one module {@code M1} accesses a class in another module
* {@code M2}, extra access checking is performed beyond the access mode bits.
* A {@code Lookup} with {@link #PUBLIC} mode and a lookup class in {@code M1}
* can access public types in {@code M2} when {@code M2} is readable to {@code M1}
* and when the type is in a package of {@code M2} that is exported to
* at least {@code M1}.
* <p>
* A {@code Lookup} on {@code C} can also <em>teleport</em> to a target class
* via {@link #in(Class) Lookup.in} and {@link MethodHandles#privateLookupIn(Class, Lookup)
* MethodHandles.privateLookupIn} methods.
* Teleporting across modules will always record the original lookup class as
* the <em>{@linkplain #previousLookupClass() previous lookup class}</em>
* and drops {@link Lookup#MODULE MODULE} access.
* If the target class is in the same module as the lookup class {@code C},
* then the target class becomes the new lookup class
* and there is no change to the previous lookup class.
* If the target class is in a different module from {@code M1} ({@code C}'s module),
* {@code C} becomes the new previous lookup class
* and the target class becomes the new lookup class.
* In that case, if there was already a previous lookup class in {@code M0},
* and it differs from {@code M1} and {@code M2}, then the resulting lookup
* drops all privileges.
* For example,
* {@snippet lang="java" :
* Lookup lookup = MethodHandles.lookup(); // in class C
* Lookup lookup2 = lookup.in(D.class);
* MethodHandle mh = lookup2.findStatic(E.class, "m", MT);
* }
* <p>
* The {@link #lookup()} factory method produces a {@code Lookup} object
* with {@code null} previous lookup class.
* {@link Lookup#in lookup.in(D.class)} transforms the {@code lookup} on class {@code C}
* to class {@code D} without elevation of privileges.
* If {@code C} and {@code D} are in the same module,
* {@code lookup2} records {@code D} as the new lookup class and keeps the
* same previous lookup class as the original {@code lookup}, or
* {@code null} if not present.
* <p>
* When a {@code Lookup} teleports from a class
* in one nest to another nest, {@code PRIVATE} access is dropped.
* When a {@code Lookup} teleports from a class in one package to
* another package, {@code PACKAGE} access is dropped.
* When a {@code Lookup} teleports from a class in one module to another module,
* {@code MODULE} access is dropped.
* Teleporting across modules drops the ability to access non-exported classes
* in both the module of the new lookup class and the module of the old lookup class
* and the resulting {@code Lookup} remains only {@code PUBLIC} access.
* A {@code Lookup} can teleport back and forth to a class in the module of
* the lookup class and the module of the previous class lookup.
* Teleporting across modules can only decrease access but cannot increase it.
* Teleporting to some third module drops all accesses.
* <p>
* In the above example, if {@code C} and {@code D} are in different modules,
* {@code lookup2} records {@code D} as its lookup class and
* {@code C} as its previous lookup class and {@code lookup2} has only
* {@code PUBLIC} access. {@code lookup2} can teleport to other class in
* {@code C}'s module and {@code D}'s module.
* If class {@code E} is in a third module, {@code lookup2.in(E.class)} creates
* a {@code Lookup} on {@code E} with no access and {@code lookup2}'s lookup
* class {@code D} is recorded as its previous lookup class.
* <p>
* Teleporting across modules restricts access to the public types that
* both the lookup class and the previous lookup class can equally access
* (see below).
* <p>
* {@link MethodHandles#privateLookupIn(Class, Lookup) MethodHandles.privateLookupIn(T.class, lookup)}
* can be used to teleport a {@code lookup} from class {@code C} to class {@code T}
* and produce a new {@code Lookup} with <a href="#privacc">private access</a>
* if the lookup class is allowed to do <em>deep reflection</em> on {@code T}.
* The {@code lookup} must have {@link #MODULE} and {@link #PRIVATE} access
* to call {@code privateLookupIn}.
* A {@code lookup} on {@code C} in module {@code M1} is allowed to do deep reflection
* on all classes in {@code M1}. If {@code T} is in {@code M1}, {@code privateLookupIn}
* produces a new {@code Lookup} on {@code T} with full capabilities.
* A {@code lookup} on {@code C} is also allowed
* to do deep reflection on {@code T} in another module {@code M2} if
* {@code M1} reads {@code M2} and {@code M2} {@link Module#isOpen(String,Module) opens}
* the package containing {@code T} to at least {@code M1}.
* {@code T} becomes the new lookup class and {@code C} becomes the new previous
* lookup class and {@code MODULE} access is dropped from the resulting {@code Lookup}.
* The resulting {@code Lookup} can be used to do member lookup or teleport
* to another lookup class by calling {@link #in Lookup::in}. But
* it cannot be used to obtain another private {@code Lookup} by calling
* {@link MethodHandles#privateLookupIn(Class, Lookup) privateLookupIn}
* because it has no {@code MODULE} access.
* <p>
* The {@code Lookup} object returned by {@code privateLookupIn} is allowed to
* {@linkplain Lookup#defineClass(byte[]) define classes} in the runtime package
* of {@code T}. Extreme caution should be taken when opening a package
* to another module as such defined classes have the same full privilege
* access as other members in {@code M2}.
*
* <h2><a id="module-access-check"></a>Cross-module access checks</h2>
*
* A {@code Lookup} with {@link #PUBLIC} or with {@link #UNCONDITIONAL} mode
* allows cross-module access. The access checking is performed with respect
* to both the lookup class and the previous lookup class if present.
* <p>
* A {@code Lookup} with {@link #UNCONDITIONAL} mode can access public type
* in all modules when the type is in a package that is {@linkplain Module#isExported(String)
* exported unconditionally}.
* <p>
* If a {@code Lookup} on {@code LC} in {@code M1} has no previous lookup class,
* the lookup with {@link #PUBLIC} mode can access all public types in modules
* that are readable to {@code M1} and the type is in a package that is exported
* at least to {@code M1}.
* <p>
* If a {@code Lookup} on {@code LC} in {@code M1} has a previous lookup class
* {@code PLC} on {@code M0}, the lookup with {@link #PUBLIC} mode can access
* the intersection of all public types that are accessible to {@code M1}
* with all public types that are accessible to {@code M0}. {@code M0}
* reads {@code M1} and hence the set of accessible types includes:
*
* <ul>
* <li>unconditional-exported packages from {@code M1}</li>
* <li>unconditional-exported packages from {@code M0} if {@code M1} reads {@code M0}</li>
* <li>
* unconditional-exported packages from a third module {@code M2}if both {@code M0}
* and {@code M1} read {@code M2}
* </li>
* <li>qualified-exported packages from {@code M1} to {@code M0}</li>
* <li>qualified-exported packages from {@code M0} to {@code M1} if {@code M1} reads {@code M0}</li>
* <li>
* qualified-exported packages from a third module {@code M2} to both {@code M0} and
* {@code M1} if both {@code M0} and {@code M1} read {@code M2}
* </li>
* </ul>
*
* <h2><a id="access-modes"></a>Access modes</h2>
*
* The table below shows the access modes of a {@code Lookup} produced by
* any of the following factory or transformation methods:
* <ul>
* <li>{@link #lookup() MethodHandles::lookup}</li>
* <li>{@link #publicLookup() MethodHandles::publicLookup}</li>
* <li>{@link #privateLookupIn(Class, Lookup) MethodHandles::privateLookupIn}</li>
* <li>{@link Lookup#in Lookup::in}</li>
* <li>{@link Lookup#dropLookupMode(int) Lookup::dropLookupMode}</li>
* </ul>
*
* <table class="striped">
* <caption style="display:none">
* Access mode summary
* </caption>
* <thead>
* <tr>
* <th scope="col">Lookup object</th>
* <th style="text-align:center">original</th>
* <th style="text-align:center">protected</th>
* <th style="text-align:center">private</th>
* <th style="text-align:center">package</th>
* <th style="text-align:center">module</th>
* <th style="text-align:center">public</th>
* </tr>
* </thead>
* <tbody>
* <tr>
* <th scope="row" style="text-align:left">{@code CL = MethodHandles.lookup()} in {@code C}</th>
* <td style="text-align:center">ORI</td>
* <td style="text-align:center">PRO</td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.in(C1)} same package</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.in(C1)} same module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.in(D)} different module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.in(D).in(C)} hop back to module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1 = privateLookupIn(C1,CL)}</th>
* <td></td>
* <td style="text-align:center">PRO</td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1a = privateLookupIn(C,PRI1)}</th>
* <td></td>
* <td style="text-align:center">PRO</td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.in(C1)} same package</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.in(C1)} different package</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.in(D)} different module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.dropLookupMode(PROTECTED)}</th>
* <td></td>
* <td></td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.dropLookupMode(PRIVATE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.dropLookupMode(PACKAGE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.dropLookupMode(MODULE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI1.dropLookupMode(PUBLIC)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2 = privateLookupIn(D,CL)}</th>
* <td></td>
* <td style="text-align:center">PRO</td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code privateLookupIn(D,PRI1)}</th>
* <td></td>
* <td style="text-align:center">PRO</td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code privateLookupIn(C,PRI2)} fails</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">IAE</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.in(D2)} same package</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.in(D2)} different package</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.in(C1)} hop back to module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.in(E)} hop to third module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.dropLookupMode(PROTECTED)}</th>
* <td></td>
* <td></td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.dropLookupMode(PRIVATE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.dropLookupMode(PACKAGE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.dropLookupMode(MODULE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">2R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PRI2.dropLookupMode(PUBLIC)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.dropLookupMode(PROTECTED)}</th>
* <td></td>
* <td></td>
* <td style="text-align:center">PRI</td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.dropLookupMode(PRIVATE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">PAC</td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.dropLookupMode(PACKAGE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">MOD</td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.dropLookupMode(MODULE)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">1R</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code CL.dropLookupMode(PUBLIC)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PUB = publicLookup()}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">U</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PUB.in(D)} different module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">U</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PUB.in(D).in(E)} third module</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">U</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code PUB.dropLookupMode(UNCONDITIONAL)}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code privateLookupIn(C1,PUB)} fails</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">IAE</td>
* </tr>
* <tr>
* <th scope="row" style="text-align:left">{@code ANY.in(X)}, for inaccessible {@code X}</th>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td></td>
* <td style="text-align:center">none</td>
* </tr>
* </tbody>
* </table>
*
* <p>
* Notes:
* <ul>
* <li>Class {@code C} and class {@code C1} are in module {@code M1},
* but {@code D} and {@code D2} are in module {@code M2}, and {@code E}
* is in module {@code M3}. {@code X} stands for class which is inaccessible
* to the lookup. {@code ANY} stands for any of the example lookups.</li>
* <li>{@code ORI} indicates {@link #ORIGINAL} bit set,
* {@code PRO} indicates {@link #PROTECTED} bit set,
* {@code PRI} indicates {@link #PRIVATE} bit set,
* {@code PAC} indicates {@link #PACKAGE} bit set,
* {@code MOD} indicates {@link #MODULE} bit set,
* {@code 1R} and {@code 2R} indicate {@link #PUBLIC} bit set,
* {@code U} indicates {@link #UNCONDITIONAL} bit set,
* {@code IAE} indicates {@code IllegalAccessException} thrown.</li>
* <li>Public access comes in three kinds:
* <ul>
* <li>unconditional ({@code U}): the lookup assumes readability.
* The lookup has {@code null} previous lookup class.
* <li>one-module-reads ({@code 1R}): the module access checking is
* performed with respect to the lookup class. The lookup has {@code null}
* previous lookup class.
* <li>two-module-reads ({@code 2R}): the module access checking is
* performed with respect to the lookup class and the previous lookup class.
* The lookup has a non-null previous lookup class which is in a
* different module from the current lookup class.
* </ul>
* <li>Any attempt to reach a third module loses all access.</li>
* <li>If a target class {@code X} is not accessible to {@code Lookup::in}
* all access modes are dropped.</li>
* </ul>
*
* <h2><a id="callsens"></a>Caller sensitive methods</h2>
* A small number of Java methods have a special property called caller sensitivity.
* A <em>caller-sensitive</em> method can behave differently depending on the
* identity of its immediate caller.
* <p>
* If a method handle for a caller-sensitive method is requested,
* the general rules for <a href="MethodHandles.Lookup.html#equiv">bytecode behaviors</a> apply,
* but they take account of the lookup class in a special way.
* The resulting method handle behaves as if it were called
* from an instruction contained in the lookup class,
* so that the caller-sensitive method detects the lookup class.
* (By contrast, the invoker of the method handle is disregarded.)
* Thus, in the case of caller-sensitive methods,
* different lookup classes may give rise to
* differently behaving method handles.
* <p>
* In cases where the lookup object is
* {@link MethodHandles#publicLookup() publicLookup()},
* or some other lookup object without the
* {@linkplain #ORIGINAL original access},
* the lookup class is disregarded.
* In such cases, no caller-sensitive method handle can be created,
* access is forbidden, and the lookup fails with an
* {@code IllegalAccessException}.
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* For example, the caller-sensitive method
* {@link java.lang.Class#forName(String) Class.forName(x)}
* can return varying classes or throw varying exceptions,
* depending on the class loader of the class that calls it.
* A public lookup of {@code Class.forName} will fail, because
* there is no reasonable way to determine its bytecode behavior.
* <p style="font-size:smaller;">
* If an application caches method handles for broad sharing,
* it should use {@code publicLookup()} to create them.
* If there is a lookup of {@code Class.forName}, it will fail,
* and the application must take appropriate action in that case.
* It may be that a later lookup, perhaps during the invocation of a
* bootstrap method, can incorporate the specific identity
* of the caller, making the method accessible.
* <p style="font-size:smaller;">
* The function {@code MethodHandles.lookup} is caller sensitive
* so that there can be a secure foundation for lookups.
* Nearly all other methods in the JSR 292 API rely on lookup
* objects to check access requests.
*/
public static final
class Lookup {
/** The class on behalf of whom the lookup is being performed. */
private final Class<?> lookupClass;
/** previous lookup class */
private final Class<?> prevLookupClass;
/** The allowed sorts of members which may be looked up (PUBLIC, etc.). */
private final int allowedModes;
static {
Reflection.registerFieldsToFilter(Lookup.class, Set.of("lookupClass", "allowedModes"));
}
/** A single-bit mask representing {@code public} access,
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value, {@code 0x01}, happens to be the same as the value of the
* {@code public} {@linkplain java.lang.reflect.Modifier#PUBLIC modifier bit}.
* <p>
* A {@code Lookup} with this lookup mode performs cross-module access check
* with respect to the {@linkplain #lookupClass() lookup class} and
* {@linkplain #previousLookupClass() previous lookup class} if present.
*/
public static final int PUBLIC = Modifier.PUBLIC;
/** A single-bit mask representing {@code private} access,
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value, {@code 0x02}, happens to be the same as the value of the
* {@code private} {@linkplain java.lang.reflect.Modifier#PRIVATE modifier bit}.
*/
public static final int PRIVATE = Modifier.PRIVATE;
/** A single-bit mask representing {@code protected} access,
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value, {@code 0x04}, happens to be the same as the value of the
* {@code protected} {@linkplain java.lang.reflect.Modifier#PROTECTED modifier bit}.
*/
public static final int PROTECTED = Modifier.PROTECTED;
/** A single-bit mask representing {@code package} access (default access),
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value is {@code 0x08}, which does not correspond meaningfully to
* any particular {@linkplain java.lang.reflect.Modifier modifier bit}.
*/
public static final int PACKAGE = Modifier.STATIC;
/** A single-bit mask representing {@code module} access,
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value is {@code 0x10}, which does not correspond meaningfully to
* any particular {@linkplain java.lang.reflect.Modifier modifier bit}.
* In conjunction with the {@code PUBLIC} modifier bit, a {@code Lookup}
* with this lookup mode can access all public types in the module of the
* lookup class and public types in packages exported by other modules
* to the module of the lookup class.
* <p>
* If this lookup mode is set, the {@linkplain #previousLookupClass()
* previous lookup class} is always {@code null}.
*
* @since 9
*/
public static final int MODULE = PACKAGE << 1;
/** A single-bit mask representing {@code unconditional} access
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value is {@code 0x20}, which does not correspond meaningfully to
* any particular {@linkplain java.lang.reflect.Modifier modifier bit}.
* A {@code Lookup} with this lookup mode assumes {@linkplain
* java.lang.Module#canRead(java.lang.Module) readability}.
* This lookup mode can access all public members of public types
* of all modules when the type is in a package that is {@link
* java.lang.Module#isExported(String) exported unconditionally}.
*
* <p>
* If this lookup mode is set, the {@linkplain #previousLookupClass()
* previous lookup class} is always {@code null}.
*
* @since 9
* @see #publicLookup()
*/
public static final int UNCONDITIONAL = PACKAGE << 2;
/** A single-bit mask representing {@code original} access
* which may contribute to the result of {@link #lookupModes lookupModes}.
* The value is {@code 0x40}, which does not correspond meaningfully to
* any particular {@linkplain java.lang.reflect.Modifier modifier bit}.
*
* <p>
* If this lookup mode is set, the {@code Lookup} object must be
* created by the original lookup class by calling
* {@link MethodHandles#lookup()} method or by a bootstrap method
* invoked by the VM. The {@code Lookup} object with this lookup
* mode has {@linkplain #hasFullPrivilegeAccess() full privilege access}.
*
* @since 16
*/
public static final int ORIGINAL = PACKAGE << 3;
private static final int ALL_MODES = (PUBLIC | PRIVATE | PROTECTED | PACKAGE | MODULE | UNCONDITIONAL | ORIGINAL);
private static final int FULL_POWER_MODES = (ALL_MODES & ~UNCONDITIONAL); // with original access
private static final int TRUSTED = -1;
/*
* Adjust PUBLIC => PUBLIC|MODULE|ORIGINAL|UNCONDITIONAL
* Adjust 0 => PACKAGE
*/
private static int fixmods(int mods) {
mods &= (ALL_MODES - PACKAGE - MODULE - ORIGINAL - UNCONDITIONAL);
if (Modifier.isPublic(mods))
mods |= UNCONDITIONAL;
return (mods != 0) ? mods : PACKAGE;
}
/** Tells which class is performing the lookup. It is this class against
* which checks are performed for visibility and access permissions.
* <p>
* If this lookup object has a {@linkplain #previousLookupClass() previous lookup class},
* access checks are performed against both the lookup class and the previous lookup class.
* <p>
* The class implies a maximum level of access permission,
* but the permissions may be additionally limited by the bitmask
* {@link #lookupModes lookupModes}, which controls whether non-public members
* can be accessed.
* @return the lookup class, on behalf of which this lookup object finds members
* @see <a href="#cross-module-lookup">Cross-module lookups</a>
*/
public Class<?> lookupClass() {
return lookupClass;
}
/** Reports a lookup class in another module that this lookup object
* was previously teleported from, or {@code null}.
* <p>
* A {@code Lookup} object produced by the factory methods, such as the
* {@link #lookup() lookup()} and {@link #publicLookup() publicLookup()} method,
* has {@code null} previous lookup class.
* A {@code Lookup} object has a non-null previous lookup class
* when this lookup was teleported from an old lookup class
* in one module to a new lookup class in another module.
*
* @return the lookup class in another module that this lookup object was
* previously teleported from, or {@code null}
* @since 14
* @see #in(Class)
* @see MethodHandles#privateLookupIn(Class, Lookup)
* @see <a href="#cross-module-lookup">Cross-module lookups</a>
*/
public Class<?> previousLookupClass() {
return prevLookupClass;
}
// This is just for calling out to MethodHandleImpl.
private Class<?> lookupClassOrNull() {
return (allowedModes == TRUSTED) ? null : lookupClass;
}
/** Tells which access-protection classes of members this lookup object can produce.
* The result is a bit-mask of the bits
* {@linkplain #PUBLIC PUBLIC (0x01)},
* {@linkplain #PRIVATE PRIVATE (0x02)},
* {@linkplain #PROTECTED PROTECTED (0x04)},
* {@linkplain #PACKAGE PACKAGE (0x08)},
* {@linkplain #MODULE MODULE (0x10)},
* {@linkplain #UNCONDITIONAL UNCONDITIONAL (0x20)},
* and {@linkplain #ORIGINAL ORIGINAL (0x40)}.
* <p>
* A freshly-created lookup object
* on the {@linkplain java.lang.invoke.MethodHandles#lookup() caller's class} has
* all possible bits set, except {@code UNCONDITIONAL}.
* A lookup object on a new lookup class
* {@linkplain java.lang.invoke.MethodHandles.Lookup#in created from a previous lookup object}
* may have some mode bits set to zero.
* Mode bits can also be
* {@linkplain java.lang.invoke.MethodHandles.Lookup#dropLookupMode directly cleared}.
* Once cleared, mode bits cannot be restored from the downgraded lookup object.
* The purpose of this is to restrict access via the new lookup object,
* so that it can access only names which can be reached by the original
* lookup object, and also by the new lookup class.
* @return the lookup modes, which limit the kinds of access performed by this lookup object
* @see #in
* @see #dropLookupMode
*/
public int lookupModes() {
return allowedModes & ALL_MODES;
}
/** Embody the current class (the lookupClass) as a lookup class
* for method handle creation.
* Must be called by from a method in this package,
* which in turn is called by a method not in this package.
*/
Lookup(Class<?> lookupClass) {
this(lookupClass, null, FULL_POWER_MODES);
}
private Lookup(Class<?> lookupClass, Class<?> prevLookupClass, int allowedModes) {
assert prevLookupClass == null || ((allowedModes & MODULE) == 0
&& prevLookupClass.getModule() != lookupClass.getModule());
assert !lookupClass.isArray() && !lookupClass.isPrimitive();
this.lookupClass = lookupClass;
this.prevLookupClass = prevLookupClass;
this.allowedModes = allowedModes;
}
private static Lookup newLookup(Class<?> lookupClass, Class<?> prevLookupClass, int allowedModes) {
// make sure we haven't accidentally picked up a privileged class:
checkUnprivilegedlookupClass(lookupClass);
return new Lookup(lookupClass, prevLookupClass, allowedModes);
}
/**
* Creates a lookup on the specified new lookup class.
* The resulting object will report the specified
* class as its own {@link #lookupClass() lookupClass}.
*
* <p>
* However, the resulting {@code Lookup} object is guaranteed
* to have no more access capabilities than the original.
* In particular, access capabilities can be lost as follows:<ul>
* <li>If the new lookup class is different from the old lookup class,
* i.e. {@link #ORIGINAL ORIGINAL} access is lost.
* <li>If the new lookup class is in a different module from the old one,
* i.e. {@link #MODULE MODULE} access is lost.
* <li>If the new lookup class is in a different package
* than the old one, protected and default (package) members will not be accessible,
* i.e. {@link #PROTECTED PROTECTED} and {@link #PACKAGE PACKAGE} access are lost.
* <li>If the new lookup class is not within the same package member
* as the old one, private members will not be accessible, and protected members
* will not be accessible by virtue of inheritance,
* i.e. {@link #PRIVATE PRIVATE} access is lost.
* (Protected members may continue to be accessible because of package sharing.)
* <li>If the new lookup class is not
* {@linkplain #accessClass(Class) accessible} to this lookup,
* then no members, not even public members, will be accessible
* i.e. all access modes are lost.
* <li>If the new lookup class, the old lookup class and the previous lookup class
* are all in different modules i.e. teleporting to a third module,
* all access modes are lost.
* </ul>
* <p>
* The new previous lookup class is chosen as follows:
* <ul>
* <li>If the new lookup object has {@link #UNCONDITIONAL UNCONDITIONAL} bit,
* the new previous lookup class is {@code null}.
* <li>If the new lookup class is in the same module as the old lookup class,
* the new previous lookup class is the old previous lookup class.
* <li>If the new lookup class is in a different module from the old lookup class,
* the new previous lookup class is the old lookup class.
*</ul>
* <p>
* The resulting lookup's capabilities for loading classes
* (used during {@link #findClass} invocations)
* are determined by the lookup class' loader,
* which may change due to this operation.
*
* @param requestedLookupClass the desired lookup class for the new lookup object
* @return a lookup object which reports the desired lookup class, or the same object
* if there is no change
* @throws IllegalArgumentException if {@code requestedLookupClass} is a primitive type or void or array class
* @throws NullPointerException if the argument is null
*
* @see #accessClass(Class)
* @see <a href="#cross-module-lookup">Cross-module lookups</a>
*/
public Lookup in(Class<?> requestedLookupClass) {
Objects.requireNonNull(requestedLookupClass);
if (requestedLookupClass.isPrimitive())
throw new IllegalArgumentException(requestedLookupClass + " is a primitive class");
if (requestedLookupClass.isArray())
throw new IllegalArgumentException(requestedLookupClass + " is an array class");
if (allowedModes == TRUSTED) // IMPL_LOOKUP can make any lookup at all
return new Lookup(requestedLookupClass, null, FULL_POWER_MODES);
if (requestedLookupClass == this.lookupClass)
return this; // keep same capabilities
int newModes = (allowedModes & FULL_POWER_MODES) & ~ORIGINAL;
Module fromModule = this.lookupClass.getModule();
Module targetModule = requestedLookupClass.getModule();
Class<?> plc = this.previousLookupClass();
if ((this.allowedModes & UNCONDITIONAL) != 0) {
assert plc == null;
newModes = UNCONDITIONAL;
} else if (fromModule != targetModule) {
if (plc != null && !VerifyAccess.isSameModule(plc, requestedLookupClass)) {
// allow hopping back and forth between fromModule and plc's module
// but not the third module
newModes = 0;
}
// drop MODULE access
newModes &= ~(MODULE|PACKAGE|PRIVATE|PROTECTED);
// teleport from this lookup class
plc = this.lookupClass;
}
if ((newModes & PACKAGE) != 0
&& !VerifyAccess.isSamePackage(this.lookupClass, requestedLookupClass)) {
newModes &= ~(PACKAGE|PRIVATE|PROTECTED);
}
// Allow nestmate lookups to be created without special privilege:
if ((newModes & PRIVATE) != 0
&& !VerifyAccess.isSamePackageMember(this.lookupClass, requestedLookupClass)) {
newModes &= ~(PRIVATE|PROTECTED);
}
if ((newModes & (PUBLIC|UNCONDITIONAL)) != 0
&& !VerifyAccess.isClassAccessible(requestedLookupClass, this.lookupClass, this.prevLookupClass, allowedModes)) {
// The requested class it not accessible from the lookup class.
// No permissions.
newModes = 0;
}
return newLookup(requestedLookupClass, plc, newModes);
}
/**
* Creates a lookup on the same lookup class which this lookup object
* finds members, but with a lookup mode that has lost the given lookup mode.
* The lookup mode to drop is one of {@link #PUBLIC PUBLIC}, {@link #MODULE
* MODULE}, {@link #PACKAGE PACKAGE}, {@link #PROTECTED PROTECTED},
* {@link #PRIVATE PRIVATE}, {@link #ORIGINAL ORIGINAL}, or
* {@link #UNCONDITIONAL UNCONDITIONAL}.
*
* <p> If this lookup is a {@linkplain MethodHandles#publicLookup() public lookup},
* this lookup has {@code UNCONDITIONAL} mode set and it has no other mode set.
* When dropping {@code UNCONDITIONAL} on a public lookup then the resulting
* lookup has no access.
*
* <p> If this lookup is not a public lookup, then the following applies
* regardless of its {@linkplain #lookupModes() lookup modes}.
* {@link #PROTECTED PROTECTED} and {@link #ORIGINAL ORIGINAL} are always
* dropped and so the resulting lookup mode will never have these access
* capabilities. When dropping {@code PACKAGE}
* then the resulting lookup will not have {@code PACKAGE} or {@code PRIVATE}
* access. When dropping {@code MODULE} then the resulting lookup will not
* have {@code MODULE}, {@code PACKAGE}, or {@code PRIVATE} access.
* When dropping {@code PUBLIC} then the resulting lookup has no access.
*
* @apiNote
* A lookup with {@code PACKAGE} but not {@code PRIVATE} mode can safely
* delegate non-public access within the package of the lookup class without
* conferring <a href="MethodHandles.Lookup.html#privacc">private access</a>.
* A lookup with {@code MODULE} but not
* {@code PACKAGE} mode can safely delegate {@code PUBLIC} access within
* the module of the lookup class without conferring package access.
* A lookup with a {@linkplain #previousLookupClass() previous lookup class}
* (and {@code PUBLIC} but not {@code MODULE} mode) can safely delegate access
* to public classes accessible to both the module of the lookup class
* and the module of the previous lookup class.
*
* @param modeToDrop the lookup mode to drop
* @return a lookup object which lacks the indicated mode, or the same object if there is no change
* @throws IllegalArgumentException if {@code modeToDrop} is not one of {@code PUBLIC},
* {@code MODULE}, {@code PACKAGE}, {@code PROTECTED}, {@code PRIVATE}, {@code ORIGINAL}
* or {@code UNCONDITIONAL}
* @see MethodHandles#privateLookupIn
* @since 9
*/
public Lookup dropLookupMode(int modeToDrop) {
int oldModes = lookupModes();
int newModes = oldModes & ~(modeToDrop | PROTECTED | ORIGINAL);
switch (modeToDrop) {
case PUBLIC: newModes &= ~(FULL_POWER_MODES); break;
case MODULE: newModes &= ~(PACKAGE | PRIVATE); break;
case PACKAGE: newModes &= ~(PRIVATE); break;
case PROTECTED:
case PRIVATE:
case ORIGINAL:
case UNCONDITIONAL: break;
default: throw new IllegalArgumentException(modeToDrop + " is not a valid mode to drop");
}
if (newModes == oldModes) return this; // return self if no change
return newLookup(lookupClass(), previousLookupClass(), newModes);
}
/**
* Creates and links a class or interface from {@code bytes}
* with the same class loader and in the same runtime package and
* {@linkplain java.security.ProtectionDomain protection domain} as this lookup's
* {@linkplain #lookupClass() lookup class} as if calling
* {@link ClassLoader#defineClass(String,byte[],int,int,ProtectionDomain)
* ClassLoader::defineClass}.
*
* <p> The {@linkplain #lookupModes() lookup modes} for this lookup must include
* {@link #PACKAGE PACKAGE} access as default (package) members will be
* accessible to the class. The {@code PACKAGE} lookup mode serves to authenticate
* that the lookup object was created by a caller in the runtime package (or derived
* from a lookup originally created by suitably privileged code to a target class in
* the runtime package). </p>
*
* <p> The {@code bytes} parameter is the class bytes of a valid class file (as defined
* by the <em>The Java Virtual Machine Specification</em>) with a class name in the
* same package as the lookup class. </p>
*
* <p> This method does not run the class initializer. The class initializer may
* run at a later time, as detailed in section 12.4 of the <em>The Java Language
* Specification</em>. </p>
*
* @param bytes the class bytes
* @return the {@code Class} object for the class
* @throws IllegalAccessException if this lookup does not have {@code PACKAGE} access
* @throws ClassFormatError if {@code bytes} is not a {@code ClassFile} structure
* @throws IllegalArgumentException if {@code bytes} denotes a class in a different package
* than the lookup class or {@code bytes} is not a class or interface
* ({@code ACC_MODULE} flag is set in the value of the {@code access_flags} item)
* @throws VerifyError if the newly created class cannot be verified
* @throws LinkageError if the newly created class cannot be linked for any other reason
* @throws NullPointerException if {@code bytes} is {@code null}
* @since 9
* @see MethodHandles#privateLookupIn
* @see Lookup#dropLookupMode
* @see ClassLoader#defineClass(String,byte[],int,int,ProtectionDomain)
*/
public Class<?> defineClass(byte[] bytes) throws IllegalAccessException {
if ((lookupModes() & PACKAGE) == 0)
throw new IllegalAccessException("Lookup does not have PACKAGE access");
return makeClassDefiner(bytes.clone()).defineClass(false);
}
/**
* The set of class options that specify whether a hidden class created by
* {@link Lookup#defineHiddenClass(byte[], boolean, ClassOption...)
* Lookup::defineHiddenClass} method is dynamically added as a new member
* to the nest of a lookup class and/or whether a hidden class has
* a strong relationship with the class loader marked as its defining loader.
*
* @since 15
*/
public enum ClassOption {
/**
* Specifies that a hidden class be added to {@linkplain Class#getNestHost nest}
* of a lookup class as a nestmate.
*
* <p> A hidden nestmate class has access to the private members of all
* classes and interfaces in the same nest.
*
* @see Class#getNestHost()
*/
NESTMATE(NESTMATE_CLASS),
/**
* Specifies that a hidden class has a <em>strong</em>
* relationship with the class loader marked as its defining loader,
* as a normal class or interface has with its own defining loader.
* This means that the hidden class may be unloaded if and only if
* its defining loader is not reachable and thus may be reclaimed
* by a garbage collector (JLS {@jls 12.7}).
*
* <p> By default, a hidden class or interface may be unloaded
* even if the class loader that is marked as its defining loader is
* <a href="../ref/package-summary.html#reachability">reachable</a>.
*
* @jls 12.7 Unloading of Classes and Interfaces
*/
STRONG(STRONG_LOADER_LINK);
/* the flag value is used by VM at define class time */
private final int flag;
ClassOption(int flag) {
this.flag = flag;
}
static int optionsToFlag(ClassOption[] options) {
int flags = 0;
for (ClassOption cp : options) {
if ((flags & cp.flag) != 0) {
throw new IllegalArgumentException("Duplicate ClassOption " + cp);
}
flags |= cp.flag;
}
return flags;
}
}
/**
* Creates a <em>hidden</em> class or interface from {@code bytes},
* returning a {@code Lookup} on the newly created class or interface.
*
* <p> Ordinarily, a class or interface {@code C} is created by a class loader,
* which either defines {@code C} directly or delegates to another class loader.
* A class loader defines {@code C} directly by invoking
* {@link ClassLoader#defineClass(String, byte[], int, int, ProtectionDomain)
* ClassLoader::defineClass}, which causes the Java Virtual Machine
* to derive {@code C} from a purported representation in {@code class} file format.
* In situations where use of a class loader is undesirable, a class or interface
* {@code C} can be created by this method instead. This method is capable of
* defining {@code C}, and thereby creating it, without invoking
* {@code ClassLoader::defineClass}.
* Instead, this method defines {@code C} as if by arranging for
* the Java Virtual Machine to derive a nonarray class or interface {@code C}
* from a purported representation in {@code class} file format
* using the following rules:
*
* <ol>
* <li> The {@linkplain #lookupModes() lookup modes} for this {@code Lookup}
* must include {@linkplain #hasFullPrivilegeAccess() full privilege} access.
* This level of access is needed to create {@code C} in the module
* of the lookup class of this {@code Lookup}.</li>
*
* <li> The purported representation in {@code bytes} must be a {@code ClassFile}
* structure (JVMS {@jvms 4.1}) of a supported major and minor version.
* The major and minor version may differ from the {@code class} file version
* of the lookup class of this {@code Lookup}.</li>
*
* <li> The value of {@code this_class} must be a valid index in the
* {@code constant_pool} table, and the entry at that index must be a valid
* {@code CONSTANT_Class_info} structure. Let {@code N} be the binary name
* encoded in internal form that is specified by this structure. {@code N} must
* denote a class or interface in the same package as the lookup class.</li>
*
* <li> Let {@code CN} be the string {@code N + "." + <suffix>},
* where {@code <suffix>} is an unqualified name.
*
* <p> Let {@code newBytes} be the {@code ClassFile} structure given by
* {@code bytes} with an additional entry in the {@code constant_pool} table,
* indicating a {@code CONSTANT_Utf8_info} structure for {@code CN}, and
* where the {@code CONSTANT_Class_info} structure indicated by {@code this_class}
* refers to the new {@code CONSTANT_Utf8_info} structure.
*
* <p> Let {@code L} be the defining class loader of the lookup class of this {@code Lookup}.
*
* <p> {@code C} is derived with name {@code CN}, class loader {@code L}, and
* purported representation {@code newBytes} as if by the rules of JVMS {@jvms 5.3.5},
* with the following adjustments:
* <ul>
* <li> The constant indicated by {@code this_class} is permitted to specify a name
* that includes a single {@code "."} character, even though this is not a valid
* binary class or interface name in internal form.</li>
*
* <li> The Java Virtual Machine marks {@code L} as the defining class loader of {@code C},
* but no class loader is recorded as an initiating class loader of {@code C}.</li>
*
* <li> {@code C} is considered to have the same runtime
* {@linkplain Class#getPackage() package}, {@linkplain Class#getModule() module}
* and {@linkplain java.security.ProtectionDomain protection domain}
* as the lookup class of this {@code Lookup}.
* <li> Let {@code GN} be the binary name obtained by taking {@code N}
* (a binary name encoded in internal form) and replacing ASCII forward slashes with
* ASCII periods. For the instance of {@link java.lang.Class} representing {@code C}:
* <ul>
* <li> {@link Class#getName()} returns the string {@code GN + "/" + <suffix>},
* even though this is not a valid binary class or interface name.</li>
* <li> {@link Class#descriptorString()} returns the string
* {@code "L" + N + "." + <suffix> + ";"},
* even though this is not a valid type descriptor name.</li>
* <li> {@link Class#describeConstable()} returns an empty optional as {@code C}
* cannot be described in {@linkplain java.lang.constant.ClassDesc nominal form}.</li>
* </ul>
* </ul>
* </li>
* </ol>
*
* <p> After {@code C} is derived, it is linked by the Java Virtual Machine.
* Linkage occurs as specified in JVMS {@jvms 5.4.3}, with the following adjustments:
* <ul>
* <li> During verification, whenever it is necessary to load the class named
* {@code CN}, the attempt succeeds, producing class {@code C}. No request is
* made of any class loader.</li>
*
* <li> On any attempt to resolve the entry in the run-time constant pool indicated
* by {@code this_class}, the symbolic reference is considered to be resolved to
* {@code C} and resolution always succeeds immediately.</li>
* </ul>
*
* <p> If the {@code initialize} parameter is {@code true},
* then {@code C} is initialized by the Java Virtual Machine.
*
* <p> The newly created class or interface {@code C} serves as the
* {@linkplain #lookupClass() lookup class} of the {@code Lookup} object
* returned by this method. {@code C} is <em>hidden</em> in the sense that
* no other class or interface can refer to {@code C} via a constant pool entry.
* That is, a hidden class or interface cannot be named as a supertype, a field type,
* a method parameter type, or a method return type by any other class.
* This is because a hidden class or interface does not have a binary name, so
* there is no internal form available to record in any class's constant pool.
* A hidden class or interface is not discoverable by {@link Class#forName(String, boolean, ClassLoader)},
* {@link ClassLoader#loadClass(String, boolean)}, or {@link #findClass(String)}, and
* is not {@linkplain java.instrument/java.lang.instrument.Instrumentation#isModifiableClass(Class)
* modifiable} by Java agents or tool agents using the <a href="{@docRoot}/../specs/jvmti.html">
* JVM Tool Interface</a>.
*
* <p> A class or interface created by
* {@linkplain ClassLoader#defineClass(String, byte[], int, int, ProtectionDomain)
* a class loader} has a strong relationship with that class loader.
* That is, every {@code Class} object contains a reference to the {@code ClassLoader}
* that {@linkplain Class#getClassLoader() defined it}.
* This means that a class created by a class loader may be unloaded if and
* only if its defining loader is not reachable and thus may be reclaimed
* by a garbage collector (JLS {@jls 12.7}).
*
* By default, however, a hidden class or interface may be unloaded even if
* the class loader that is marked as its defining loader is
* <a href="../ref/package-summary.html#reachability">reachable</a>.
* This behavior is useful when a hidden class or interface serves multiple
* classes defined by arbitrary class loaders. In other cases, a hidden
* class or interface may be linked to a single class (or a small number of classes)
* with the same defining loader as the hidden class or interface.
* In such cases, where the hidden class or interface must be coterminous
* with a normal class or interface, the {@link ClassOption#STRONG STRONG}
* option may be passed in {@code options}.
* This arranges for a hidden class to have the same strong relationship
* with the class loader marked as its defining loader,
* as a normal class or interface has with its own defining loader.
*
* If {@code STRONG} is not used, then the invoker of {@code defineHiddenClass}
* may still prevent a hidden class or interface from being
* unloaded by ensuring that the {@code Class} object is reachable.
*
* <p> The unloading characteristics are set for each hidden class when it is
* defined, and cannot be changed later. An advantage of allowing hidden classes
* to be unloaded independently of the class loader marked as their defining loader
* is that a very large number of hidden classes may be created by an application.
* In contrast, if {@code STRONG} is used, then the JVM may run out of memory,
* just as if normal classes were created by class loaders.
*
* <p> Classes and interfaces in a nest are allowed to have mutual access to
* their private members. The nest relationship is determined by
* the {@code NestHost} attribute (JVMS {@jvms 4.7.28}) and
* the {@code NestMembers} attribute (JVMS {@jvms 4.7.29}) in a {@code class} file.
* By default, a hidden class belongs to a nest consisting only of itself
* because a hidden class has no binary name.
* The {@link ClassOption#NESTMATE NESTMATE} option can be passed in {@code options}
* to create a hidden class or interface {@code C} as a member of a nest.
* The nest to which {@code C} belongs is not based on any {@code NestHost} attribute
* in the {@code ClassFile} structure from which {@code C} was derived.
* Instead, the following rules determine the nest host of {@code C}:
* <ul>
* <li>If the nest host of the lookup class of this {@code Lookup} has previously
* been determined, then let {@code H} be the nest host of the lookup class.
* Otherwise, the nest host of the lookup class is determined using the
* algorithm in JVMS {@jvms 5.4.4}, yielding {@code H}.</li>
* <li>The nest host of {@code C} is determined to be {@code H},
* the nest host of the lookup class.</li>
* </ul>
*
* <p> A hidden class or interface may be serializable, but this requires a custom
* serialization mechanism in order to ensure that instances are properly serialized
* and deserialized. The default serialization mechanism supports only classes and
* interfaces that are discoverable by their class name.
*
* @param bytes the bytes that make up the class data,
* in the format of a valid {@code class} file as defined by
* <cite>The Java Virtual Machine Specification</cite>.
* @param initialize if {@code true} the class will be initialized.
* @param options {@linkplain ClassOption class options}
* @return the {@code Lookup} object on the hidden class,
* with {@linkplain #ORIGINAL original} and
* {@linkplain Lookup#hasFullPrivilegeAccess() full privilege} access
*
* @throws IllegalAccessException if this {@code Lookup} does not have
* {@linkplain #hasFullPrivilegeAccess() full privilege} access
* @throws ClassFormatError if {@code bytes} is not a {@code ClassFile} structure
* @throws UnsupportedClassVersionError if {@code bytes} is not of a supported major or minor version
* @throws IllegalArgumentException if {@code bytes} denotes a class in a different package
* than the lookup class or {@code bytes} is not a class or interface
* ({@code ACC_MODULE} flag is set in the value of the {@code access_flags} item)
* @throws IncompatibleClassChangeError if the class or interface named as
* the direct superclass of {@code C} is in fact an interface, or if any of the classes
* or interfaces named as direct superinterfaces of {@code C} are not in fact interfaces
* @throws ClassCircularityError if any of the superclasses or superinterfaces of
* {@code C} is {@code C} itself
* @throws VerifyError if the newly created class cannot be verified
* @throws LinkageError if the newly created class cannot be linked for any other reason
* @throws NullPointerException if any parameter is {@code null}
*
* @since 15
* @see Class#isHidden()
* @jvms 4.2.1 Binary Class and Interface Names
* @jvms 4.2.2 Unqualified Names
* @jvms 4.7.28 The {@code NestHost} Attribute
* @jvms 4.7.29 The {@code NestMembers} Attribute
* @jvms 5.4.3.1 Class and Interface Resolution
* @jvms 5.4.4 Access Control
* @jvms 5.3.5 Deriving a {@code Class} from a {@code class} File Representation
* @jvms 5.4 Linking
* @jvms 5.5 Initialization
* @jls 12.7 Unloading of Classes and Interfaces
*/
@SuppressWarnings("doclint:reference") // cross-module links
public Lookup defineHiddenClass(byte[] bytes, boolean initialize, ClassOption... options)
throws IllegalAccessException
{
Objects.requireNonNull(bytes);
int flags = ClassOption.optionsToFlag(options);
if (!hasFullPrivilegeAccess()) {
throw new IllegalAccessException(this + " does not have full privilege access");
}
return makeHiddenClassDefiner(bytes.clone(), false, flags).defineClassAsLookup(initialize);
}
/**
* Creates a <em>hidden</em> class or interface from {@code bytes} with associated
* {@linkplain MethodHandles#classData(Lookup, String, Class) class data},
* returning a {@code Lookup} on the newly created class or interface.
*
* <p> This method is equivalent to calling
* {@link #defineHiddenClass(byte[], boolean, ClassOption...) defineHiddenClass(bytes, initialize, options)}
* as if the hidden class is injected with a private static final <i>unnamed</i>
* field which is initialized with the given {@code classData} at
* the first instruction of the class initializer.
* The newly created class is linked by the Java Virtual Machine.
*
* <p> The {@link MethodHandles#classData(Lookup, String, Class) MethodHandles::classData}
* and {@link MethodHandles#classDataAt(Lookup, String, Class, int) MethodHandles::classDataAt}
* methods can be used to retrieve the {@code classData}.
*
* @apiNote
* A framework can create a hidden class with class data with one or more
* objects and load the class data as dynamically-computed constant(s)
* via a bootstrap method. {@link MethodHandles#classData(Lookup, String, Class)
* Class data} is accessible only to the lookup object created by the newly
* defined hidden class but inaccessible to other members in the same nest
* (unlike private static fields that are accessible to nestmates).
* Care should be taken w.r.t. mutability for example when passing
* an array or other mutable structure through the class data.
* Changing any value stored in the class data at runtime may lead to
* unpredictable behavior.
* If the class data is a {@code List}, it is good practice to make it
* unmodifiable for example via {@link List#of List::of}.
*
* @param bytes the class bytes
* @param classData pre-initialized class data
* @param initialize if {@code true} the class will be initialized.
* @param options {@linkplain ClassOption class options}
* @return the {@code Lookup} object on the hidden class,
* with {@linkplain #ORIGINAL original} and
* {@linkplain Lookup#hasFullPrivilegeAccess() full privilege} access
*
* @throws IllegalAccessException if this {@code Lookup} does not have
* {@linkplain #hasFullPrivilegeAccess() full privilege} access
* @throws ClassFormatError if {@code bytes} is not a {@code ClassFile} structure
* @throws UnsupportedClassVersionError if {@code bytes} is not of a supported major or minor version
* @throws IllegalArgumentException if {@code bytes} denotes a class in a different package
* than the lookup class or {@code bytes} is not a class or interface
* ({@code ACC_MODULE} flag is set in the value of the {@code access_flags} item)
* @throws IncompatibleClassChangeError if the class or interface named as
* the direct superclass of {@code C} is in fact an interface, or if any of the classes
* or interfaces named as direct superinterfaces of {@code C} are not in fact interfaces
* @throws ClassCircularityError if any of the superclasses or superinterfaces of
* {@code C} is {@code C} itself
* @throws VerifyError if the newly created class cannot be verified
* @throws LinkageError if the newly created class cannot be linked for any other reason
* @throws NullPointerException if any parameter is {@code null}
*
* @since 16
* @see Lookup#defineHiddenClass(byte[], boolean, ClassOption...)
* @see Class#isHidden()
* @see MethodHandles#classData(Lookup, String, Class)
* @see MethodHandles#classDataAt(Lookup, String, Class, int)
* @jvms 4.2.1 Binary Class and Interface Names
* @jvms 4.2.2 Unqualified Names
* @jvms 4.7.28 The {@code NestHost} Attribute
* @jvms 4.7.29 The {@code NestMembers} Attribute
* @jvms 5.4.3.1 Class and Interface Resolution
* @jvms 5.4.4 Access Control
* @jvms 5.3.5 Deriving a {@code Class} from a {@code class} File Representation
* @jvms 5.4 Linking
* @jvms 5.5 Initialization
* @jls 12.7 Unloading of Classes and Interfaces
*/
public Lookup defineHiddenClassWithClassData(byte[] bytes, Object classData, boolean initialize, ClassOption... options)
throws IllegalAccessException
{
Objects.requireNonNull(bytes);
Objects.requireNonNull(classData);
int flags = ClassOption.optionsToFlag(options);
if (!hasFullPrivilegeAccess()) {
throw new IllegalAccessException(this + " does not have full privilege access");
}
return makeHiddenClassDefiner(bytes.clone(), false, flags)
.defineClassAsLookup(initialize, classData);
}
// A default dumper for writing class files passed to Lookup::defineClass
// and Lookup::defineHiddenClass to disk for debugging purposes. To enable,
// set -Djdk.invoke.MethodHandle.dumpHiddenClassFiles or
// -Djdk.invoke.MethodHandle.dumpHiddenClassFiles=true
//
// This default dumper does not dump hidden classes defined by LambdaMetafactory
// and LambdaForms and method handle internals. They are dumped via
// different ClassFileDumpers.
private static ClassFileDumper defaultDumper() {
return DEFAULT_DUMPER;
}
private static final ClassFileDumper DEFAULT_DUMPER = ClassFileDumper.getInstance(
"jdk.invoke.MethodHandle.dumpClassFiles", "DUMP_CLASS_FILES");
/**
* This method checks the class file version and the structure of `this_class`.
* and checks if the bytes is a class or interface (ACC_MODULE flag not set)
* that is in the named package.
*
* @throws IllegalArgumentException if ACC_MODULE flag is set in access flags
* or the class is not in the given package name.
*/
static String validateAndFindInternalName(byte[] bytes, String pkgName) {
int magic = readInt(bytes, 0);
if (magic != ClassFile.MAGIC_NUMBER) {
throw new ClassFormatError("Incompatible magic value: " + magic);
}
// We have to read major and minor this way as ClassFile API throws IAE
// yet we want distinct ClassFormatError and UnsupportedClassVersionError
int minor = readUnsignedShort(bytes, 4);
int major = readUnsignedShort(bytes, 6);
if (!VM.isSupportedClassFileVersion(major, minor)) {
throw new UnsupportedClassVersionError("Unsupported class file version " + major + "." + minor);
}
String name;
ClassDesc sym;
int accessFlags;
try {
ClassModel cm = ClassFile.of().parse(bytes);
var thisClass = cm.thisClass();
name = thisClass.asInternalName();
sym = thisClass.asSymbol();
accessFlags = cm.flags().flagsMask();
} catch (IllegalArgumentException e) {
ClassFormatError cfe = new ClassFormatError();
cfe.initCause(e);
throw cfe;
}
// must be a class or interface
if ((accessFlags & ACC_MODULE) != 0) {
throw newIllegalArgumentException("Not a class or interface: ACC_MODULE flag is set");
}
String pn = sym.packageName();
if (!pn.equals(pkgName)) {
throw newIllegalArgumentException(name + " not in same package as lookup class");
}
return name;
}
private static int readInt(byte[] bytes, int offset) {
if ((offset + 4) > bytes.length) {
throw new ClassFormatError("Invalid ClassFile structure");
}
return ((bytes[offset] & 0xFF) << 24)
| ((bytes[offset + 1] & 0xFF) << 16)
| ((bytes[offset + 2] & 0xFF) << 8)
| (bytes[offset + 3] & 0xFF);
}
private static int readUnsignedShort(byte[] bytes, int offset) {
if ((offset+2) > bytes.length) {
throw new ClassFormatError("Invalid ClassFile structure");
}
return ((bytes[offset] & 0xFF) << 8) | (bytes[offset + 1] & 0xFF);
}
/*
* Returns a ClassDefiner that creates a {@code Class} object of a normal class
* from the given bytes.
*
* Caller should make a defensive copy of the arguments if needed
* before calling this factory method.
*
* @throws IllegalArgumentException if {@code bytes} is not a class or interface or
* {@code bytes} denotes a class in a different package than the lookup class
*/
private ClassDefiner makeClassDefiner(byte[] bytes) {
var internalName = validateAndFindInternalName(bytes, lookupClass().getPackageName());
return new ClassDefiner(this, internalName, bytes, STRONG_LOADER_LINK, defaultDumper());
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a normal class
* from the given bytes. No package name check on the given bytes.
*
* @param internalName internal name
* @param bytes class bytes
* @param dumper dumper to write the given bytes to the dumper's output directory
* @return ClassDefiner that defines a normal class of the given bytes.
*/
ClassDefiner makeClassDefiner(String internalName, byte[] bytes, ClassFileDumper dumper) {
// skip package name validation
return new ClassDefiner(this, internalName, bytes, STRONG_LOADER_LINK, dumper);
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a hidden class
* from the given bytes. The name must be in the same package as the lookup class.
*
* Caller should make a defensive copy of the arguments if needed
* before calling this factory method.
*
* @param bytes class bytes
* @param dumper dumper to write the given bytes to the dumper's output directory
* @return ClassDefiner that defines a hidden class of the given bytes.
*
* @throws IllegalArgumentException if {@code bytes} is not a class or interface or
* {@code bytes} denotes a class in a different package than the lookup class
*/
ClassDefiner makeHiddenClassDefiner(byte[] bytes, ClassFileDumper dumper) {
var internalName = validateAndFindInternalName(bytes, lookupClass().getPackageName());
return makeHiddenClassDefiner(internalName, bytes, false, dumper, 0);
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a hidden class
* from the given bytes and options.
* The name must be in the same package as the lookup class.
*
* Caller should make a defensive copy of the arguments if needed
* before calling this factory method.
*
* @param bytes class bytes
* @param flags class option flag mask
* @param accessVmAnnotations true to give the hidden class access to VM annotations
* @return ClassDefiner that defines a hidden class of the given bytes and options
*
* @throws IllegalArgumentException if {@code bytes} is not a class or interface or
* {@code bytes} denotes a class in a different package than the lookup class
*/
private ClassDefiner makeHiddenClassDefiner(byte[] bytes,
boolean accessVmAnnotations,
int flags) {
var internalName = validateAndFindInternalName(bytes, lookupClass().getPackageName());
return makeHiddenClassDefiner(internalName, bytes, accessVmAnnotations, defaultDumper(), flags);
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a hidden class
* from the given bytes and the given options. No package name check on the given bytes.
*
* @param internalName internal name that specifies the prefix of the hidden class
* @param bytes class bytes
* @param dumper dumper to write the given bytes to the dumper's output directory
* @return ClassDefiner that defines a hidden class of the given bytes and options.
*/
ClassDefiner makeHiddenClassDefiner(String internalName, byte[] bytes, ClassFileDumper dumper) {
Objects.requireNonNull(dumper);
// skip name and access flags validation
return makeHiddenClassDefiner(internalName, bytes, false, dumper, 0);
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a hidden class
* from the given bytes and the given options. No package name check on the given bytes.
*
* @param internalName internal name that specifies the prefix of the hidden class
* @param bytes class bytes
* @param flags class options flag mask
* @param dumper dumper to write the given bytes to the dumper's output directory
* @return ClassDefiner that defines a hidden class of the given bytes and options.
*/
ClassDefiner makeHiddenClassDefiner(String internalName, byte[] bytes, ClassFileDumper dumper, int flags) {
Objects.requireNonNull(dumper);
// skip name and access flags validation
return makeHiddenClassDefiner(internalName, bytes, false, dumper, flags);
}
/**
* Returns a ClassDefiner that creates a {@code Class} object of a hidden class
* from the given class file and options.
*
* @param internalName internal name
* @param bytes Class byte array
* @param flags class option flag mask
* @param accessVmAnnotations true to give the hidden class access to VM annotations
* @param dumper dumper to write the given bytes to the dumper's output directory
*/
private ClassDefiner makeHiddenClassDefiner(String internalName,
byte[] bytes,
boolean accessVmAnnotations,
ClassFileDumper dumper,
int flags) {
flags |= HIDDEN_CLASS;
if (accessVmAnnotations | VM.isSystemDomainLoader(lookupClass.getClassLoader())) {
// jdk.internal.vm.annotations are permitted for classes
// defined to boot loader and platform loader
flags |= ACCESS_VM_ANNOTATIONS;
}
return new ClassDefiner(this, internalName, bytes, flags, dumper);
}
record ClassDefiner(Lookup lookup, String internalName, byte[] bytes, int classFlags, ClassFileDumper dumper) {
ClassDefiner {
assert ((classFlags & HIDDEN_CLASS) != 0 || (classFlags & STRONG_LOADER_LINK) == STRONG_LOADER_LINK);
}
Class<?> defineClass(boolean initialize) {
return defineClass(initialize, null);
}
Lookup defineClassAsLookup(boolean initialize) {
Class<?> c = defineClass(initialize, null);
return new Lookup(c, null, FULL_POWER_MODES);
}
/**
* Defines the class of the given bytes and the given classData.
* If {@code initialize} parameter is true, then the class will be initialized.
*
* @param initialize true if the class to be initialized
* @param classData classData or null
* @return the class
*
* @throws LinkageError linkage error
*/
Class<?> defineClass(boolean initialize, Object classData) {
Class<?> lookupClass = lookup.lookupClass();
ClassLoader loader = lookupClass.getClassLoader();
ProtectionDomain pd = (loader != null) ? lookup.lookupClassProtectionDomain() : null;
Class<?> c = null;
try {
c = SharedSecrets.getJavaLangAccess()
.defineClass(loader, lookupClass, internalName, bytes, pd, initialize, classFlags, classData);
assert !isNestmate() || c.getNestHost() == lookupClass.getNestHost();
return c;
} finally {
// dump the classfile for debugging
if (dumper.isEnabled()) {
String name = internalName();
if (c != null) {
dumper.dumpClass(name, c, bytes);
} else {
dumper.dumpFailedClass(name, bytes);
}
}
}
}
/**
* Defines the class of the given bytes and the given classData.
* If {@code initialize} parameter is true, then the class will be initialized.
*
* @param initialize true if the class to be initialized
* @param classData classData or null
* @return a Lookup for the defined class
*
* @throws LinkageError linkage error
*/
Lookup defineClassAsLookup(boolean initialize, Object classData) {
Class<?> c = defineClass(initialize, classData);
return new Lookup(c, null, FULL_POWER_MODES);
}
private boolean isNestmate() {
return (classFlags & NESTMATE_CLASS) != 0;
}
}
private ProtectionDomain lookupClassProtectionDomain() {
ProtectionDomain pd = cachedProtectionDomain;
if (pd == null) {
cachedProtectionDomain = pd = SharedSecrets.getJavaLangAccess().protectionDomain(lookupClass);
}
return pd;
}
// cached protection domain
private volatile ProtectionDomain cachedProtectionDomain;
// Make sure outer class is initialized first.
static { IMPL_NAMES.getClass(); }
/** Package-private version of lookup which is trusted. */
static final Lookup IMPL_LOOKUP = new Lookup(Object.class, null, TRUSTED);
/** Version of lookup which is trusted minimally.
* It can only be used to create method handles to publicly accessible
* members in packages that are exported unconditionally.
*/
static final Lookup PUBLIC_LOOKUP = new Lookup(Object.class, null, UNCONDITIONAL);
private static void checkUnprivilegedlookupClass(Class<?> lookupClass) {
String name = lookupClass.getName();
if (name.startsWith("java.lang.invoke."))
throw newIllegalArgumentException("illegal lookupClass: "+lookupClass);
}
/**
* Displays the name of the class from which lookups are to be made,
* followed by "/" and the name of the {@linkplain #previousLookupClass()
* previous lookup class} if present.
* (The name is the one reported by {@link java.lang.Class#getName() Class.getName}.)
* If there are restrictions on the access permitted to this lookup,
* this is indicated by adding a suffix to the class name, consisting
* of a slash and a keyword. The keyword represents the strongest
* allowed access, and is chosen as follows:
* <ul>
* <li>If no access is allowed, the suffix is "/noaccess".
* <li>If only unconditional access is allowed, the suffix is "/publicLookup".
* <li>If only public access to types in exported packages is allowed, the suffix is "/public".
* <li>If only public and module access are allowed, the suffix is "/module".
* <li>If public and package access are allowed, the suffix is "/package".
* <li>If public, package, and private access are allowed, the suffix is "/private".
* </ul>
* If none of the above cases apply, it is the case that
* {@linkplain #hasFullPrivilegeAccess() full privilege access}
* (public, module, package, private, and protected) is allowed.
* In this case, no suffix is added.
* This is true only of an object obtained originally from
* {@link java.lang.invoke.MethodHandles#lookup MethodHandles.lookup}.
* Objects created by {@link java.lang.invoke.MethodHandles.Lookup#in Lookup.in}
* always have restricted access, and will display a suffix.
* <p>
* (It may seem strange that protected access should be
* stronger than private access. Viewed independently from
* package access, protected access is the first to be lost,
* because it requires a direct subclass relationship between
* caller and callee.)
* @see #in
*/
@Override
public String toString() {
String cname = lookupClass.getName();
if (prevLookupClass != null)
cname += "/" + prevLookupClass.getName();
switch (allowedModes) {
case 0: // no privileges
return cname + "/noaccess";
case UNCONDITIONAL:
return cname + "/publicLookup";
case PUBLIC:
return cname + "/public";
case PUBLIC|MODULE:
return cname + "/module";
case PUBLIC|PACKAGE:
case PUBLIC|MODULE|PACKAGE:
return cname + "/package";
case PUBLIC|PACKAGE|PRIVATE:
case PUBLIC|MODULE|PACKAGE|PRIVATE:
return cname + "/private";
case PUBLIC|PACKAGE|PRIVATE|PROTECTED:
case PUBLIC|MODULE|PACKAGE|PRIVATE|PROTECTED:
case FULL_POWER_MODES:
return cname;
case TRUSTED:
return "/trusted"; // internal only; not exported
default: // Should not happen, but it's a bitfield...
cname = cname + "/" + Integer.toHexString(allowedModes);
assert(false) : cname;
return cname;
}
}
/**
* Produces a method handle for a static method.
* The type of the method handle will be that of the method.
* (Since static methods do not take receivers, there is no
* additional receiver argument inserted into the method handle type,
* as there would be with {@link #findVirtual findVirtual} or {@link #findSpecial findSpecial}.)
* The method and all its argument types must be accessible to the lookup object.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set.
* <p>
* If the returned method handle is invoked, the method's class will
* be initialized, if it has not already been initialized.
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle MH_asList = publicLookup().findStatic(Arrays.class,
"asList", methodType(List.class, Object[].class));
assertEquals("[x, y]", MH_asList.invoke("x", "y").toString());
* }
* @param refc the class from which the method is accessed
* @param name the name of the method
* @param type the type of the method
* @return the desired method handle
* @throws NoSuchMethodException if the method does not exist
* @throws IllegalAccessException if access checking fails,
* or if the method is not {@code static},
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
*/
public MethodHandle findStatic(Class<?> refc, String name, MethodType type) throws NoSuchMethodException, IllegalAccessException {
MemberName method = resolveOrFail(REF_invokeStatic, refc, name, type);
return getDirectMethod(REF_invokeStatic, refc, method, findBoundCallerLookup(method));
}
/**
* Produces a method handle for a virtual method.
* The type of the method handle will be that of the method,
* with the receiver type (usually {@code refc}) prepended.
* The method and all its argument types must be accessible to the lookup object.
* <p>
* When called, the handle will treat the first argument as a receiver
* and, for non-private methods, dispatch on the receiver's type to determine which method
* implementation to enter.
* For private methods the named method in {@code refc} will be invoked on the receiver.
* (The dispatching action is identical with that performed by an
* {@code invokevirtual} or {@code invokeinterface} instruction.)
* <p>
* The first argument will be of type {@code refc} if the lookup
* class has full privileges to access the member. Otherwise
* the member must be {@code protected} and the first argument
* will be restricted in type to the lookup class.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set.
* <p>
* Because of the general <a href="MethodHandles.Lookup.html#equiv">equivalence</a> between {@code invokevirtual}
* instructions and method handles produced by {@code findVirtual},
* if the class is {@code MethodHandle} and the name string is
* {@code invokeExact} or {@code invoke}, the resulting
* method handle is equivalent to one produced by
* {@link java.lang.invoke.MethodHandles#exactInvoker MethodHandles.exactInvoker} or
* {@link java.lang.invoke.MethodHandles#invoker MethodHandles.invoker}
* with the same {@code type} argument.
* <p>
* If the class is {@code VarHandle} and the name string corresponds to
* the name of a signature-polymorphic access mode method, the resulting
* method handle is equivalent to one produced by
* {@link java.lang.invoke.MethodHandles#varHandleInvoker} with
* the access mode corresponding to the name string and with the same
* {@code type} arguments.
* <p>
* <b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle MH_concat = publicLookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
MethodHandle MH_hashCode = publicLookup().findVirtual(Object.class,
"hashCode", methodType(int.class));
MethodHandle MH_hashCode_String = publicLookup().findVirtual(String.class,
"hashCode", methodType(int.class));
assertEquals("xy", (String) MH_concat.invokeExact("x", "y"));
assertEquals("xy".hashCode(), (int) MH_hashCode.invokeExact((Object)"xy"));
assertEquals("xy".hashCode(), (int) MH_hashCode_String.invokeExact("xy"));
// interface method:
MethodHandle MH_subSequence = publicLookup().findVirtual(CharSequence.class,
"subSequence", methodType(CharSequence.class, int.class, int.class));
assertEquals("def", MH_subSequence.invoke("abcdefghi", 3, 6).toString());
// constructor "internal method" must be accessed differently:
MethodType MT_newString = methodType(void.class); //()V for new String()
try { assertEquals("impossible", lookup()
.findVirtual(String.class, "<init>", MT_newString));
} catch (NoSuchMethodException ex) { } // OK
MethodHandle MH_newString = publicLookup()
.findConstructor(String.class, MT_newString);
assertEquals("", (String) MH_newString.invokeExact());
* }
*
* @param refc the class or interface from which the method is accessed
* @param name the name of the method
* @param type the type of the method, with the receiver argument omitted
* @return the desired method handle
* @throws NoSuchMethodException if the method does not exist
* @throws IllegalAccessException if access checking fails,
* or if the method is {@code static},
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
*/
public MethodHandle findVirtual(Class<?> refc, String name, MethodType type) throws NoSuchMethodException, IllegalAccessException {
if (refc == MethodHandle.class) {
MethodHandle mh = findVirtualForMH(name, type);
if (mh != null) return mh;
} else if (refc == VarHandle.class) {
MethodHandle mh = findVirtualForVH(name, type);
if (mh != null) return mh;
}
byte refKind = (refc.isInterface() ? REF_invokeInterface : REF_invokeVirtual);
MemberName method = resolveOrFail(refKind, refc, name, type);
return getDirectMethod(refKind, refc, method, findBoundCallerLookup(method));
}
private MethodHandle findVirtualForMH(String name, MethodType type) {
// these names require special lookups because of the implicit MethodType argument
if ("invoke".equals(name))
return invoker(type);
if ("invokeExact".equals(name))
return exactInvoker(type);
assert(!MemberName.isMethodHandleInvokeName(name));
return null;
}
private MethodHandle findVirtualForVH(String name, MethodType type) {
try {
return varHandleInvoker(VarHandle.AccessMode.valueFromMethodName(name), type);
} catch (IllegalArgumentException e) {
return null;
}
}
/**
* Produces a method handle which creates an object and initializes it, using
* the constructor of the specified type.
* The parameter types of the method handle will be those of the constructor,
* while the return type will be a reference to the constructor's class.
* The constructor and all its argument types must be accessible to the lookup object.
* <p>
* The requested type must have a return type of {@code void}.
* (This is consistent with the JVM's treatment of constructor type descriptors.)
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the constructor's variable arity modifier bit ({@code 0x0080}) is set.
* <p>
* If the returned method handle is invoked, the constructor's class will
* be initialized, if it has not already been initialized.
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle MH_newArrayList = publicLookup().findConstructor(
ArrayList.class, methodType(void.class, Collection.class));
Collection orig = Arrays.asList("x", "y");
Collection copy = (ArrayList) MH_newArrayList.invokeExact(orig);
assert(orig != copy);
assertEquals(orig, copy);
// a variable-arity constructor:
MethodHandle MH_newProcessBuilder = publicLookup().findConstructor(
ProcessBuilder.class, methodType(void.class, String[].class));
ProcessBuilder pb = (ProcessBuilder)
MH_newProcessBuilder.invoke("x", "y", "z");
assertEquals("[x, y, z]", pb.command().toString());
* }
* @param refc the class or interface from which the method is accessed
* @param type the type of the method, with the receiver argument omitted, and a void return type
* @return the desired method handle
* @throws NoSuchMethodException if the constructor does not exist
* @throws IllegalAccessException if access checking fails
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
*/
public MethodHandle findConstructor(Class<?> refc, MethodType type) throws NoSuchMethodException, IllegalAccessException {
if (refc.isArray()) {
throw new NoSuchMethodException("no constructor for array class: " + refc.getName());
}
String name = ConstantDescs.INIT_NAME;
MemberName ctor = resolveOrFail(REF_newInvokeSpecial, refc, name, type);
return getDirectConstructor(refc, ctor);
}
/**
* Looks up a class by name from the lookup context defined by this {@code Lookup} object,
* <a href="MethodHandles.Lookup.html#equiv">as if resolved</a> by an {@code ldc} instruction.
* Such a resolution, as specified in JVMS {@jvms 5.4.3.1}, attempts to locate and load the class,
* and then determines whether the class is accessible to this lookup object.
* <p>
* For a class or an interface, the name is the {@linkplain ClassLoader##binary-name binary name}.
* For an array class of {@code n} dimensions, the name begins with {@code n} occurrences
* of {@code '['} and followed by the element type as encoded in the
* {@linkplain Class##nameFormat table} specified in {@link Class#getName}.
* <p>
* The lookup context here is determined by the {@linkplain #lookupClass() lookup class},
* its class loader, and the {@linkplain #lookupModes() lookup modes}.
*
* @param targetName the {@linkplain ClassLoader##binary-name binary name} of the class
* or the string representing an array class
* @return the requested class.
* @throws LinkageError if the linkage fails
* @throws ClassNotFoundException if the class cannot be loaded by the lookup class' loader.
* @throws IllegalAccessException if the class is not accessible, using the allowed access
* modes.
* @throws NullPointerException if {@code targetName} is null
* @since 9
* @jvms 5.4.3.1 Class and Interface Resolution
*/
public Class<?> findClass(String targetName) throws ClassNotFoundException, IllegalAccessException {
Class<?> targetClass = Class.forName(targetName, false, lookupClass.getClassLoader());
return accessClass(targetClass);
}
/**
* Ensures that {@code targetClass} has been initialized. The class
* to be initialized must be {@linkplain #accessClass accessible}
* to this {@code Lookup} object. This method causes {@code targetClass}
* to be initialized if it has not been already initialized,
* as specified in JVMS {@jvms 5.5}.
*
* <p>
* This method returns when {@code targetClass} is fully initialized, or
* when {@code targetClass} is being initialized by the current thread.
*
* @param <T> the type of the class to be initialized
* @param targetClass the class to be initialized
* @return {@code targetClass} that has been initialized, or that is being
* initialized by the current thread.
*
* @throws IllegalArgumentException if {@code targetClass} is a primitive type or {@code void}
* or array class
* @throws IllegalAccessException if {@code targetClass} is not
* {@linkplain #accessClass accessible} to this lookup
* @throws ExceptionInInitializerError if the class initialization provoked
* by this method fails
* @since 15
* @jvms 5.5 Initialization
*/
public <T> Class<T> ensureInitialized(Class<T> targetClass) throws IllegalAccessException {
if (targetClass.isPrimitive())
throw new IllegalArgumentException(targetClass + " is a primitive class");
if (targetClass.isArray())
throw new IllegalArgumentException(targetClass + " is an array class");
if (!VerifyAccess.isClassAccessible(targetClass, lookupClass, prevLookupClass, allowedModes)) {
throw makeAccessException(targetClass);
}
// ensure class initialization
Unsafe.getUnsafe().ensureClassInitialized(targetClass);
return targetClass;
}
/*
* Returns IllegalAccessException due to access violation to the given targetClass.
*
* This method is called by {@link Lookup#accessClass} and {@link Lookup#ensureInitialized}
* which verifies access to a class rather a member.
*/
private IllegalAccessException makeAccessException(Class<?> targetClass) {
String message = "access violation: "+ targetClass;
if (this == MethodHandles.publicLookup()) {
message += ", from public Lookup";
} else {
Module m = lookupClass().getModule();
message += ", from " + lookupClass() + " (" + m + ")";
if (prevLookupClass != null) {
message += ", previous lookup " +
prevLookupClass.getName() + " (" + prevLookupClass.getModule() + ")";
}
}
return new IllegalAccessException(message);
}
/**
* Determines if a class can be accessed from the lookup context defined by
* this {@code Lookup} object. The static initializer of the class is not run.
* If {@code targetClass} is an array class, {@code targetClass} is accessible
* if the element type of the array class is accessible. Otherwise,
* {@code targetClass} is determined as accessible as follows.
*
* <p>
* If {@code targetClass} is in the same module as the lookup class,
* the lookup class is {@code LC} in module {@code M1} and
* the previous lookup class is in module {@code M0} or
* {@code null} if not present,
* {@code targetClass} is accessible if and only if one of the following is true:
* <ul>
* <li>If this lookup has {@link #PRIVATE} access, {@code targetClass} is
* {@code LC} or other class in the same nest of {@code LC}.</li>
* <li>If this lookup has {@link #PACKAGE} access, {@code targetClass} is
* in the same runtime package of {@code LC}.</li>
* <li>If this lookup has {@link #MODULE} access, {@code targetClass} is
* a public type in {@code M1}.</li>
* <li>If this lookup has {@link #PUBLIC} access, {@code targetClass} is
* a public type in a package exported by {@code M1} to at least {@code M0}
* if the previous lookup class is present; otherwise, {@code targetClass}
* is a public type in a package exported by {@code M1} unconditionally.</li>
* </ul>
*
* <p>
* Otherwise, if this lookup has {@link #UNCONDITIONAL} access, this lookup
* can access public types in all modules when the type is in a package
* that is exported unconditionally.
* <p>
* Otherwise, {@code targetClass} is in a different module from {@code lookupClass},
* and if this lookup does not have {@code PUBLIC} access, {@code lookupClass}
* is inaccessible.
* <p>
* Otherwise, if this lookup has no {@linkplain #previousLookupClass() previous lookup class},
* {@code M1} is the module containing {@code lookupClass} and
* {@code M2} is the module containing {@code targetClass},
* then {@code targetClass} is accessible if and only if
* <ul>
* <li>{@code M1} reads {@code M2}, and
* <li>{@code targetClass} is public and in a package exported by
* {@code M2} at least to {@code M1}.
* </ul>
* <p>
* Otherwise, if this lookup has a {@linkplain #previousLookupClass() previous lookup class},
* {@code M1} and {@code M2} are as before, and {@code M0} is the module
* containing the previous lookup class, then {@code targetClass} is accessible
* if and only if one of the following is true:
* <ul>
* <li>{@code targetClass} is in {@code M0} and {@code M1}
* {@linkplain Module#canRead(Module)} reads} {@code M0} and the type is
* in a package that is exported to at least {@code M1}.
* <li>{@code targetClass} is in {@code M1} and {@code M0}
* {@linkplain Module#canRead(Module)} reads} {@code M1} and the type is
* in a package that is exported to at least {@code M0}.
* <li>{@code targetClass} is in a third module {@code M2} and both {@code M0}
* and {@code M1} reads {@code M2} and the type is in a package
* that is exported to at least both {@code M0} and {@code M2}.
* </ul>
* <p>
* Otherwise, {@code targetClass} is not accessible.
*
* @param <T> the type of the class to be access-checked
* @param targetClass the class to be access-checked
* @return {@code targetClass} that has been access-checked
* @throws IllegalAccessException if the class is not accessible from the lookup class
* and previous lookup class, if present, using the allowed access modes.
* @throws NullPointerException if {@code targetClass} is {@code null}
* @since 9
* @see <a href="#cross-module-lookup">Cross-module lookups</a>
*/
public <T> Class<T> accessClass(Class<T> targetClass) throws IllegalAccessException {
if (!isClassAccessible(targetClass)) {
throw makeAccessException(targetClass);
}
return targetClass;
}
/**
* Produces an early-bound method handle for a virtual method.
* It will bypass checks for overriding methods on the receiver,
* <a href="MethodHandles.Lookup.html#equiv">as if called</a> from an {@code invokespecial}
* instruction from within the explicitly specified {@code specialCaller}.
* The type of the method handle will be that of the method,
* with a suitably restricted receiver type prepended.
* (The receiver type will be {@code specialCaller} or a subtype.)
* The method and all its argument types must be accessible
* to the lookup object.
* <p>
* Before method resolution,
* if the explicitly specified caller class is not identical with the
* lookup class, or if this lookup object does not have
* <a href="MethodHandles.Lookup.html#privacc">private access</a>
* privileges, the access fails.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set.
* <p style="font-size:smaller;">
* <em>(Note: JVM internal methods named {@value ConstantDescs#INIT_NAME}
* are not visible to this API,
* even though the {@code invokespecial} instruction can refer to them
* in special circumstances. Use {@link #findConstructor findConstructor}
* to access instance initialization methods in a safe manner.)</em>
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
static class Listie extends ArrayList {
public String toString() { return "[wee Listie]"; }
static Lookup lookup() { return MethodHandles.lookup(); }
}
...
// no access to constructor via invokeSpecial:
MethodHandle MH_newListie = Listie.lookup()
.findConstructor(Listie.class, methodType(void.class));
Listie l = (Listie) MH_newListie.invokeExact();
try { assertEquals("impossible", Listie.lookup().findSpecial(
Listie.class, "<init>", methodType(void.class), Listie.class));
} catch (NoSuchMethodException ex) { } // OK
// access to super and self methods via invokeSpecial:
MethodHandle MH_super = Listie.lookup().findSpecial(
ArrayList.class, "toString" , methodType(String.class), Listie.class);
MethodHandle MH_this = Listie.lookup().findSpecial(
Listie.class, "toString" , methodType(String.class), Listie.class);
MethodHandle MH_duper = Listie.lookup().findSpecial(
Object.class, "toString" , methodType(String.class), Listie.class);
assertEquals("[]", (String) MH_super.invokeExact(l));
assertEquals(""+l, (String) MH_this.invokeExact(l));
assertEquals("[]", (String) MH_duper.invokeExact(l)); // ArrayList method
try { assertEquals("inaccessible", Listie.lookup().findSpecial(
String.class, "toString", methodType(String.class), Listie.class));
} catch (IllegalAccessException ex) { } // OK
Listie subl = new Listie() { public String toString() { return "[subclass]"; } };
assertEquals(""+l, (String) MH_this.invokeExact(subl)); // Listie method
* }
*
* @param refc the class or interface from which the method is accessed
* @param name the name of the method (which must not be "&lt;init&gt;")
* @param type the type of the method, with the receiver argument omitted
* @param specialCaller the proposed calling class to perform the {@code invokespecial}
* @return the desired method handle
* @throws NoSuchMethodException if the method does not exist
* @throws IllegalAccessException if access checking fails,
* or if the method is {@code static},
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
*/
public MethodHandle findSpecial(Class<?> refc, String name, MethodType type,
Class<?> specialCaller) throws NoSuchMethodException, IllegalAccessException {
checkSpecialCaller(specialCaller, refc);
Lookup specialLookup = this.in(specialCaller);
MemberName method = specialLookup.resolveOrFail(REF_invokeSpecial, refc, name, type);
return specialLookup.getDirectMethod(REF_invokeSpecial, refc, method, findBoundCallerLookup(method));
}
/**
* Produces a method handle giving read access to a non-static field.
* The type of the method handle will have a return type of the field's
* value type.
* The method handle's single argument will be the instance containing
* the field.
* Access checking is performed immediately on behalf of the lookup class.
* @param refc the class or interface from which the method is accessed
* @param name the field's name
* @param type the field's type
* @return a method handle which can load values from the field
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is {@code static}
* @throws NullPointerException if any argument is null
* @see #findVarHandle(Class, String, Class)
*/
public MethodHandle findGetter(Class<?> refc, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName field = resolveOrFail(REF_getField, refc, name, type);
return getDirectField(REF_getField, refc, field);
}
/**
* Produces a method handle giving write access to a non-static field.
* The type of the method handle will have a void return type.
* The method handle will take two arguments, the instance containing
* the field, and the value to be stored.
* The second argument will be of the field's value type.
* Access checking is performed immediately on behalf of the lookup class.
* @param refc the class or interface from which the method is accessed
* @param name the field's name
* @param type the field's type
* @return a method handle which can store values into the field
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is {@code static}
* or {@code final}
* @throws NullPointerException if any argument is null
* @see #findVarHandle(Class, String, Class)
*/
public MethodHandle findSetter(Class<?> refc, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName field = resolveOrFail(REF_putField, refc, name, type);
return getDirectField(REF_putField, refc, field);
}
/**
* Produces a VarHandle giving access to a non-static field {@code name}
* of type {@code type} declared in a class of type {@code recv}.
* The VarHandle's variable type is {@code type} and it has one
* coordinate type, {@code recv}.
* <p>
* Access checking is performed immediately on behalf of the lookup
* class.
* <p>
* Certain access modes of the returned VarHandle are unsupported under
* the following conditions:
* <ul>
* <li>if the field is declared {@code final}, then the write, atomic
* update, numeric atomic update, and bitwise atomic update access
* modes are unsupported.
* <li>if the field type is anything other than {@code byte},
* {@code short}, {@code char}, {@code int}, {@code long},
* {@code float}, or {@code double} then numeric atomic update
* access modes are unsupported.
* <li>if the field type is anything other than {@code boolean},
* {@code byte}, {@code short}, {@code char}, {@code int} or
* {@code long} then bitwise atomic update access modes are
* unsupported.
* </ul>
* <p>
* If the field is declared {@code volatile} then the returned VarHandle
* will override access to the field (effectively ignore the
* {@code volatile} declaration) in accordance to its specified
* access modes.
* <p>
* If the field type is {@code float} or {@code double} then numeric
* and atomic update access modes compare values using their bitwise
* representation (see {@link Float#floatToRawIntBits} and
* {@link Double#doubleToRawLongBits}, respectively).
* @apiNote
* Bitwise comparison of {@code float} values or {@code double} values,
* as performed by the numeric and atomic update access modes, differ
* from the primitive {@code ==} operator and the {@link Float#equals}
* and {@link Double#equals} methods, specifically with respect to
* comparing NaN values or comparing {@code -0.0} with {@code +0.0}.
* Care should be taken when performing a compare and set or a compare
* and exchange operation with such values since the operation may
* unexpectedly fail.
* There are many possible NaN values that are considered to be
* {@code NaN} in Java, although no IEEE 754 floating-point operation
* provided by Java can distinguish between them. Operation failure can
* occur if the expected or witness value is a NaN value and it is
* transformed (perhaps in a platform specific manner) into another NaN
* value, and thus has a different bitwise representation (see
* {@link Float#intBitsToFloat} or {@link Double#longBitsToDouble} for more
* details).
* The values {@code -0.0} and {@code +0.0} have different bitwise
* representations but are considered equal when using the primitive
* {@code ==} operator. Operation failure can occur if, for example, a
* numeric algorithm computes an expected value to be say {@code -0.0}
* and previously computed the witness value to be say {@code +0.0}.
* @param recv the receiver class, of type {@code R}, that declares the
* non-static field
* @param name the field's name
* @param type the field's type, of type {@code T}
* @return a VarHandle giving access to non-static fields.
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is {@code static}
* @throws NullPointerException if any argument is null
* @since 9
*/
public VarHandle findVarHandle(Class<?> recv, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName getField = resolveOrFail(REF_getField, recv, name, type);
MemberName putField = resolveOrFail(REF_putField, recv, name, type);
return getFieldVarHandle(REF_getField, REF_putField, recv, getField, putField);
}
/**
* Produces a method handle giving read access to a static field.
* The type of the method handle will have a return type of the field's
* value type.
* The method handle will take no arguments.
* Access checking is performed immediately on behalf of the lookup class.
* <p>
* If the returned method handle is invoked, the field's class will
* be initialized, if it has not already been initialized.
* @param refc the class or interface from which the method is accessed
* @param name the field's name
* @param type the field's type
* @return a method handle which can load values from the field
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is not {@code static}
* @throws NullPointerException if any argument is null
*/
public MethodHandle findStaticGetter(Class<?> refc, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName field = resolveOrFail(REF_getStatic, refc, name, type);
return getDirectField(REF_getStatic, refc, field);
}
/**
* Produces a method handle giving write access to a static field.
* The type of the method handle will have a void return type.
* The method handle will take a single
* argument, of the field's value type, the value to be stored.
* Access checking is performed immediately on behalf of the lookup class.
* <p>
* If the returned method handle is invoked, the field's class will
* be initialized, if it has not already been initialized.
* @param refc the class or interface from which the method is accessed
* @param name the field's name
* @param type the field's type
* @return a method handle which can store values into the field
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is not {@code static}
* or is {@code final}
* @throws NullPointerException if any argument is null
*/
public MethodHandle findStaticSetter(Class<?> refc, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName field = resolveOrFail(REF_putStatic, refc, name, type);
return getDirectField(REF_putStatic, refc, field);
}
/**
* Produces a VarHandle giving access to a static field {@code name} of
* type {@code type} declared in a class of type {@code decl}.
* The VarHandle's variable type is {@code type} and it has no
* coordinate types.
* <p>
* Access checking is performed immediately on behalf of the lookup
* class.
* <p>
* If the returned VarHandle is operated on, the declaring class will be
* initialized, if it has not already been initialized.
* <p>
* Certain access modes of the returned VarHandle are unsupported under
* the following conditions:
* <ul>
* <li>if the field is declared {@code final}, then the write, atomic
* update, numeric atomic update, and bitwise atomic update access
* modes are unsupported.
* <li>if the field type is anything other than {@code byte},
* {@code short}, {@code char}, {@code int}, {@code long},
* {@code float}, or {@code double}, then numeric atomic update
* access modes are unsupported.
* <li>if the field type is anything other than {@code boolean},
* {@code byte}, {@code short}, {@code char}, {@code int} or
* {@code long} then bitwise atomic update access modes are
* unsupported.
* </ul>
* <p>
* If the field is declared {@code volatile} then the returned VarHandle
* will override access to the field (effectively ignore the
* {@code volatile} declaration) in accordance to its specified
* access modes.
* <p>
* If the field type is {@code float} or {@code double} then numeric
* and atomic update access modes compare values using their bitwise
* representation (see {@link Float#floatToRawIntBits} and
* {@link Double#doubleToRawLongBits}, respectively).
* @apiNote
* Bitwise comparison of {@code float} values or {@code double} values,
* as performed by the numeric and atomic update access modes, differ
* from the primitive {@code ==} operator and the {@link Float#equals}
* and {@link Double#equals} methods, specifically with respect to
* comparing NaN values or comparing {@code -0.0} with {@code +0.0}.
* Care should be taken when performing a compare and set or a compare
* and exchange operation with such values since the operation may
* unexpectedly fail.
* There are many possible NaN values that are considered to be
* {@code NaN} in Java, although no IEEE 754 floating-point operation
* provided by Java can distinguish between them. Operation failure can
* occur if the expected or witness value is a NaN value and it is
* transformed (perhaps in a platform specific manner) into another NaN
* value, and thus has a different bitwise representation (see
* {@link Float#intBitsToFloat} or {@link Double#longBitsToDouble} for more
* details).
* The values {@code -0.0} and {@code +0.0} have different bitwise
* representations but are considered equal when using the primitive
* {@code ==} operator. Operation failure can occur if, for example, a
* numeric algorithm computes an expected value to be say {@code -0.0}
* and previously computed the witness value to be say {@code +0.0}.
* @param decl the class that declares the static field
* @param name the field's name
* @param type the field's type, of type {@code T}
* @return a VarHandle giving access to a static field
* @throws NoSuchFieldException if the field does not exist
* @throws IllegalAccessException if access checking fails, or if the field is not {@code static}
* @throws NullPointerException if any argument is null
* @since 9
*/
public VarHandle findStaticVarHandle(Class<?> decl, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
MemberName getField = resolveOrFail(REF_getStatic, decl, name, type);
MemberName putField = resolveOrFail(REF_putStatic, decl, name, type);
return getFieldVarHandle(REF_getStatic, REF_putStatic, decl, getField, putField);
}
/**
* Produces an early-bound method handle for a non-static method.
* The receiver must have a supertype {@code defc} in which a method
* of the given name and type is accessible to the lookup class.
* The method and all its argument types must be accessible to the lookup object.
* The type of the method handle will be that of the method,
* without any insertion of an additional receiver parameter.
* The given receiver will be bound into the method handle,
* so that every call to the method handle will invoke the
* requested method on the given receiver.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set
* <em>and</em> the trailing array argument is not the only argument.
* (If the trailing array argument is the only argument,
* the given receiver value will be bound to it.)
* <p>
* This is almost equivalent to the following code, with some differences noted below:
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle mh0 = lookup().findVirtual(defc, name, type);
MethodHandle mh1 = mh0.bindTo(receiver);
mh1 = mh1.withVarargs(mh0.isVarargsCollector());
return mh1;
* }
* where {@code defc} is either {@code receiver.getClass()} or a super
* type of that class, in which the requested method is accessible
* to the lookup class.
* (Unlike {@code bind}, {@code bindTo} does not preserve variable arity.
* Also, {@code bindTo} may throw a {@code ClassCastException} in instances where {@code bind} would
* throw an {@code IllegalAccessException}, as in the case where the member is {@code protected} and
* the receiver is restricted by {@code findVirtual} to the lookup class.)
* @param receiver the object from which the method is accessed
* @param name the name of the method
* @param type the type of the method, with the receiver argument omitted
* @return the desired method handle
* @throws NoSuchMethodException if the method does not exist
* @throws IllegalAccessException if access checking fails
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
* @see MethodHandle#bindTo
* @see #findVirtual
*/
public MethodHandle bind(Object receiver, String name, MethodType type) throws NoSuchMethodException, IllegalAccessException {
Class<? extends Object> refc = receiver.getClass(); // may get NPE
MemberName method = resolveOrFail(REF_invokeSpecial, refc, name, type);
MethodHandle mh = getDirectMethodNoRestrictInvokeSpecial(refc, method, findBoundCallerLookup(method));
if (!mh.type().leadingReferenceParameter().isAssignableFrom(receiver.getClass())) {
throw new IllegalAccessException("The restricted defining class " +
mh.type().leadingReferenceParameter().getName() +
" is not assignable from receiver class " +
receiver.getClass().getName());
}
return mh.bindArgumentL(0, receiver).setVarargs(method);
}
/**
* Makes a <a href="MethodHandleInfo.html#directmh">direct method handle</a>
* to <i>m</i>, if the lookup class has permission.
* If <i>m</i> is non-static, the receiver argument is treated as an initial argument.
* If <i>m</i> is virtual, overriding is respected on every call.
* Unlike the Core Reflection API, exceptions are <em>not</em> wrapped.
* The type of the method handle will be that of the method,
* with the receiver type prepended (but only if it is non-static).
* If the method's {@code accessible} flag is not set,
* access checking is performed immediately on behalf of the lookup class.
* If <i>m</i> is not public, do not share the resulting handle with untrusted parties.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set.
* <p>
* If <i>m</i> is static, and
* if the returned method handle is invoked, the method's class will
* be initialized, if it has not already been initialized.
* @param m the reflected method
* @return a method handle which can invoke the reflected method
* @throws IllegalAccessException if access checking fails
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if the argument is null
*/
public MethodHandle unreflect(Method m) throws IllegalAccessException {
if (m.getDeclaringClass() == MethodHandle.class) {
MethodHandle mh = unreflectForMH(m);
if (mh != null) return mh;
}
if (m.getDeclaringClass() == VarHandle.class) {
MethodHandle mh = unreflectForVH(m);
if (mh != null) return mh;
}
MemberName method = new MemberName(m);
byte refKind = method.getReferenceKind();
if (refKind == REF_invokeSpecial)
refKind = REF_invokeVirtual;
assert(method.isMethod());
@SuppressWarnings("deprecation")
Lookup lookup = m.isAccessible() ? IMPL_LOOKUP : this;
return lookup.getDirectMethod(refKind, method.getDeclaringClass(), method, findBoundCallerLookup(method));
}
private MethodHandle unreflectForMH(Method m) {
// these names require special lookups because they throw UnsupportedOperationException
if (MemberName.isMethodHandleInvokeName(m.getName()))
return MethodHandleImpl.fakeMethodHandleInvoke(new MemberName(m));
return null;
}
private MethodHandle unreflectForVH(Method m) {
// these names require special lookups because they throw UnsupportedOperationException
if (MemberName.isVarHandleMethodInvokeName(m.getName()))
return MethodHandleImpl.fakeVarHandleInvoke(new MemberName(m));
return null;
}
/**
* Produces a method handle for a reflected method.
* It will bypass checks for overriding methods on the receiver,
* <a href="MethodHandles.Lookup.html#equiv">as if called</a> from an {@code invokespecial}
* instruction from within the explicitly specified {@code specialCaller}.
* The type of the method handle will be that of the method,
* with a suitably restricted receiver type prepended.
* (The receiver type will be {@code specialCaller} or a subtype.)
* If the method's {@code accessible} flag is not set,
* access checking is performed immediately on behalf of the lookup class,
* as if {@code invokespecial} instruction were being linked.
* <p>
* Before method resolution,
* if the explicitly specified caller class is not identical with the
* lookup class, or if this lookup object does not have
* <a href="MethodHandles.Lookup.html#privacc">private access</a>
* privileges, the access fails.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the method's variable arity modifier bit ({@code 0x0080}) is set.
* @param m the reflected method
* @param specialCaller the class nominally calling the method
* @return a method handle which can invoke the reflected method
* @throws IllegalAccessException if access checking fails,
* or if the method is {@code static},
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if any argument is null
*/
public MethodHandle unreflectSpecial(Method m, Class<?> specialCaller) throws IllegalAccessException {
checkSpecialCaller(specialCaller, m.getDeclaringClass());
Lookup specialLookup = this.in(specialCaller);
MemberName method = new MemberName(m, true);
assert(method.isMethod());
// ignore m.isAccessible: this is a new kind of access
return specialLookup.getDirectMethod(REF_invokeSpecial, method.getDeclaringClass(), method, findBoundCallerLookup(method));
}
/**
* Produces a method handle for a reflected constructor.
* The type of the method handle will be that of the constructor,
* with the return type changed to the declaring class.
* The method handle will perform a {@code newInstance} operation,
* creating a new instance of the constructor's class on the
* arguments passed to the method handle.
* <p>
* If the constructor's {@code accessible} flag is not set,
* access checking is performed immediately on behalf of the lookup class.
* <p>
* The returned method handle will have
* {@linkplain MethodHandle#asVarargsCollector variable arity} if and only if
* the constructor's variable arity modifier bit ({@code 0x0080}) is set.
* <p>
* If the returned method handle is invoked, the constructor's class will
* be initialized, if it has not already been initialized.
* @param c the reflected constructor
* @return a method handle which can invoke the reflected constructor
* @throws IllegalAccessException if access checking fails
* or if the method's variable arity modifier bit
* is set and {@code asVarargsCollector} fails
* @throws NullPointerException if the argument is null
*/
public MethodHandle unreflectConstructor(Constructor<?> c) throws IllegalAccessException {
MemberName ctor = new MemberName(c);
assert(ctor.isConstructor());
@SuppressWarnings("deprecation")
Lookup lookup = c.isAccessible() ? IMPL_LOOKUP : this;
return lookup.getDirectConstructor(ctor.getDeclaringClass(), ctor);
}
/*
* Produces a method handle that is capable of creating instances of the given class
* and instantiated by the given constructor.
*
* This method should only be used by ReflectionFactory::newConstructorForSerialization.
*/
/* package-private */ MethodHandle serializableConstructor(Class<?> decl, Constructor<?> c) throws IllegalAccessException {
MemberName ctor = new MemberName(c);
assert(ctor.isConstructor() && constructorInSuperclass(decl, c));
checkAccess(REF_newInvokeSpecial, decl, ctor);
assert(!MethodHandleNatives.isCallerSensitive(ctor)); // maybeBindCaller not relevant here
return DirectMethodHandle.makeAllocator(decl, ctor).setVarargs(ctor);
}
private static boolean constructorInSuperclass(Class<?> decl, Constructor<?> ctor) {
if (decl == ctor.getDeclaringClass())
return true;
Class<?> cl = decl;
while ((cl = cl.getSuperclass()) != null) {
if (cl == ctor.getDeclaringClass()) {
return true;
}
}
return false;
}
/**
* Produces a method handle giving read access to a reflected field.
* The type of the method handle will have a return type of the field's
* value type.
* If the field is {@code static}, the method handle will take no arguments.
* Otherwise, its single argument will be the instance containing
* the field.
* If the {@code Field} object's {@code accessible} flag is not set,
* access checking is performed immediately on behalf of the lookup class.
* <p>
* If the field is static, and
* if the returned method handle is invoked, the field's class will
* be initialized, if it has not already been initialized.
* @param f the reflected field
* @return a method handle which can load values from the reflected field
* @throws IllegalAccessException if access checking fails
* @throws NullPointerException if the argument is null
*/
public MethodHandle unreflectGetter(Field f) throws IllegalAccessException {
return unreflectField(f, false);
}
/**
* Produces a method handle giving write access to a reflected field.
* The type of the method handle will have a void return type.
* If the field is {@code static}, the method handle will take a single
* argument, of the field's value type, the value to be stored.
* Otherwise, the two arguments will be the instance containing
* the field, and the value to be stored.
* If the {@code Field} object's {@code accessible} flag is not set,
* access checking is performed immediately on behalf of the lookup class.
* <p>
* If the field is {@code final}, write access will not be
* allowed and access checking will fail, except under certain
* narrow circumstances documented for {@link Field#set Field.set}.
* A method handle is returned only if a corresponding call to
* the {@code Field} object's {@code set} method could return
* normally. In particular, fields which are both {@code static}
* and {@code final} may never be set.
* <p>
* If the field is {@code static}, and
* if the returned method handle is invoked, the field's class will
* be initialized, if it has not already been initialized.
* @param f the reflected field
* @return a method handle which can store values into the reflected field
* @throws IllegalAccessException if access checking fails,
* or if the field is {@code final} and write access
* is not enabled on the {@code Field} object
* @throws NullPointerException if the argument is null
*/
public MethodHandle unreflectSetter(Field f) throws IllegalAccessException {
return unreflectField(f, true);
}
private MethodHandle unreflectField(Field f, boolean isSetter) throws IllegalAccessException {
MemberName field = new MemberName(f, isSetter);
if (isSetter && field.isFinal()) {
if (field.isTrustedFinalField()) {
String msg = field.isStatic() ? "static final field has no write access"
: "final field has no write access";
throw field.makeAccessException(msg, this);
}
}
assert(isSetter
? MethodHandleNatives.refKindIsSetter(field.getReferenceKind())
: MethodHandleNatives.refKindIsGetter(field.getReferenceKind()));
@SuppressWarnings("deprecation")
Lookup lookup = f.isAccessible() ? IMPL_LOOKUP : this;
return lookup.getDirectField(field.getReferenceKind(), f.getDeclaringClass(), field);
}
/**
* Produces a VarHandle giving access to a reflected field {@code f}
* of type {@code T} declared in a class of type {@code R}.
* The VarHandle's variable type is {@code T}.
* If the field is non-static the VarHandle has one coordinate type,
* {@code R}. Otherwise, the field is static, and the VarHandle has no
* coordinate types.
* <p>
* Access checking is performed immediately on behalf of the lookup
* class, regardless of the value of the field's {@code accessible}
* flag.
* <p>
* If the field is static, and if the returned VarHandle is operated
* on, the field's declaring class will be initialized, if it has not
* already been initialized.
* <p>
* Certain access modes of the returned VarHandle are unsupported under
* the following conditions:
* <ul>
* <li>if the field is declared {@code final}, then the write, atomic
* update, numeric atomic update, and bitwise atomic update access
* modes are unsupported.
* <li>if the field type is anything other than {@code byte},
* {@code short}, {@code char}, {@code int}, {@code long},
* {@code float}, or {@code double} then numeric atomic update
* access modes are unsupported.
* <li>if the field type is anything other than {@code boolean},
* {@code byte}, {@code short}, {@code char}, {@code int} or
* {@code long} then bitwise atomic update access modes are
* unsupported.
* </ul>
* <p>
* If the field is declared {@code volatile} then the returned VarHandle
* will override access to the field (effectively ignore the
* {@code volatile} declaration) in accordance to its specified
* access modes.
* <p>
* If the field type is {@code float} or {@code double} then numeric
* and atomic update access modes compare values using their bitwise
* representation (see {@link Float#floatToRawIntBits} and
* {@link Double#doubleToRawLongBits}, respectively).
* @apiNote
* Bitwise comparison of {@code float} values or {@code double} values,
* as performed by the numeric and atomic update access modes, differ
* from the primitive {@code ==} operator and the {@link Float#equals}
* and {@link Double#equals} methods, specifically with respect to
* comparing NaN values or comparing {@code -0.0} with {@code +0.0}.
* Care should be taken when performing a compare and set or a compare
* and exchange operation with such values since the operation may
* unexpectedly fail.
* There are many possible NaN values that are considered to be
* {@code NaN} in Java, although no IEEE 754 floating-point operation
* provided by Java can distinguish between them. Operation failure can
* occur if the expected or witness value is a NaN value and it is
* transformed (perhaps in a platform specific manner) into another NaN
* value, and thus has a different bitwise representation (see
* {@link Float#intBitsToFloat} or {@link Double#longBitsToDouble} for more
* details).
* The values {@code -0.0} and {@code +0.0} have different bitwise
* representations but are considered equal when using the primitive
* {@code ==} operator. Operation failure can occur if, for example, a
* numeric algorithm computes an expected value to be say {@code -0.0}
* and previously computed the witness value to be say {@code +0.0}.
* @param f the reflected field, with a field of type {@code T}, and
* a declaring class of type {@code R}
* @return a VarHandle giving access to non-static fields or a static
* field
* @throws IllegalAccessException if access checking fails
* @throws NullPointerException if the argument is null
* @since 9
*/
public VarHandle unreflectVarHandle(Field f) throws IllegalAccessException {
MemberName getField = new MemberName(f, false);
MemberName putField = new MemberName(f, true);
return getFieldVarHandle(getField.getReferenceKind(), putField.getReferenceKind(),
f.getDeclaringClass(), getField, putField);
}
/**
* Cracks a <a href="MethodHandleInfo.html#directmh">direct method handle</a>
* created by this lookup object or a similar one.
* Security and access checks are performed to ensure that this lookup object
* is capable of reproducing the target method handle.
* This means that the cracking may fail if target is a direct method handle
* but was created by an unrelated lookup object.
* This can happen if the method handle is <a href="MethodHandles.Lookup.html#callsens">caller sensitive</a>
* and was created by a lookup object for a different class.
* @param target a direct method handle to crack into symbolic reference components
* @return a symbolic reference which can be used to reconstruct this method handle from this lookup object
* @throws IllegalArgumentException if the target is not a direct method handle or if access checking fails
* @throws NullPointerException if the target is {@code null}
* @see MethodHandleInfo
* @since 1.8
*/
public MethodHandleInfo revealDirect(MethodHandle target) {
if (!target.isCrackable()) {
throw newIllegalArgumentException("not a direct method handle");
}
MemberName member = target.internalMemberName();
Class<?> defc = member.getDeclaringClass();
byte refKind = member.getReferenceKind();
assert(MethodHandleNatives.refKindIsValid(refKind));
if (refKind == REF_invokeSpecial && !target.isInvokeSpecial())
// Devirtualized method invocation is usually formally virtual.
// To avoid creating extra MemberName objects for this common case,
// we encode this extra degree of freedom using MH.isInvokeSpecial.
refKind = REF_invokeVirtual;
if (refKind == REF_invokeVirtual && defc.isInterface())
// Symbolic reference is through interface but resolves to Object method (toString, etc.)
refKind = REF_invokeInterface;
// Check member access before cracking.
try {
checkAccess(refKind, defc, member);
} catch (IllegalAccessException ex) {
throw new IllegalArgumentException(ex);
}
if (allowedModes != TRUSTED && member.isCallerSensitive()) {
Class<?> callerClass = target.internalCallerClass();
if ((lookupModes() & ORIGINAL) == 0 || callerClass != lookupClass())
throw new IllegalArgumentException("method handle is caller sensitive: "+callerClass);
}
// Produce the handle to the results.
return new InfoFromMemberName(this, member, refKind);
}
//--- Helper methods, all package-private.
MemberName resolveOrFail(byte refKind, Class<?> refc, String name, Class<?> type) throws NoSuchFieldException, IllegalAccessException {
checkSymbolicClass(refc); // do this before attempting to resolve
Objects.requireNonNull(name);
Objects.requireNonNull(type);
return IMPL_NAMES.resolveOrFail(refKind, new MemberName(refc, name, type, refKind), lookupClassOrNull(), allowedModes,
NoSuchFieldException.class);
}
MemberName resolveOrFail(byte refKind, Class<?> refc, String name, MethodType type) throws NoSuchMethodException, IllegalAccessException {
checkSymbolicClass(refc); // do this before attempting to resolve
Objects.requireNonNull(type);
checkMethodName(refKind, name); // implicit null-check of name
return IMPL_NAMES.resolveOrFail(refKind, new MemberName(refc, name, type, refKind), lookupClassOrNull(), allowedModes,
NoSuchMethodException.class);
}
MemberName resolveOrFail(byte refKind, MemberName member) throws ReflectiveOperationException {
checkSymbolicClass(member.getDeclaringClass()); // do this before attempting to resolve
Objects.requireNonNull(member.getName());
Objects.requireNonNull(member.getType());
return IMPL_NAMES.resolveOrFail(refKind, member, lookupClassOrNull(), allowedModes,
ReflectiveOperationException.class);
}
MemberName resolveOrNull(byte refKind, MemberName member) {
// do this before attempting to resolve
if (!isClassAccessible(member.getDeclaringClass())) {
return null;
}
Objects.requireNonNull(member.getName());
Objects.requireNonNull(member.getType());
return IMPL_NAMES.resolveOrNull(refKind, member, lookupClassOrNull(), allowedModes);
}
MemberName resolveOrNull(byte refKind, Class<?> refc, String name, MethodType type) {
// do this before attempting to resolve
if (!isClassAccessible(refc)) {
return null;
}
Objects.requireNonNull(type);
// implicit null-check of name
if (name.startsWith("<") && refKind != REF_newInvokeSpecial) {
return null;
}
return IMPL_NAMES.resolveOrNull(refKind, new MemberName(refc, name, type, refKind), lookupClassOrNull(), allowedModes);
}
void checkSymbolicClass(Class<?> refc) throws IllegalAccessException {
if (!isClassAccessible(refc)) {
throw new MemberName(refc).makeAccessException("symbolic reference class is not accessible", this);
}
}
boolean isClassAccessible(Class<?> refc) {
Objects.requireNonNull(refc);
Class<?> caller = lookupClassOrNull();
Class<?> type = refc;
while (type.isArray()) {
type = type.getComponentType();
}
return caller == null || VerifyAccess.isClassAccessible(type, caller, prevLookupClass, allowedModes);
}
/** Check name for an illegal leading "&lt;" character. */
void checkMethodName(byte refKind, String name) throws NoSuchMethodException {
if (name.startsWith("<") && refKind != REF_newInvokeSpecial)
throw new NoSuchMethodException("illegal method name: "+name);
}
/**
* Find my trustable caller class if m is a caller sensitive method.
* If this lookup object has original full privilege access, then the caller class is the lookupClass.
* Otherwise, if m is caller-sensitive, throw IllegalAccessException.
*/
Lookup findBoundCallerLookup(MemberName m) throws IllegalAccessException {
if (MethodHandleNatives.isCallerSensitive(m) && (lookupModes() & ORIGINAL) == 0) {
// Only lookups with full privilege access are allowed to resolve caller-sensitive methods
throw new IllegalAccessException("Attempt to lookup caller-sensitive method using restricted lookup object");
}
return this;
}
/**
* Returns {@code true} if this lookup has {@code PRIVATE} and {@code MODULE} access.
* @return {@code true} if this lookup has {@code PRIVATE} and {@code MODULE} access.
*
* @deprecated This method was originally designed to test {@code PRIVATE} access
* that implies full privilege access but {@code MODULE} access has since become
* independent of {@code PRIVATE} access. It is recommended to call
* {@link #hasFullPrivilegeAccess()} instead.
* @since 9
*/
@Deprecated(since="14")
public boolean hasPrivateAccess() {
return hasFullPrivilegeAccess();
}
/**
* Returns {@code true} if this lookup has <em>full privilege access</em>,
* i.e. {@code PRIVATE} and {@code MODULE} access.
* A {@code Lookup} object must have full privilege access in order to
* access all members that are allowed to the
* {@linkplain #lookupClass() lookup class}.
*
* @return {@code true} if this lookup has full privilege access.
* @since 14
* @see <a href="MethodHandles.Lookup.html#privacc">private and module access</a>
*/
public boolean hasFullPrivilegeAccess() {
return (allowedModes & (PRIVATE|MODULE)) == (PRIVATE|MODULE);
}
void checkMethod(byte refKind, Class<?> refc, MemberName m) throws IllegalAccessException {
boolean wantStatic = (refKind == REF_invokeStatic);
String message;
if (m.isConstructor())
message = "expected a method, not a constructor";
else if (!m.isMethod())
message = "expected a method";
else if (wantStatic != m.isStatic())
message = wantStatic ? "expected a static method" : "expected a non-static method";
else
{ checkAccess(refKind, refc, m); return; }
throw m.makeAccessException(message, this);
}
void checkField(byte refKind, Class<?> refc, MemberName m) throws IllegalAccessException {
boolean wantStatic = !MethodHandleNatives.refKindHasReceiver(refKind);
String message;
if (wantStatic != m.isStatic())
message = wantStatic ? "expected a static field" : "expected a non-static field";
else
{ checkAccess(refKind, refc, m); return; }
throw m.makeAccessException(message, this);
}
private boolean isArrayClone(byte refKind, Class<?> refc, MemberName m) {
return Modifier.isProtected(m.getModifiers()) &&
refKind == REF_invokeVirtual &&
m.getDeclaringClass() == Object.class &&
m.getName().equals("clone") &&
refc.isArray();
}
/** Check public/protected/private bits on the symbolic reference class and its member. */
void checkAccess(byte refKind, Class<?> refc, MemberName m) throws IllegalAccessException {
assert(m.referenceKindIsConsistentWith(refKind) &&
MethodHandleNatives.refKindIsValid(refKind) &&
(MethodHandleNatives.refKindIsField(refKind) == m.isField()));
int allowedModes = this.allowedModes;
if (allowedModes == TRUSTED) return;
int mods = m.getModifiers();
if (isArrayClone(refKind, refc, m)) {
// The JVM does this hack also.
// (See ClassVerifier::verify_invoke_instructions
// and LinkResolver::check_method_accessability.)
// Because the JVM does not allow separate methods on array types,
// there is no separate method for int[].clone.
// All arrays simply inherit Object.clone.
// But for access checking logic, we make Object.clone
// (normally protected) appear to be public.
// Later on, when the DirectMethodHandle is created,
// its leading argument will be restricted to the
// requested array type.
// N.B. The return type is not adjusted, because
// that is *not* the bytecode behavior.
mods ^= Modifier.PROTECTED | Modifier.PUBLIC;
}
if (Modifier.isProtected(mods) && refKind == REF_newInvokeSpecial) {
// cannot "new" a protected ctor in a different package
mods ^= Modifier.PROTECTED;
}
if (Modifier.isFinal(mods) &&
MethodHandleNatives.refKindIsSetter(refKind))
throw m.makeAccessException("unexpected set of a final field", this);
int requestedModes = fixmods(mods); // adjust 0 => PACKAGE
if ((requestedModes & allowedModes) != 0) {
if (VerifyAccess.isMemberAccessible(refc, m.getDeclaringClass(),
mods, lookupClass(), previousLookupClass(), allowedModes))
return;
} else {
// Protected members can also be checked as if they were package-private.
if ((requestedModes & PROTECTED) != 0 && (allowedModes & PACKAGE) != 0
&& VerifyAccess.isSamePackage(m.getDeclaringClass(), lookupClass()))
return;
}
throw m.makeAccessException(accessFailedMessage(refc, m), this);
}
String accessFailedMessage(Class<?> refc, MemberName m) {
Class<?> defc = m.getDeclaringClass();
int mods = m.getModifiers();
// check the class first:
boolean classOK = (Modifier.isPublic(defc.getModifiers()) &&
(defc == refc ||
Modifier.isPublic(refc.getModifiers())));
if (!classOK && (allowedModes & PACKAGE) != 0) {
// ignore previous lookup class to check if default package access
classOK = (VerifyAccess.isClassAccessible(defc, lookupClass(), null, FULL_POWER_MODES) &&
(defc == refc ||
VerifyAccess.isClassAccessible(refc, lookupClass(), null, FULL_POWER_MODES)));
}
if (!classOK)
return "class is not public";
if (Modifier.isPublic(mods))
return "access to public member failed"; // (how?, module not readable?)
if (Modifier.isPrivate(mods))
return "member is private";
if (Modifier.isProtected(mods))
return "member is protected";
return "member is private to package";
}
private void checkSpecialCaller(Class<?> specialCaller, Class<?> refc) throws IllegalAccessException {
int allowedModes = this.allowedModes;
if (allowedModes == TRUSTED) return;
if ((lookupModes() & PRIVATE) == 0
|| (specialCaller != lookupClass()
// ensure non-abstract methods in superinterfaces can be special-invoked
&& !(refc != null && refc.isInterface() && refc.isAssignableFrom(specialCaller))))
throw new MemberName(specialCaller).
makeAccessException("no private access for invokespecial", this);
}
private boolean restrictProtectedReceiver(MemberName method) {
// The accessing class only has the right to use a protected member
// on itself or a subclass. Enforce that restriction, from JVMS 5.4.4, etc.
if (!method.isProtected() || method.isStatic()
|| allowedModes == TRUSTED
|| method.getDeclaringClass() == lookupClass()
|| VerifyAccess.isSamePackage(method.getDeclaringClass(), lookupClass()))
return false;
return true;
}
private MethodHandle restrictReceiver(MemberName method, DirectMethodHandle mh, Class<?> caller) throws IllegalAccessException {
assert(!method.isStatic());
// receiver type of mh is too wide; narrow to caller
if (!method.getDeclaringClass().isAssignableFrom(caller)) {
throw method.makeAccessException("caller class must be a subclass below the method", caller);
}
MethodType rawType = mh.type();
if (caller.isAssignableFrom(rawType.parameterType(0))) return mh; // no need to restrict; already narrow
MethodType narrowType = rawType.changeParameterType(0, caller);
assert(!mh.isVarargsCollector()); // viewAsType will lose varargs-ness
assert(mh.viewAsTypeChecks(narrowType, true));
return mh.copyWith(narrowType, mh.form);
}
/** Check access and get the requested method. */
private MethodHandle getDirectMethod(byte refKind, Class<?> refc, MemberName method, Lookup callerLookup) throws IllegalAccessException {
final boolean doRestrict = true;
return getDirectMethodCommon(refKind, refc, method, doRestrict, callerLookup);
}
/** Check access and get the requested method, for invokespecial with no restriction on the application of narrowing rules. */
private MethodHandle getDirectMethodNoRestrictInvokeSpecial(Class<?> refc, MemberName method, Lookup callerLookup) throws IllegalAccessException {
final boolean doRestrict = false;
return getDirectMethodCommon(REF_invokeSpecial, refc, method, doRestrict, callerLookup);
}
/** Common code for all methods; do not call directly except from immediately above. */
private MethodHandle getDirectMethodCommon(byte refKind, Class<?> refc, MemberName method,
boolean doRestrict,
Lookup boundCaller) throws IllegalAccessException {
checkMethod(refKind, refc, method);
assert(!method.isMethodHandleInvoke());
if (refKind == REF_invokeSpecial &&
refc != lookupClass() &&
!refc.isInterface() && !lookupClass().isInterface() &&
refc != lookupClass().getSuperclass() &&
refc.isAssignableFrom(lookupClass())) {
assert(!method.getName().equals(ConstantDescs.INIT_NAME)); // not this code path
// Per JVMS 6.5, desc. of invokespecial instruction:
// If the method is in a superclass of the LC,
// and if our original search was above LC.super,
// repeat the search (symbolic lookup) from LC.super
// and continue with the direct superclass of that class,
// and so forth, until a match is found or no further superclasses exist.
// FIXME: MemberName.resolve should handle this instead.
Class<?> refcAsSuper = lookupClass();
MemberName m2;
do {
refcAsSuper = refcAsSuper.getSuperclass();
m2 = new MemberName(refcAsSuper,
method.getName(),
method.getMethodType(),
REF_invokeSpecial);
m2 = IMPL_NAMES.resolveOrNull(refKind, m2, lookupClassOrNull(), allowedModes);
} while (m2 == null && // no method is found yet
refc != refcAsSuper); // search up to refc
if (m2 == null) throw new InternalError(method.toString());
method = m2;
refc = refcAsSuper;
// redo basic checks
checkMethod(refKind, refc, method);
}
DirectMethodHandle dmh = DirectMethodHandle.make(refKind, refc, method, lookupClass());
MethodHandle mh = dmh;
// Optionally narrow the receiver argument to lookupClass using restrictReceiver.
if ((doRestrict && refKind == REF_invokeSpecial) ||
(MethodHandleNatives.refKindHasReceiver(refKind) &&
restrictProtectedReceiver(method) &&
// All arrays simply inherit the protected Object.clone method.
// The leading argument is already restricted to the requested
// array type (not the lookup class).
!isArrayClone(refKind, refc, method))) {
mh = restrictReceiver(method, dmh, lookupClass());
}
mh = maybeBindCaller(method, mh, boundCaller);
mh = mh.setVarargs(method);
return mh;
}
private MethodHandle maybeBindCaller(MemberName method, MethodHandle mh, Lookup boundCaller)
throws IllegalAccessException {
if (boundCaller.allowedModes == TRUSTED || !MethodHandleNatives.isCallerSensitive(method))
return mh;
// boundCaller must have full privilege access.
// It should have been checked by findBoundCallerLookup. Safe to check this again.
if ((boundCaller.lookupModes() & ORIGINAL) == 0)
throw new IllegalAccessException("Attempt to lookup caller-sensitive method using restricted lookup object");
assert boundCaller.hasFullPrivilegeAccess();
MethodHandle cbmh = MethodHandleImpl.bindCaller(mh, boundCaller.lookupClass);
// Note: caller will apply varargs after this step happens.
return cbmh;
}
/** Check access and get the requested field. */
private MethodHandle getDirectField(byte refKind, Class<?> refc, MemberName field) throws IllegalAccessException {
return getDirectFieldCommon(refKind, refc, field);
}
/** Common code for all fields; do not call directly except from immediately above. */
private MethodHandle getDirectFieldCommon(byte refKind, Class<?> refc, MemberName field) throws IllegalAccessException {
checkField(refKind, refc, field);
DirectMethodHandle dmh = DirectMethodHandle.make(refc, field);
boolean doRestrict = (MethodHandleNatives.refKindHasReceiver(refKind) &&
restrictProtectedReceiver(field));
if (doRestrict)
return restrictReceiver(field, dmh, lookupClass());
return dmh;
}
private VarHandle getFieldVarHandle(byte getRefKind, byte putRefKind,
Class<?> refc, MemberName getField, MemberName putField)
throws IllegalAccessException {
return getFieldVarHandleCommon(getRefKind, putRefKind, refc, getField, putField);
}
private VarHandle getFieldVarHandleCommon(byte getRefKind, byte putRefKind,
Class<?> refc, MemberName getField,
MemberName putField) throws IllegalAccessException {
assert getField.isStatic() == putField.isStatic();
assert getField.isGetter() && putField.isSetter();
assert MethodHandleNatives.refKindIsStatic(getRefKind) == MethodHandleNatives.refKindIsStatic(putRefKind);
assert MethodHandleNatives.refKindIsGetter(getRefKind) && MethodHandleNatives.refKindIsSetter(putRefKind);
checkField(getRefKind, refc, getField);
if (!putField.isFinal()) {
// A VarHandle does not support updates to final fields, any
// such VarHandle to a final field will be read-only and
// therefore the following write-based accessibility checks are
// only required for non-final fields
checkField(putRefKind, refc, putField);
}
boolean doRestrict = (MethodHandleNatives.refKindHasReceiver(getRefKind) &&
restrictProtectedReceiver(getField));
if (doRestrict) {
assert !getField.isStatic();
// receiver type of VarHandle is too wide; narrow to caller
if (!getField.getDeclaringClass().isAssignableFrom(lookupClass())) {
throw getField.makeAccessException("caller class must be a subclass below the method", lookupClass());
}
refc = lookupClass();
}
return VarHandles.makeFieldHandle(getField, refc,
this.allowedModes == TRUSTED && !getField.isTrustedFinalField());
}
/** Check access and get the requested constructor. */
private MethodHandle getDirectConstructor(Class<?> refc, MemberName ctor) throws IllegalAccessException {
return getDirectConstructorCommon(refc, ctor);
}
/** Common code for all constructors; do not call directly except from immediately above. */
private MethodHandle getDirectConstructorCommon(Class<?> refc, MemberName ctor) throws IllegalAccessException {
assert(ctor.isConstructor());
checkAccess(REF_newInvokeSpecial, refc, ctor);
assert(!MethodHandleNatives.isCallerSensitive(ctor)); // maybeBindCaller not relevant here
return DirectMethodHandle.make(ctor).setVarargs(ctor);
}
/** Hook called from the JVM (via MethodHandleNatives) to link MH constants:
*/
/*non-public*/
MethodHandle linkMethodHandleConstant(byte refKind, Class<?> defc, String name, Object type)
throws ReflectiveOperationException {
if (!(type instanceof Class || type instanceof MethodType))
throw new InternalError("unresolved MemberName");
MemberName member = new MemberName(refKind, defc, name, type);
MethodHandle mh = LOOKASIDE_TABLE.get(member);
if (mh != null) {
checkSymbolicClass(defc);
return mh;
}
if (defc == MethodHandle.class && refKind == REF_invokeVirtual) {
// Treat MethodHandle.invoke and invokeExact specially.
mh = findVirtualForMH(member.getName(), member.getMethodType());
if (mh != null) {
return mh;
}
} else if (defc == VarHandle.class && refKind == REF_invokeVirtual) {
// Treat signature-polymorphic methods on VarHandle specially.
mh = findVirtualForVH(member.getName(), member.getMethodType());
if (mh != null) {
return mh;
}
}
MemberName resolved = resolveOrFail(refKind, member);
mh = getDirectMethodForConstant(refKind, defc, resolved);
if (mh instanceof DirectMethodHandle dmh
&& canBeCached(refKind, defc, resolved)) {
MemberName key = mh.internalMemberName();
if (key != null) {
key = key.asNormalOriginal();
}
if (member.equals(key)) { // better safe than sorry
LOOKASIDE_TABLE.put(key, dmh);
}
}
return mh;
}
private boolean canBeCached(byte refKind, Class<?> defc, MemberName member) {
if (refKind == REF_invokeSpecial) {
return false;
}
if (!Modifier.isPublic(defc.getModifiers()) ||
!Modifier.isPublic(member.getDeclaringClass().getModifiers()) ||
!member.isPublic() ||
member.isCallerSensitive()) {
return false;
}
ClassLoader loader = defc.getClassLoader();
if (loader != null) {
ClassLoader sysl = ClassLoader.getSystemClassLoader();
boolean found = false;
while (sysl != null) {
if (loader == sysl) { found = true; break; }
sysl = sysl.getParent();
}
if (!found) {
return false;
}
}
MemberName resolved2 = publicLookup().resolveOrNull(refKind,
new MemberName(refKind, defc, member.getName(), member.getType()));
if (resolved2 == null) {
return false;
}
return true;
}
private MethodHandle getDirectMethodForConstant(byte refKind, Class<?> defc, MemberName member)
throws ReflectiveOperationException {
if (MethodHandleNatives.refKindIsField(refKind)) {
return getDirectField(refKind, defc, member);
} else if (MethodHandleNatives.refKindIsMethod(refKind)) {
return getDirectMethod(refKind, defc, member, findBoundCallerLookup(member));
} else if (refKind == REF_newInvokeSpecial) {
return getDirectConstructor(defc, member);
}
// oops
throw newIllegalArgumentException("bad MethodHandle constant #"+member);
}
static ConcurrentHashMap<MemberName, DirectMethodHandle> LOOKASIDE_TABLE = new ConcurrentHashMap<>();
}
/**
* Produces a method handle constructing arrays of a desired type,
* as if by the {@code anewarray} bytecode.
* The return type of the method handle will be the array type.
* The type of its sole argument will be {@code int}, which specifies the size of the array.
*
* <p> If the returned method handle is invoked with a negative
* array size, a {@code NegativeArraySizeException} will be thrown.
*
* @param arrayClass an array type
* @return a method handle which can create arrays of the given type
* @throws NullPointerException if the argument is {@code null}
* @throws IllegalArgumentException if {@code arrayClass} is not an array type
* @see java.lang.reflect.Array#newInstance(Class, int)
* @jvms 6.5 {@code anewarray} Instruction
* @since 9
*/
public static MethodHandle arrayConstructor(Class<?> arrayClass) throws IllegalArgumentException {
if (!arrayClass.isArray()) {
throw newIllegalArgumentException("not an array class: " + arrayClass.getName());
}
MethodHandle ani = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_Array_newInstance).
bindTo(arrayClass.getComponentType());
return ani.asType(ani.type().changeReturnType(arrayClass));
}
/**
* Produces a method handle returning the length of an array,
* as if by the {@code arraylength} bytecode.
* The type of the method handle will have {@code int} as return type,
* and its sole argument will be the array type.
*
* <p> If the returned method handle is invoked with a {@code null}
* array reference, a {@code NullPointerException} will be thrown.
*
* @param arrayClass an array type
* @return a method handle which can retrieve the length of an array of the given array type
* @throws NullPointerException if the argument is {@code null}
* @throws IllegalArgumentException if arrayClass is not an array type
* @jvms 6.5 {@code arraylength} Instruction
* @since 9
*/
public static MethodHandle arrayLength(Class<?> arrayClass) throws IllegalArgumentException {
return MethodHandleImpl.makeArrayElementAccessor(arrayClass, MethodHandleImpl.ArrayAccess.LENGTH);
}
/**
* Produces a method handle giving read access to elements of an array,
* as if by the {@code aaload} bytecode.
* The type of the method handle will have a return type of the array's
* element type. Its first argument will be the array type,
* and the second will be {@code int}.
*
* <p> When the returned method handle is invoked,
* the array reference and array index are checked.
* A {@code NullPointerException} will be thrown if the array reference
* is {@code null} and an {@code ArrayIndexOutOfBoundsException} will be
* thrown if the index is negative or if it is greater than or equal to
* the length of the array.
*
* @param arrayClass an array type
* @return a method handle which can load values from the given array type
* @throws NullPointerException if the argument is null
* @throws IllegalArgumentException if arrayClass is not an array type
* @jvms 6.5 {@code aaload} Instruction
*/
public static MethodHandle arrayElementGetter(Class<?> arrayClass) throws IllegalArgumentException {
return MethodHandleImpl.makeArrayElementAccessor(arrayClass, MethodHandleImpl.ArrayAccess.GET);
}
/**
* Produces a method handle giving write access to elements of an array,
* as if by the {@code astore} bytecode.
* The type of the method handle will have a void return type.
* Its last argument will be the array's element type.
* The first and second arguments will be the array type and int.
*
* <p> When the returned method handle is invoked,
* the array reference and array index are checked.
* A {@code NullPointerException} will be thrown if the array reference
* is {@code null} and an {@code ArrayIndexOutOfBoundsException} will be
* thrown if the index is negative or if it is greater than or equal to
* the length of the array.
*
* @param arrayClass the class of an array
* @return a method handle which can store values into the array type
* @throws NullPointerException if the argument is null
* @throws IllegalArgumentException if arrayClass is not an array type
* @jvms 6.5 {@code aastore} Instruction
*/
public static MethodHandle arrayElementSetter(Class<?> arrayClass) throws IllegalArgumentException {
return MethodHandleImpl.makeArrayElementAccessor(arrayClass, MethodHandleImpl.ArrayAccess.SET);
}
/**
* Produces a VarHandle giving access to elements of an array of type
* {@code arrayClass}. The VarHandle's variable type is the component type
* of {@code arrayClass} and the list of coordinate types is
* {@code (arrayClass, int)}, where the {@code int} coordinate type
* corresponds to an argument that is an index into an array.
* <p>
* Certain access modes of the returned VarHandle are unsupported under
* the following conditions:
* <ul>
* <li>if the component type is anything other than {@code byte},
* {@code short}, {@code char}, {@code int}, {@code long},
* {@code float}, or {@code double} then numeric atomic update access
* modes are unsupported.
* <li>if the component type is anything other than {@code boolean},
* {@code byte}, {@code short}, {@code char}, {@code int} or
* {@code long} then bitwise atomic update access modes are
* unsupported.
* </ul>
* <p>
* If the component type is {@code float} or {@code double} then numeric
* and atomic update access modes compare values using their bitwise
* representation (see {@link Float#floatToRawIntBits} and
* {@link Double#doubleToRawLongBits}, respectively).
*
* <p> When the returned {@code VarHandle} is invoked,
* the array reference and array index are checked.
* A {@code NullPointerException} will be thrown if the array reference
* is {@code null} and an {@code ArrayIndexOutOfBoundsException} will be
* thrown if the index is negative or if it is greater than or equal to
* the length of the array.
*
* @apiNote
* Bitwise comparison of {@code float} values or {@code double} values,
* as performed by the numeric and atomic update access modes, differ
* from the primitive {@code ==} operator and the {@link Float#equals}
* and {@link Double#equals} methods, specifically with respect to
* comparing NaN values or comparing {@code -0.0} with {@code +0.0}.
* Care should be taken when performing a compare and set or a compare
* and exchange operation with such values since the operation may
* unexpectedly fail.
* There are many possible NaN values that are considered to be
* {@code NaN} in Java, although no IEEE 754 floating-point operation
* provided by Java can distinguish between them. Operation failure can
* occur if the expected or witness value is a NaN value and it is
* transformed (perhaps in a platform specific manner) into another NaN
* value, and thus has a different bitwise representation (see
* {@link Float#intBitsToFloat} or {@link Double#longBitsToDouble} for more
* details).
* The values {@code -0.0} and {@code +0.0} have different bitwise
* representations but are considered equal when using the primitive
* {@code ==} operator. Operation failure can occur if, for example, a
* numeric algorithm computes an expected value to be say {@code -0.0}
* and previously computed the witness value to be say {@code +0.0}.
* @param arrayClass the class of an array, of type {@code T[]}
* @return a VarHandle giving access to elements of an array
* @throws NullPointerException if the arrayClass is null
* @throws IllegalArgumentException if arrayClass is not an array type
* @since 9
*/
public static VarHandle arrayElementVarHandle(Class<?> arrayClass) throws IllegalArgumentException {
return VarHandles.makeArrayElementHandle(arrayClass);
}
/**
* Produces a VarHandle giving access to elements of a {@code byte[]} array
* viewed as if it were a different primitive array type, such as
* {@code int[]} or {@code long[]}.
* The VarHandle's variable type is the component type of
* {@code viewArrayClass} and the list of coordinate types is
* {@code (byte[], int)}, where the {@code int} coordinate type
* corresponds to an argument that is an index into a {@code byte[]} array.
* The returned VarHandle accesses bytes at an index in a {@code byte[]}
* array, composing bytes to or from a value of the component type of
* {@code viewArrayClass} according to the given endianness.
* <p>
* The supported component types (variables types) are {@code short},
* {@code char}, {@code int}, {@code long}, {@code float} and
* {@code double}.
* <p>
* Access of bytes at a given index will result in an
* {@code ArrayIndexOutOfBoundsException} if the index is less than {@code 0}
* or greater than the {@code byte[]} array length minus the size (in bytes)
* of {@code T}.
* <p>
* Only plain {@linkplain VarHandle.AccessMode#GET get} and {@linkplain VarHandle.AccessMode#SET set}
* access modes are supported by the returned var handle. For all other access modes, an
* {@link UnsupportedOperationException} will be thrown.
*
* @apiNote if access modes other than plain access are required, clients should
* consider using off-heap memory through
* {@linkplain java.nio.ByteBuffer#allocateDirect(int) direct byte buffers} or
* off-heap {@linkplain java.lang.foreign.MemorySegment memory segments},
* or memory segments backed by a
* {@linkplain java.lang.foreign.MemorySegment#ofArray(long[]) {@code long[]}},
* for which stronger alignment guarantees can be made.
*
* @param viewArrayClass the view array class, with a component type of
* type {@code T}
* @param byteOrder the endianness of the view array elements, as
* stored in the underlying {@code byte} array
* @return a VarHandle giving access to elements of a {@code byte[]} array
* viewed as if elements corresponding to the components type of the view
* array class
* @throws NullPointerException if viewArrayClass or byteOrder is null
* @throws IllegalArgumentException if viewArrayClass is not an array type
* @throws UnsupportedOperationException if the component type of
* viewArrayClass is not supported as a variable type
* @since 9
*/
public static VarHandle byteArrayViewVarHandle(Class<?> viewArrayClass,
ByteOrder byteOrder) throws IllegalArgumentException {
Objects.requireNonNull(byteOrder);
return VarHandles.byteArrayViewHandle(viewArrayClass,
byteOrder == ByteOrder.BIG_ENDIAN);
}
/**
* Produces a VarHandle giving access to elements of a {@code ByteBuffer}
* viewed as if it were an array of elements of a different primitive
* component type to that of {@code byte}, such as {@code int[]} or
* {@code long[]}.
* The VarHandle's variable type is the component type of
* {@code viewArrayClass} and the list of coordinate types is
* {@code (ByteBuffer, int)}, where the {@code int} coordinate type
* corresponds to an argument that is an index into a {@code byte[]} array.
* The returned VarHandle accesses bytes at an index in a
* {@code ByteBuffer}, composing bytes to or from a value of the component
* type of {@code viewArrayClass} according to the given endianness.
* <p>
* The supported component types (variables types) are {@code short},
* {@code char}, {@code int}, {@code long}, {@code float} and
* {@code double}.
* <p>
* Access will result in a {@code ReadOnlyBufferException} for anything
* other than the read access modes if the {@code ByteBuffer} is read-only.
* <p>
* Access of bytes at a given index will result in an
* {@code IndexOutOfBoundsException} if the index is less than {@code 0}
* or greater than the {@code ByteBuffer} limit minus the size (in bytes) of
* {@code T}.
* <p>
* For heap byte buffers, access is always unaligned. As a result, only the plain
* {@linkplain VarHandle.AccessMode#GET get}
* and {@linkplain VarHandle.AccessMode#SET set} access modes are supported by the
* returned var handle. For all other access modes, an {@link IllegalStateException}
* will be thrown.
* <p>
* For direct buffers only, access of bytes at an index may be aligned or misaligned for {@code T},
* with respect to the underlying memory address, {@code A} say, associated
* with the {@code ByteBuffer} and index.
* If access is misaligned then access for anything other than the
* {@code get} and {@code set} access modes will result in an
* {@code IllegalStateException}. In such cases atomic access is only
* guaranteed with respect to the largest power of two that divides the GCD
* of {@code A} and the size (in bytes) of {@code T}.
* If access is aligned then following access modes are supported and are
* guaranteed to support atomic access:
* <ul>
* <li>read write access modes for all {@code T}. Access modes {@code get}
* and {@code set} for {@code long} and {@code double} are supported but
* have no atomicity guarantee, as described in Section {@jls 17.7} of
* <cite>The Java Language Specification</cite>.
* <li>atomic update access modes for {@code int}, {@code long},
* {@code float} or {@code double}.
* (Future major platform releases of the JDK may support additional
* types for certain currently unsupported access modes.)
* <li>numeric atomic update access modes for {@code int} and {@code long}.
* (Future major platform releases of the JDK may support additional
* numeric types for certain currently unsupported access modes.)
* <li>bitwise atomic update access modes for {@code int} and {@code long}.
* (Future major platform releases of the JDK may support additional
* numeric types for certain currently unsupported access modes.)
* </ul>
* <p>
* Misaligned access, and therefore atomicity guarantees, may be determined
* for a {@code ByteBuffer}, {@code bb} (direct or otherwise), an
* {@code index}, {@code T} and its corresponding boxed type,
* {@code T_BOX}, as follows:
* <pre>{@code
* int sizeOfT = T_BOX.BYTES; // size in bytes of T
* ByteBuffer bb = ...
* int misalignedAtIndex = bb.alignmentOffset(index, sizeOfT);
* boolean isMisaligned = misalignedAtIndex != 0;
* }</pre>
* <p>
* If the variable type is {@code float} or {@code double} then atomic
* update access modes compare values using their bitwise representation
* (see {@link Float#floatToRawIntBits} and
* {@link Double#doubleToRawLongBits}, respectively).
* @param viewArrayClass the view array class, with a component type of
* type {@code T}
* @param byteOrder the endianness of the view array elements, as
* stored in the underlying {@code ByteBuffer} (Note this overrides the
* endianness of a {@code ByteBuffer})
* @return a VarHandle giving access to elements of a {@code ByteBuffer}
* viewed as if elements corresponding to the components type of the view
* array class
* @throws NullPointerException if viewArrayClass or byteOrder is null
* @throws IllegalArgumentException if viewArrayClass is not an array type
* @throws UnsupportedOperationException if the component type of
* viewArrayClass is not supported as a variable type
* @since 9
*/
public static VarHandle byteBufferViewVarHandle(Class<?> viewArrayClass,
ByteOrder byteOrder) throws IllegalArgumentException {
Objects.requireNonNull(byteOrder);
return VarHandles.makeByteBufferViewHandle(viewArrayClass,
byteOrder == ByteOrder.BIG_ENDIAN);
}
//--- method handle invocation (reflective style)
/**
* Produces a method handle which will invoke any method handle of the
* given {@code type}, with a given number of trailing arguments replaced by
* a single trailing {@code Object[]} array.
* The resulting invoker will be a method handle with the following
* arguments:
* <ul>
* <li>a single {@code MethodHandle} target
* <li>zero or more leading values (counted by {@code leadingArgCount})
* <li>an {@code Object[]} array containing trailing arguments
* </ul>
* <p>
* The invoker will invoke its target like a call to {@link MethodHandle#invoke invoke} with
* the indicated {@code type}.
* That is, if the target is exactly of the given {@code type}, it will behave
* like {@code invokeExact}; otherwise it behave as if {@link MethodHandle#asType asType}
* is used to convert the target to the required {@code type}.
* <p>
* The type of the returned invoker will not be the given {@code type}, but rather
* will have all parameters except the first {@code leadingArgCount}
* replaced by a single array of type {@code Object[]}, which will be
* the final parameter.
* <p>
* Before invoking its target, the invoker will spread the final array, apply
* reference casts as necessary, and unbox and widen primitive arguments.
* If, when the invoker is called, the supplied array argument does
* not have the correct number of elements, the invoker will throw
* an {@link IllegalArgumentException} instead of invoking the target.
* <p>
* This method is equivalent to the following code (though it may be more efficient):
* {@snippet lang="java" :
MethodHandle invoker = MethodHandles.invoker(type);
int spreadArgCount = type.parameterCount() - leadingArgCount;
invoker = invoker.asSpreader(Object[].class, spreadArgCount);
return invoker;
* }
* This method throws no reflective exceptions.
* @param type the desired target type
* @param leadingArgCount number of fixed arguments, to be passed unchanged to the target
* @return a method handle suitable for invoking any method handle of the given type
* @throws NullPointerException if {@code type} is null
* @throws IllegalArgumentException if {@code leadingArgCount} is not in
* the range from 0 to {@code type.parameterCount()} inclusive,
* or if the resulting method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
*/
public static MethodHandle spreadInvoker(MethodType type, int leadingArgCount) {
if (leadingArgCount < 0 || leadingArgCount > type.parameterCount())
throw newIllegalArgumentException("bad argument count", leadingArgCount);
type = type.asSpreaderType(Object[].class, leadingArgCount, type.parameterCount() - leadingArgCount);
return type.invokers().spreadInvoker(leadingArgCount);
}
/**
* Produces a special <em>invoker method handle</em> which can be used to
* invoke any method handle of the given type, as if by {@link MethodHandle#invokeExact invokeExact}.
* The resulting invoker will have a type which is
* exactly equal to the desired type, except that it will accept
* an additional leading argument of type {@code MethodHandle}.
* <p>
* This method is equivalent to the following code (though it may be more efficient):
* {@code publicLookup().findVirtual(MethodHandle.class, "invokeExact", type)}
*
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* Invoker method handles can be useful when working with variable method handles
* of unknown types.
* For example, to emulate an {@code invokeExact} call to a variable method
* handle {@code M}, extract its type {@code T},
* look up the invoker method {@code X} for {@code T},
* and call the invoker method, as {@code X.invoke(T, A...)}.
* (It would not work to call {@code X.invokeExact}, since the type {@code T}
* is unknown.)
* If spreading, collecting, or other argument transformations are required,
* they can be applied once to the invoker {@code X} and reused on many {@code M}
* method handle values, as long as they are compatible with the type of {@code X}.
* <p style="font-size:smaller;">
* <em>(Note: The invoker method is not available via the Core Reflection API.
* An attempt to call {@linkplain java.lang.reflect.Method#invoke java.lang.reflect.Method.invoke}
* on the declared {@code invokeExact} or {@code invoke} method will raise an
* {@link java.lang.UnsupportedOperationException UnsupportedOperationException}.)</em>
* <p>
* This method throws no reflective exceptions.
* @param type the desired target type
* @return a method handle suitable for invoking any method handle of the given type
* @throws IllegalArgumentException if the resulting method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
*/
public static MethodHandle exactInvoker(MethodType type) {
return type.invokers().exactInvoker();
}
/**
* Produces a special <em>invoker method handle</em> which can be used to
* invoke any method handle compatible with the given type, as if by {@link MethodHandle#invoke invoke}.
* The resulting invoker will have a type which is
* exactly equal to the desired type, except that it will accept
* an additional leading argument of type {@code MethodHandle}.
* <p>
* Before invoking its target, if the target differs from the expected type,
* the invoker will apply reference casts as
* necessary and box, unbox, or widen primitive values, as if by {@link MethodHandle#asType asType}.
* Similarly, the return value will be converted as necessary.
* If the target is a {@linkplain MethodHandle#asVarargsCollector variable arity method handle},
* the required arity conversion will be made, again as if by {@link MethodHandle#asType asType}.
* <p>
* This method is equivalent to the following code (though it may be more efficient):
* {@code publicLookup().findVirtual(MethodHandle.class, "invoke", type)}
* <p style="font-size:smaller;">
* <em>Discussion:</em>
* A {@linkplain MethodType#genericMethodType general method type} is one which
* mentions only {@code Object} arguments and return values.
* An invoker for such a type is capable of calling any method handle
* of the same arity as the general type.
* <p style="font-size:smaller;">
* <em>(Note: The invoker method is not available via the Core Reflection API.
* An attempt to call {@linkplain java.lang.reflect.Method#invoke java.lang.reflect.Method.invoke}
* on the declared {@code invokeExact} or {@code invoke} method will raise an
* {@link java.lang.UnsupportedOperationException UnsupportedOperationException}.)</em>
* <p>
* This method throws no reflective exceptions.
* @param type the desired target type
* @return a method handle suitable for invoking any method handle convertible to the given type
* @throws IllegalArgumentException if the resulting method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
*/
public static MethodHandle invoker(MethodType type) {
return type.invokers().genericInvoker();
}
/**
* Produces a special <em>invoker method handle</em> which can be used to
* invoke a signature-polymorphic access mode method on any VarHandle whose
* associated access mode type is compatible with the given type.
* The resulting invoker will have a type which is exactly equal to the
* desired given type, except that it will accept an additional leading
* argument of type {@code VarHandle}.
*
* @param accessMode the VarHandle access mode
* @param type the desired target type
* @return a method handle suitable for invoking an access mode method of
* any VarHandle whose access mode type is of the given type.
* @since 9
*/
public static MethodHandle varHandleExactInvoker(VarHandle.AccessMode accessMode, MethodType type) {
return type.invokers().varHandleMethodExactInvoker(accessMode);
}
/**
* Produces a special <em>invoker method handle</em> which can be used to
* invoke a signature-polymorphic access mode method on any VarHandle whose
* associated access mode type is compatible with the given type.
* The resulting invoker will have a type which is exactly equal to the
* desired given type, except that it will accept an additional leading
* argument of type {@code VarHandle}.
* <p>
* Before invoking its target, if the access mode type differs from the
* desired given type, the invoker will apply reference casts as necessary
* and box, unbox, or widen primitive values, as if by
* {@link MethodHandle#asType asType}. Similarly, the return value will be
* converted as necessary.
* <p>
* This method is equivalent to the following code (though it may be more
* efficient): {@code publicLookup().findVirtual(VarHandle.class, accessMode.name(), type)}
*
* @param accessMode the VarHandle access mode
* @param type the desired target type
* @return a method handle suitable for invoking an access mode method of
* any VarHandle whose access mode type is convertible to the given
* type.
* @since 9
*/
public static MethodHandle varHandleInvoker(VarHandle.AccessMode accessMode, MethodType type) {
return type.invokers().varHandleMethodInvoker(accessMode);
}
/*non-public*/
static MethodHandle basicInvoker(MethodType type) {
return type.invokers().basicInvoker();
}
//--- method handle modification (creation from other method handles)
/**
* Produces a method handle which adapts the type of the
* given method handle to a new type by pairwise argument and return type conversion.
* The original type and new type must have the same number of arguments.
* The resulting method handle is guaranteed to report a type
* which is equal to the desired new type.
* <p>
* If the original type and new type are equal, returns target.
* <p>
* The same conversions are allowed as for {@link MethodHandle#asType MethodHandle.asType},
* and some additional conversions are also applied if those conversions fail.
* Given types <em>T0</em>, <em>T1</em>, one of the following conversions is applied
* if possible, before or instead of any conversions done by {@code asType}:
* <ul>
* <li>If <em>T0</em> and <em>T1</em> are references, and <em>T1</em> is an interface type,
* then the value of type <em>T0</em> is passed as a <em>T1</em> without a cast.
* (This treatment of interfaces follows the usage of the bytecode verifier.)
* <li>If <em>T0</em> is boolean and <em>T1</em> is another primitive,
* the boolean is converted to a byte value, 1 for true, 0 for false.
* (This treatment follows the usage of the bytecode verifier.)
* <li>If <em>T1</em> is boolean and <em>T0</em> is another primitive,
* <em>T0</em> is converted to byte via Java casting conversion (JLS {@jls 5.5}),
* and the low order bit of the result is tested, as if by {@code (x & 1) != 0}.
* <li>If <em>T0</em> and <em>T1</em> are primitives other than boolean,
* then a Java casting conversion (JLS {@jls 5.5}) is applied.
* (Specifically, <em>T0</em> will convert to <em>T1</em> by
* widening and/or narrowing.)
* <li>If <em>T0</em> is a reference and <em>T1</em> a primitive, an unboxing
* conversion will be applied at runtime, possibly followed
* by a Java casting conversion (JLS {@jls 5.5}) on the primitive value,
* possibly followed by a conversion from byte to boolean by testing
* the low-order bit.
* <li>If <em>T0</em> is a reference and <em>T1</em> a primitive,
* and if the reference is null at runtime, a zero value is introduced.
* </ul>
* @param target the method handle to invoke after arguments are retyped
* @param newType the expected type of the new method handle
* @return a method handle which delegates to the target after performing
* any necessary argument conversions, and arranges for any
* necessary return value conversions
* @throws NullPointerException if either argument is null
* @throws WrongMethodTypeException if the conversion cannot be made
* @see MethodHandle#asType
*/
public static MethodHandle explicitCastArguments(MethodHandle target, MethodType newType) {
explicitCastArgumentsChecks(target, newType);
// use the asTypeCache when possible:
MethodType oldType = target.type();
if (oldType == newType) return target;
if (oldType.explicitCastEquivalentToAsType(newType)) {
return target.asFixedArity().asType(newType);
}
return MethodHandleImpl.makePairwiseConvert(target, newType, false);
}
private static void explicitCastArgumentsChecks(MethodHandle target, MethodType newType) {
if (target.type().parameterCount() != newType.parameterCount()) {
throw new WrongMethodTypeException("cannot explicitly cast " + target + " to " + newType);
}
}
/**
* Produces a method handle which adapts the calling sequence of the
* given method handle to a new type, by reordering the arguments.
* The resulting method handle is guaranteed to report a type
* which is equal to the desired new type.
* <p>
* The given array controls the reordering.
* Call {@code #I} the number of incoming parameters (the value
* {@code newType.parameterCount()}, and call {@code #O} the number
* of outgoing parameters (the value {@code target.type().parameterCount()}).
* Then the length of the reordering array must be {@code #O},
* and each element must be a non-negative number less than {@code #I}.
* For every {@code N} less than {@code #O}, the {@code N}-th
* outgoing argument will be taken from the {@code I}-th incoming
* argument, where {@code I} is {@code reorder[N]}.
* <p>
* No argument or return value conversions are applied.
* The type of each incoming argument, as determined by {@code newType},
* must be identical to the type of the corresponding outgoing parameter
* or parameters in the target method handle.
* The return type of {@code newType} must be identical to the return
* type of the original target.
* <p>
* The reordering array need not specify an actual permutation.
* An incoming argument will be duplicated if its index appears
* more than once in the array, and an incoming argument will be dropped
* if its index does not appear in the array.
* As in the case of {@link #dropArguments(MethodHandle,int,List) dropArguments},
* incoming arguments which are not mentioned in the reordering array
* may be of any type, as determined only by {@code newType}.
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodType intfn1 = methodType(int.class, int.class);
MethodType intfn2 = methodType(int.class, int.class, int.class);
MethodHandle sub = ... (int x, int y) -> (x-y) ...;
assert(sub.type().equals(intfn2));
MethodHandle sub1 = permuteArguments(sub, intfn2, 0, 1);
MethodHandle rsub = permuteArguments(sub, intfn2, 1, 0);
assert((int)rsub.invokeExact(1, 100) == 99);
MethodHandle add = ... (int x, int y) -> (x+y) ...;
assert(add.type().equals(intfn2));
MethodHandle twice = permuteArguments(add, intfn1, 0, 0);
assert(twice.type().equals(intfn1));
assert((int)twice.invokeExact(21) == 42);
* }
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
* @param target the method handle to invoke after arguments are reordered
* @param newType the expected type of the new method handle
* @param reorder an index array which controls the reordering
* @return a method handle which delegates to the target after it
* drops unused arguments and moves and/or duplicates the other arguments
* @throws NullPointerException if any argument is null
* @throws IllegalArgumentException if the index array length is not equal to
* the arity of the target, or if any index array element
* not a valid index for a parameter of {@code newType},
* or if two corresponding parameter types in
* {@code target.type()} and {@code newType} are not identical,
*/
public static MethodHandle permuteArguments(MethodHandle target, MethodType newType, int... reorder) {
reorder = reorder.clone(); // get a private copy
MethodType oldType = target.type();
permuteArgumentChecks(reorder, newType, oldType);
// first detect dropped arguments and handle them separately
int[] originalReorder = reorder;
BoundMethodHandle result = target.rebind();
LambdaForm form = result.form;
int newArity = newType.parameterCount();
// Normalize the reordering into a real permutation,
// by removing duplicates and adding dropped elements.
// This somewhat improves lambda form caching, as well
// as simplifying the transform by breaking it up into steps.
for (int ddIdx; (ddIdx = findFirstDupOrDrop(reorder, newArity)) != 0; ) {
if (ddIdx > 0) {
// We found a duplicated entry at reorder[ddIdx].
// Example: (x,y,z)->asList(x,y,z)
// permuted by [1*,0,1] => (a0,a1)=>asList(a1,a0,a1)
// permuted by [0,1,0*] => (a0,a1)=>asList(a0,a1,a0)
// The starred element corresponds to the argument
// deleted by the dupArgumentForm transform.
int srcPos = ddIdx, dstPos = srcPos, dupVal = reorder[srcPos];
boolean killFirst = false;
for (int val; (val = reorder[--dstPos]) != dupVal; ) {
// Set killFirst if the dup is larger than an intervening position.
// This will remove at least one inversion from the permutation.
if (dupVal > val) killFirst = true;
}
if (!killFirst) {
srcPos = dstPos;
dstPos = ddIdx;
}
form = form.editor().dupArgumentForm(1 + srcPos, 1 + dstPos);
assert (reorder[srcPos] == reorder[dstPos]);
oldType = oldType.dropParameterTypes(dstPos, dstPos + 1);
// contract the reordering by removing the element at dstPos
int tailPos = dstPos + 1;
System.arraycopy(reorder, tailPos, reorder, dstPos, reorder.length - tailPos);
reorder = Arrays.copyOf(reorder, reorder.length - 1);
} else {
int dropVal = ~ddIdx, insPos = 0;
while (insPos < reorder.length && reorder[insPos] < dropVal) {
// Find first element of reorder larger than dropVal.
// This is where we will insert the dropVal.
insPos += 1;
}
Class<?> ptype = newType.parameterType(dropVal);
form = form.editor().addArgumentForm(1 + insPos, BasicType.basicType(ptype));
oldType = oldType.insertParameterTypes(insPos, ptype);
// expand the reordering by inserting an element at insPos
int tailPos = insPos + 1;
reorder = Arrays.copyOf(reorder, reorder.length + 1);
System.arraycopy(reorder, insPos, reorder, tailPos, reorder.length - tailPos);
reorder[insPos] = dropVal;
}
assert (permuteArgumentChecks(reorder, newType, oldType));
}
assert (reorder.length == newArity); // a perfect permutation
// Note: This may cache too many distinct LFs. Consider backing off to varargs code.
form = form.editor().permuteArgumentsForm(1, reorder);
if (newType == result.type() && form == result.internalForm())
return result;
return result.copyWith(newType, form);
}
/**
* Return an indication of any duplicate or omission in reorder.
* If the reorder contains a duplicate entry, return the index of the second occurrence.
* Otherwise, return ~(n), for the first n in [0..newArity-1] that is not present in reorder.
* Otherwise, return zero.
* If an element not in [0..newArity-1] is encountered, return reorder.length.
*/
private static int findFirstDupOrDrop(int[] reorder, int newArity) {
final int BIT_LIMIT = 63; // max number of bits in bit mask
if (newArity < BIT_LIMIT) {
long mask = 0;
for (int i = 0; i < reorder.length; i++) {
int arg = reorder[i];
if (arg >= newArity) {
return reorder.length;
}
long bit = 1L << arg;
if ((mask & bit) != 0) {
return i; // >0 indicates a dup
}
mask |= bit;
}
if (mask == (1L << newArity) - 1) {
assert(Long.numberOfTrailingZeros(Long.lowestOneBit(~mask)) == newArity);
return 0;
}
// find first zero
long zeroBit = Long.lowestOneBit(~mask);
int zeroPos = Long.numberOfTrailingZeros(zeroBit);
assert(zeroPos <= newArity);
if (zeroPos == newArity) {
return 0;
}
return ~zeroPos;
} else {
// same algorithm, different bit set
BitSet mask = new BitSet(newArity);
for (int i = 0; i < reorder.length; i++) {
int arg = reorder[i];
if (arg >= newArity) {
return reorder.length;
}
if (mask.get(arg)) {
return i; // >0 indicates a dup
}
mask.set(arg);
}
int zeroPos = mask.nextClearBit(0);
assert(zeroPos <= newArity);
if (zeroPos == newArity) {
return 0;
}
return ~zeroPos;
}
}
static boolean permuteArgumentChecks(int[] reorder, MethodType newType, MethodType oldType) {
if (newType.returnType() != oldType.returnType())
throw newIllegalArgumentException("return types do not match",
oldType, newType);
if (reorder.length != oldType.parameterCount())
throw newIllegalArgumentException("old type parameter count and reorder array length do not match",
oldType, Arrays.toString(reorder));
int limit = newType.parameterCount();
for (int j = 0; j < reorder.length; j++) {
int i = reorder[j];
if (i < 0 || i >= limit) {
throw newIllegalArgumentException("index is out of bounds for new type",
i, newType);
}
Class<?> src = newType.parameterType(i);
Class<?> dst = oldType.parameterType(j);
if (src != dst)
throw newIllegalArgumentException("parameter types do not match after reorder",
oldType, newType);
}
return true;
}
/**
* Produces a method handle of the requested return type which returns the given
* constant value every time it is invoked.
* <p>
* Before the method handle is returned, the passed-in value is converted to the requested type.
* If the requested type is primitive, widening primitive conversions are attempted,
* else reference conversions are attempted.
* <p>The returned method handle is equivalent to {@code identity(type).bindTo(value)}.
* @param type the return type of the desired method handle
* @param value the value to return
* @return a method handle of the given return type and no arguments, which always returns the given value
* @throws NullPointerException if the {@code type} argument is null
* @throws ClassCastException if the value cannot be converted to the required return type
* @throws IllegalArgumentException if the given type is {@code void.class}
*/
public static MethodHandle constant(Class<?> type, Object value) {
if (Objects.requireNonNull(type) == void.class)
throw newIllegalArgumentException("void type");
return MethodHandleImpl.makeConstantReturning(type, value);
}
/**
* Produces a method handle which returns its sole argument when invoked.
* @param type the type of the sole parameter and return value of the desired method handle
* @return a unary method handle which accepts and returns the given type
* @throws NullPointerException if the argument is null
* @throws IllegalArgumentException if the given type is {@code void.class}
*/
public static MethodHandle identity(Class<?> type) {
Wrapper btw = (type.isPrimitive() ? Wrapper.forPrimitiveType(type) : Wrapper.OBJECT);
int pos = btw.ordinal();
MethodHandle ident = IDENTITY_MHS[pos];
if (ident == null) {
ident = setCachedMethodHandle(IDENTITY_MHS, pos, makeIdentity(btw.primitiveType()));
}
if (ident.type().returnType() == type)
return ident;
// something like identity(Foo.class); do not bother to intern these
assert (btw == Wrapper.OBJECT);
return makeIdentity(type);
}
/**
* Produces a constant method handle of the requested return type which
* returns the default value for that type every time it is invoked.
* The resulting constant method handle will have no side effects.
* <p>The returned method handle is equivalent to {@code empty(methodType(type))}.
* It is also equivalent to {@code explicitCastArguments(constant(Object.class, null), methodType(type))},
* since {@code explicitCastArguments} converts {@code null} to default values.
* @param type the expected return type of the desired method handle
* @return a constant method handle that takes no arguments
* and returns the default value of the given type (or void, if the type is void)
* @throws NullPointerException if the argument is null
* @see MethodHandles#constant
* @see MethodHandles#empty
* @see MethodHandles#explicitCastArguments
* @since 9
*/
public static MethodHandle zero(Class<?> type) {
Objects.requireNonNull(type);
return type.isPrimitive() ? primitiveZero(Wrapper.forPrimitiveType(type))
: MethodHandleImpl.makeConstantReturning(type, null);
}
private static MethodHandle identityOrVoid(Class<?> type) {
return type == void.class ? zero(type) : identity(type);
}
/**
* Produces a method handle of the requested type which ignores any arguments, does nothing,
* and returns a suitable default depending on the return type.
* That is, it returns a zero primitive value, a {@code null}, or {@code void}.
* <p>The returned method handle is equivalent to
* {@code dropArguments(zero(type.returnType()), 0, type.parameterList())}.
*
* @apiNote Given a predicate and target, a useful "if-then" construct can be produced as
* {@code guardWithTest(pred, target, empty(target.type())}.
* @param type the type of the desired method handle
* @return a constant method handle of the given type, which returns a default value of the given return type
* @throws NullPointerException if the argument is null
* @see MethodHandles#zero(Class)
* @see MethodHandles#constant
* @since 9
*/
public static MethodHandle empty(MethodType type) {
Objects.requireNonNull(type);
return dropArgumentsTrusted(zero(type.returnType()), 0, type.ptypes());
}
private static final MethodHandle[] IDENTITY_MHS = new MethodHandle[Wrapper.COUNT];
private static MethodHandle makeIdentity(Class<?> ptype) {
MethodType mtype = methodType(ptype, ptype); // throws IAE for void
LambdaForm lform = LambdaForm.identityForm(BasicType.basicType(ptype));
return MethodHandleImpl.makeIntrinsic(mtype, lform, Intrinsic.IDENTITY);
}
private static MethodHandle primitiveZero(Wrapper w) {
assert w != Wrapper.OBJECT : w;
int pos = w.ordinal();
MethodHandle mh = PRIMITIVE_ZERO_MHS[pos];
if (mh == null) {
mh = setCachedMethodHandle(PRIMITIVE_ZERO_MHS, pos, makePrimitiveZero(w));
}
assert (mh.type().returnType() == w.primitiveType()) : mh;
return mh;
}
private static MethodHandle makePrimitiveZero(Wrapper w) {
if (w == Wrapper.VOID) {
var lf = LambdaForm.identityForm(V_TYPE); // ensures BMH & SimpleMH are initialized
return SimpleMethodHandle.make(MethodType.methodType(void.class), lf);
} else {
return MethodHandleImpl.makeConstantReturning(w.primitiveType(), w.zero());
}
}
private static final @Stable MethodHandle[] PRIMITIVE_ZERO_MHS = new MethodHandle[Wrapper.COUNT];
private static synchronized MethodHandle setCachedMethodHandle(MethodHandle[] cache, int pos, MethodHandle value) {
// Simulate a CAS, to avoid racy duplication of results.
MethodHandle prev = cache[pos];
if (prev != null) return prev;
return cache[pos] = value;
}
/**
* Provides a target method handle with one or more <em>bound arguments</em>
* in advance of the method handle's invocation.
* The formal parameters to the target corresponding to the bound
* arguments are called <em>bound parameters</em>.
* Returns a new method handle which saves away the bound arguments.
* When it is invoked, it receives arguments for any non-bound parameters,
* binds the saved arguments to their corresponding parameters,
* and calls the original target.
* <p>
* The type of the new method handle will drop the types for the bound
* parameters from the original target type, since the new method handle
* will no longer require those arguments to be supplied by its callers.
* <p>
* Each given argument object must match the corresponding bound parameter type.
* If a bound parameter type is a primitive, the argument object
* must be a wrapper, and will be unboxed to produce the primitive value.
* <p>
* The {@code pos} argument selects which parameters are to be bound.
* It may range between zero and <i>N-L</i> (inclusively),
* where <i>N</i> is the arity of the target method handle
* and <i>L</i> is the length of the values array.
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
* @param target the method handle to invoke after the argument is inserted
* @param pos where to insert the argument (zero for the first)
* @param values the series of arguments to insert
* @return a method handle which inserts an additional argument,
* before calling the original method handle
* @throws NullPointerException if the target or the {@code values} array is null
* @throws IllegalArgumentException if {@code pos} is less than {@code 0} or greater than
* {@code N - L} where {@code N} is the arity of the target method handle and {@code L}
* is the length of the values array.
* @throws ClassCastException if an argument does not match the corresponding bound parameter
* type.
* @see MethodHandle#bindTo
*/
public static MethodHandle insertArguments(MethodHandle target, int pos, Object... values) {
int insCount = values.length;
Class<?>[] ptypes = insertArgumentsChecks(target, insCount, pos);
if (insCount == 0) return target;
BoundMethodHandle result = target.rebind();
for (int i = 0; i < insCount; i++) {
Object value = values[i];
Class<?> ptype = ptypes[pos+i];
if (ptype.isPrimitive()) {
result = insertArgumentPrimitive(result, pos, ptype, value);
} else {
value = ptype.cast(value); // throw CCE if needed
result = result.bindArgumentL(pos, value);
}
}
return result;
}
private static BoundMethodHandle insertArgumentPrimitive(BoundMethodHandle result, int pos,
Class<?> ptype, Object value) {
Wrapper w = Wrapper.forPrimitiveType(ptype);
// perform unboxing and/or primitive conversion
value = w.convert(value, ptype);
return switch (w) {
case INT -> result.bindArgumentI(pos, (int) value);
case LONG -> result.bindArgumentJ(pos, (long) value);
case FLOAT -> result.bindArgumentF(pos, (float) value);
case DOUBLE -> result.bindArgumentD(pos, (double) value);
default -> result.bindArgumentI(pos, ValueConversions.widenSubword(value));
};
}
private static Class<?>[] insertArgumentsChecks(MethodHandle target, int insCount, int pos) throws RuntimeException {
MethodType oldType = target.type();
int outargs = oldType.parameterCount();
int inargs = outargs - insCount;
if (inargs < 0)
throw newIllegalArgumentException("too many values to insert");
if (pos < 0 || pos > inargs)
throw newIllegalArgumentException("no argument type to append");
return oldType.ptypes();
}
/**
* Produces a method handle which will discard some dummy arguments
* before calling some other specified <i>target</i> method handle.
* The type of the new method handle will be the same as the target's type,
* except it will also include the dummy argument types,
* at some given position.
* <p>
* The {@code pos} argument may range between zero and <i>N</i>,
* where <i>N</i> is the arity of the target.
* If {@code pos} is zero, the dummy arguments will precede
* the target's real arguments; if {@code pos} is <i>N</i>
* they will come after.
* <p>
* <b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
assertEquals("xy", (String) cat.invokeExact("x", "y"));
MethodType bigType = cat.type().insertParameterTypes(0, int.class, String.class);
MethodHandle d0 = dropArguments(cat, 0, bigType.parameterList().subList(0,2));
assertEquals(bigType, d0.type());
assertEquals("yz", (String) d0.invokeExact(123, "x", "y", "z"));
* }
* <p>
* This method is also equivalent to the following code:
* <blockquote><pre>
* {@link #dropArguments(MethodHandle,int,Class...) dropArguments}{@code (target, pos, valueTypes.toArray(new Class[0]))}
* </pre></blockquote>
* @param target the method handle to invoke after the arguments are dropped
* @param pos position of first argument to drop (zero for the leftmost)
* @param valueTypes the type(s) of the argument(s) to drop
* @return a method handle which drops arguments of the given types,
* before calling the original method handle
* @throws NullPointerException if the target is null,
* or if the {@code valueTypes} list or any of its elements is null
* @throws IllegalArgumentException if any element of {@code valueTypes} is {@code void.class},
* or if {@code pos} is negative or greater than the arity of the target,
* or if the new method handle's type would have too many parameters
*/
public static MethodHandle dropArguments(MethodHandle target, int pos, List<Class<?>> valueTypes) {
return dropArgumentsTrusted(target, pos, valueTypes.toArray(new Class<?>[0]).clone());
}
static MethodHandle dropArgumentsTrusted(MethodHandle target, int pos, Class<?>[] valueTypes) {
MethodType oldType = target.type(); // get NPE
int dropped = dropArgumentChecks(oldType, pos, valueTypes);
MethodType newType = oldType.insertParameterTypes(pos, valueTypes);
if (dropped == 0) return target;
BoundMethodHandle result = target.rebind();
LambdaForm lform = result.form;
int insertFormArg = 1 + pos;
for (Class<?> ptype : valueTypes) {
lform = lform.editor().addArgumentForm(insertFormArg++, BasicType.basicType(ptype));
}
result = result.copyWith(newType, lform);
return result;
}
private static int dropArgumentChecks(MethodType oldType, int pos, Class<?>[] valueTypes) {
int dropped = valueTypes.length;
MethodType.checkSlotCount(dropped);
int outargs = oldType.parameterCount();
int inargs = outargs + dropped;
if (pos < 0 || pos > outargs)
throw newIllegalArgumentException("no argument type to remove"
+ Arrays.asList(oldType, pos, valueTypes, inargs, outargs)
);
return dropped;
}
/**
* Produces a method handle which will discard some dummy arguments
* before calling some other specified <i>target</i> method handle.
* The type of the new method handle will be the same as the target's type,
* except it will also include the dummy argument types,
* at some given position.
* <p>
* The {@code pos} argument may range between zero and <i>N</i>,
* where <i>N</i> is the arity of the target.
* If {@code pos} is zero, the dummy arguments will precede
* the target's real arguments; if {@code pos} is <i>N</i>
* they will come after.
* @apiNote
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
assertEquals("xy", (String) cat.invokeExact("x", "y"));
MethodHandle d0 = dropArguments(cat, 0, String.class);
assertEquals("yz", (String) d0.invokeExact("x", "y", "z"));
MethodHandle d1 = dropArguments(cat, 1, String.class);
assertEquals("xz", (String) d1.invokeExact("x", "y", "z"));
MethodHandle d2 = dropArguments(cat, 2, String.class);
assertEquals("xy", (String) d2.invokeExact("x", "y", "z"));
MethodHandle d12 = dropArguments(cat, 1, int.class, boolean.class);
assertEquals("xz", (String) d12.invokeExact("x", 12, true, "z"));
* }
* <p>
* This method is also equivalent to the following code:
* <blockquote><pre>
* {@link #dropArguments(MethodHandle,int,List) dropArguments}{@code (target, pos, Arrays.asList(valueTypes))}
* </pre></blockquote>
* @param target the method handle to invoke after the arguments are dropped
* @param pos position of first argument to drop (zero for the leftmost)
* @param valueTypes the type(s) of the argument(s) to drop
* @return a method handle which drops arguments of the given types,
* before calling the original method handle
* @throws NullPointerException if the target is null,
* or if the {@code valueTypes} array or any of its elements is null
* @throws IllegalArgumentException if any element of {@code valueTypes} is {@code void.class},
* or if {@code pos} is negative or greater than the arity of the target,
* or if the new method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
*/
public static MethodHandle dropArguments(MethodHandle target, int pos, Class<?>... valueTypes) {
return dropArgumentsTrusted(target, pos, valueTypes.clone());
}
/* Convenience overloads for trusting internal low-arity call-sites */
static MethodHandle dropArguments(MethodHandle target, int pos, Class<?> valueType1) {
return dropArgumentsTrusted(target, pos, new Class<?>[] { valueType1 });
}
static MethodHandle dropArguments(MethodHandle target, int pos, Class<?> valueType1, Class<?> valueType2) {
return dropArgumentsTrusted(target, pos, new Class<?>[] { valueType1, valueType2 });
}
// private version which allows caller some freedom with error handling
private static MethodHandle dropArgumentsToMatch(MethodHandle target, int skip, Class<?>[] newTypes, int pos,
boolean nullOnFailure) {
Class<?>[] oldTypes = target.type().ptypes();
int match = oldTypes.length;
if (skip != 0) {
if (skip < 0 || skip > match) {
throw newIllegalArgumentException("illegal skip", skip, target);
}
oldTypes = Arrays.copyOfRange(oldTypes, skip, match);
match -= skip;
}
Class<?>[] addTypes = newTypes;
int add = addTypes.length;
if (pos != 0) {
if (pos < 0 || pos > add) {
throw newIllegalArgumentException("illegal pos", pos, Arrays.toString(newTypes));
}
addTypes = Arrays.copyOfRange(addTypes, pos, add);
add -= pos;
assert(addTypes.length == add);
}
// Do not add types which already match the existing arguments.
if (match > add || !Arrays.equals(oldTypes, 0, oldTypes.length, addTypes, 0, match)) {
if (nullOnFailure) {
return null;
}
throw newIllegalArgumentException("argument lists do not match",
Arrays.toString(oldTypes), Arrays.toString(newTypes));
}
addTypes = Arrays.copyOfRange(addTypes, match, add);
add -= match;
assert(addTypes.length == add);
// newTypes: ( P*[pos], M*[match], A*[add] )
// target: ( S*[skip], M*[match] )
MethodHandle adapter = target;
if (add > 0) {
adapter = dropArgumentsTrusted(adapter, skip+ match, addTypes);
}
// adapter: (S*[skip], M*[match], A*[add] )
if (pos > 0) {
adapter = dropArgumentsTrusted(adapter, skip, Arrays.copyOfRange(newTypes, 0, pos));
}
// adapter: (S*[skip], P*[pos], M*[match], A*[add] )
return adapter;
}
/**
* Adapts a target method handle to match the given parameter type list. If necessary, adds dummy arguments. Some
* leading parameters can be skipped before matching begins. The remaining types in the {@code target}'s parameter
* type list must be a sub-list of the {@code newTypes} type list at the starting position {@code pos}. The
* resulting handle will have the target handle's parameter type list, with any non-matching parameter types (before
* or after the matching sub-list) inserted in corresponding positions of the target's original parameters, as if by
* {@link #dropArguments(MethodHandle, int, Class[])}.
* <p>
* The resulting handle will have the same return type as the target handle.
* <p>
* In more formal terms, assume these two type lists:<ul>
* <li>The target handle has the parameter type list {@code S..., M...}, with as many types in {@code S} as
* indicated by {@code skip}. The {@code M} types are those that are supposed to match part of the given type list,
* {@code newTypes}.
* <li>The {@code newTypes} list contains types {@code P..., M..., A...}, with as many types in {@code P} as
* indicated by {@code pos}. The {@code M} types are precisely those that the {@code M} types in the target handle's
* parameter type list are supposed to match. The types in {@code A} are additional types found after the matching
* sub-list.
* </ul>
* Given these assumptions, the result of an invocation of {@code dropArgumentsToMatch} will have the parameter type
* list {@code S..., P..., M..., A...}, with the {@code P} and {@code A} types inserted as if by
* {@link #dropArguments(MethodHandle, int, Class[])}.
*
* @apiNote
* Two method handles whose argument lists are "effectively identical" (i.e., identical in a common prefix) may be
* mutually converted to a common type by two calls to {@code dropArgumentsToMatch}, as follows:
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
...
MethodHandle h0 = constant(boolean.class, true);
MethodHandle h1 = lookup().findVirtual(String.class, "concat", methodType(String.class, String.class));
MethodType bigType = h1.type().insertParameterTypes(1, String.class, int.class);
MethodHandle h2 = dropArguments(h1, 0, bigType.parameterList());
if (h1.type().parameterCount() < h2.type().parameterCount())
h1 = dropArgumentsToMatch(h1, 0, h2.type().parameterList(), 0); // lengthen h1
else
h2 = dropArgumentsToMatch(h2, 0, h1.type().parameterList(), 0); // lengthen h2
MethodHandle h3 = guardWithTest(h0, h1, h2);
assertEquals("xy", h3.invoke("x", "y", 1, "a", "b", "c"));
* }
* @param target the method handle to adapt
* @param skip number of targets parameters to disregard (they will be unchanged)
* @param newTypes the list of types to match {@code target}'s parameter type list to
* @param pos place in {@code newTypes} where the non-skipped target parameters must occur
* @return a possibly adapted method handle
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if any element of {@code newTypes} is {@code void.class},
* or if {@code skip} is negative or greater than the arity of the target,
* or if {@code pos} is negative or greater than the newTypes list size,
* or if {@code newTypes} does not contain the {@code target}'s non-skipped parameter types at position
* {@code pos}.
* @since 9
*/
public static MethodHandle dropArgumentsToMatch(MethodHandle target, int skip, List<Class<?>> newTypes, int pos) {
Objects.requireNonNull(target);
Objects.requireNonNull(newTypes);
return dropArgumentsToMatch(target, skip, newTypes.toArray(new Class<?>[0]).clone(), pos, false);
}
/**
* Drop the return value of the target handle (if any).
* The returned method handle will have a {@code void} return type.
*
* @param target the method handle to adapt
* @return a possibly adapted method handle
* @throws NullPointerException if {@code target} is null
* @since 16
*/
public static MethodHandle dropReturn(MethodHandle target) {
Objects.requireNonNull(target);
MethodType oldType = target.type();
Class<?> oldReturnType = oldType.returnType();
if (oldReturnType == void.class)
return target;
MethodType newType = oldType.changeReturnType(void.class);
BoundMethodHandle result = target.rebind();
LambdaForm lform = result.editor().filterReturnForm(V_TYPE, true);
result = result.copyWith(newType, lform);
return result;
}
/**
* Adapts a target method handle by pre-processing
* one or more of its arguments, each with its own unary filter function,
* and then calling the target with each pre-processed argument
* replaced by the result of its corresponding filter function.
* <p>
* The pre-processing is performed by one or more method handles,
* specified in the elements of the {@code filters} array.
* The first element of the filter array corresponds to the {@code pos}
* argument of the target, and so on in sequence.
* The filter functions are invoked in left to right order.
* <p>
* Null arguments in the array are treated as identity functions,
* and the corresponding arguments left unchanged.
* (If there are no non-null elements in the array, the original target is returned.)
* Each filter is applied to the corresponding argument of the adapter.
* <p>
* If a filter {@code F} applies to the {@code N}th argument of
* the target, then {@code F} must be a method handle which
* takes exactly one argument. The type of {@code F}'s sole argument
* replaces the corresponding argument type of the target
* in the resulting adapted method handle.
* The return type of {@code F} must be identical to the corresponding
* parameter type of the target.
* <p>
* It is an error if there are elements of {@code filters}
* (null or not)
* which do not correspond to argument positions in the target.
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
MethodHandle upcase = lookup().findVirtual(String.class,
"toUpperCase", methodType(String.class));
assertEquals("xy", (String) cat.invokeExact("x", "y"));
MethodHandle f0 = filterArguments(cat, 0, upcase);
assertEquals("Xy", (String) f0.invokeExact("x", "y")); // Xy
MethodHandle f1 = filterArguments(cat, 1, upcase);
assertEquals("xY", (String) f1.invokeExact("x", "y")); // xY
MethodHandle f2 = filterArguments(cat, 0, upcase, upcase);
assertEquals("XY", (String) f2.invokeExact("x", "y")); // XY
* }
* <p>Here is pseudocode for the resulting adapter. In the code, {@code T}
* denotes the return type of both the {@code target} and resulting adapter.
* {@code P}/{@code p} and {@code B}/{@code b} represent the types and values
* of the parameters and arguments that precede and follow the filter position
* {@code pos}, respectively. {@code A[i]}/{@code a[i]} stand for the types and
* values of the filtered parameters and arguments; they also represent the
* return types of the {@code filter[i]} handles. The latter accept arguments
* {@code v[i]} of type {@code V[i]}, which also appear in the signature of
* the resulting adapter.
* {@snippet lang="java" :
* T target(P... p, A[i]... a[i], B... b);
* A[i] filter[i](V[i]);
* T adapter(P... p, V[i]... v[i], B... b) {
* return target(p..., filter[i](v[i])..., b...);
* }
* }
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
*
* @param target the method handle to invoke after arguments are filtered
* @param pos the position of the first argument to filter
* @param filters method handles to call initially on filtered arguments
* @return method handle which incorporates the specified argument filtering logic
* @throws NullPointerException if the target is null
* or if the {@code filters} array is null
* @throws IllegalArgumentException if a non-null element of {@code filters}
* does not match a corresponding argument type of target as described above,
* or if the {@code pos+filters.length} is greater than {@code target.type().parameterCount()},
* or if the resulting method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
*/
public static MethodHandle filterArguments(MethodHandle target, int pos, MethodHandle... filters) {
// In method types arguments start at index 0, while the LF
// editor have the MH receiver at position 0 - adjust appropriately.
final int MH_RECEIVER_OFFSET = 1;
filterArgumentsCheckArity(target, pos, filters);
MethodHandle adapter = target;
// keep track of currently matched filters, as to optimize repeated filters
int index = 0;
int[] positions = new int[filters.length];
MethodHandle filter = null;
// process filters in reverse order so that the invocation of
// the resulting adapter will invoke the filters in left-to-right order
for (int i = filters.length - 1; i >= 0; --i) {
MethodHandle newFilter = filters[i];
if (newFilter == null) continue; // ignore null elements of filters
// flush changes on update
if (filter != newFilter) {
if (filter != null) {
if (index > 1) {
adapter = filterRepeatedArgument(adapter, filter, Arrays.copyOf(positions, index));
} else {
adapter = filterArgument(adapter, positions[0] - 1, filter);
}
}
filter = newFilter;
index = 0;
}
filterArgumentChecks(target, pos + i, newFilter);
positions[index++] = pos + i + MH_RECEIVER_OFFSET;
}
if (index > 1) {
adapter = filterRepeatedArgument(adapter, filter, Arrays.copyOf(positions, index));
} else if (index == 1) {
adapter = filterArgument(adapter, positions[0] - 1, filter);
}
return adapter;
}
private static MethodHandle filterRepeatedArgument(MethodHandle adapter, MethodHandle filter, int[] positions) {
MethodType targetType = adapter.type();
MethodType filterType = filter.type();
BoundMethodHandle result = adapter.rebind();
Class<?> newParamType = filterType.parameterType(0);
Class<?>[] ptypes = targetType.ptypes().clone();
for (int pos : positions) {
ptypes[pos - 1] = newParamType;
}
MethodType newType = MethodType.methodType(targetType.rtype(), ptypes, true);
LambdaForm lform = result.editor().filterRepeatedArgumentForm(BasicType.basicType(newParamType), positions);
return result.copyWithExtendL(newType, lform, filter);
}
/*non-public*/
static MethodHandle filterArgument(MethodHandle target, int pos, MethodHandle filter) {
filterArgumentChecks(target, pos, filter);
MethodType targetType = target.type();
MethodType filterType = filter.type();
BoundMethodHandle result = target.rebind();
Class<?> newParamType = filterType.parameterType(0);
LambdaForm lform = result.editor().filterArgumentForm(1 + pos, BasicType.basicType(newParamType));
MethodType newType = targetType.changeParameterType(pos, newParamType);
result = result.copyWithExtendL(newType, lform, filter);
return result;
}
private static void filterArgumentsCheckArity(MethodHandle target, int pos, MethodHandle[] filters) {
MethodType targetType = target.type();
int maxPos = targetType.parameterCount();
if (pos + filters.length > maxPos)
throw newIllegalArgumentException("too many filters");
}
private static void filterArgumentChecks(MethodHandle target, int pos, MethodHandle filter) throws RuntimeException {
MethodType targetType = target.type();
MethodType filterType = filter.type();
if (filterType.parameterCount() != 1
|| filterType.returnType() != targetType.parameterType(pos))
throw newIllegalArgumentException("target and filter types do not match", targetType, filterType);
}
/**
* Adapts a target method handle by pre-processing
* a sub-sequence of its arguments with a filter (another method handle).
* The pre-processed arguments are replaced by the result (if any) of the
* filter function.
* The target is then called on the modified (usually shortened) argument list.
* <p>
* If the filter returns a value, the target must accept that value as
* its argument in position {@code pos}, preceded and/or followed by
* any arguments not passed to the filter.
* If the filter returns void, the target must accept all arguments
* not passed to the filter.
* No arguments are reordered, and a result returned from the filter
* replaces (in order) the whole subsequence of arguments originally
* passed to the adapter.
* <p>
* The argument types (if any) of the filter
* replace zero or one argument types of the target, at position {@code pos},
* in the resulting adapted method handle.
* The return type of the filter (if any) must be identical to the
* argument type of the target at position {@code pos}, and that target argument
* is supplied by the return value of the filter.
* <p>
* In all cases, {@code pos} must be greater than or equal to zero, and
* {@code pos} must also be less than or equal to the target's arity.
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle deepToString = publicLookup()
.findStatic(Arrays.class, "deepToString", methodType(String.class, Object[].class));
MethodHandle ts1 = deepToString.asCollector(String[].class, 1);
assertEquals("[strange]", (String) ts1.invokeExact("strange"));
MethodHandle ts2 = deepToString.asCollector(String[].class, 2);
assertEquals("[up, down]", (String) ts2.invokeExact("up", "down"));
MethodHandle ts3 = deepToString.asCollector(String[].class, 3);
MethodHandle ts3_ts2 = collectArguments(ts3, 1, ts2);
assertEquals("[top, [up, down], strange]",
(String) ts3_ts2.invokeExact("top", "up", "down", "strange"));
MethodHandle ts3_ts2_ts1 = collectArguments(ts3_ts2, 3, ts1);
assertEquals("[top, [up, down], [strange]]",
(String) ts3_ts2_ts1.invokeExact("top", "up", "down", "strange"));
MethodHandle ts3_ts2_ts3 = collectArguments(ts3_ts2, 1, ts3);
assertEquals("[top, [[up, down, strange], charm], bottom]",
(String) ts3_ts2_ts3.invokeExact("top", "up", "down", "strange", "charm", "bottom"));
* }
* <p>Here is pseudocode for the resulting adapter. In the code, {@code T}
* represents the return type of the {@code target} and resulting adapter.
* {@code V}/{@code v} stand for the return type and value of the
* {@code filter}, which are also found in the signature and arguments of
* the {@code target}, respectively, unless {@code V} is {@code void}.
* {@code A}/{@code a} and {@code C}/{@code c} represent the parameter types
* and values preceding and following the collection position, {@code pos},
* in the {@code target}'s signature. They also turn up in the resulting
* adapter's signature and arguments, where they surround
* {@code B}/{@code b}, which represent the parameter types and arguments
* to the {@code filter} (if any).
* {@snippet lang="java" :
* T target(A...,V,C...);
* V filter(B...);
* T adapter(A... a,B... b,C... c) {
* V v = filter(b...);
* return target(a...,v,c...);
* }
* // and if the filter has no arguments:
* T target2(A...,V,C...);
* V filter2();
* T adapter2(A... a,C... c) {
* V v = filter2();
* return target2(a...,v,c...);
* }
* // and if the filter has a void return:
* T target3(A...,C...);
* void filter3(B...);
* T adapter3(A... a,B... b,C... c) {
* filter3(b...);
* return target3(a...,c...);
* }
* }
* <p>
* A collection adapter {@code collectArguments(mh, 0, coll)} is equivalent to
* one which first "folds" the affected arguments, and then drops them, in separate
* steps as follows:
* {@snippet lang="java" :
* mh = MethodHandles.dropArguments(mh, 1, coll.type().parameterList()); //step 2
* mh = MethodHandles.foldArguments(mh, coll); //step 1
* }
* If the target method handle consumes no arguments besides than the result
* (if any) of the filter {@code coll}, then {@code collectArguments(mh, 0, coll)}
* is equivalent to {@code filterReturnValue(coll, mh)}.
* If the filter method handle {@code coll} consumes one argument and produces
* a non-void result, then {@code collectArguments(mh, N, coll)}
* is equivalent to {@code filterArguments(mh, N, coll)}.
* Other equivalences are possible but would require argument permutation.
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
*
* @param target the method handle to invoke after filtering the subsequence of arguments
* @param pos the position of the first adapter argument to pass to the filter,
* and/or the target argument which receives the result of the filter
* @param filter method handle to call on the subsequence of arguments
* @return method handle which incorporates the specified argument subsequence filtering logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if the return type of {@code filter}
* is non-void and is not the same as the {@code pos} argument of the target,
* or if {@code pos} is not between 0 and the target's arity, inclusive,
* or if the resulting method handle's type would have
* <a href="MethodHandle.html#maxarity">too many parameters</a>
* @see MethodHandles#foldArguments
* @see MethodHandles#filterArguments
* @see MethodHandles#filterReturnValue
*/
public static MethodHandle collectArguments(MethodHandle target, int pos, MethodHandle filter) {
MethodType newType = collectArgumentsChecks(target, pos, filter);
MethodType collectorType = filter.type();
BoundMethodHandle result = target.rebind();
LambdaForm lform = result.editor().collectArgumentsForm(1 + pos, collectorType.basicType());
return result.copyWithExtendL(newType, lform, filter);
}
private static MethodType collectArgumentsChecks(MethodHandle target, int pos, MethodHandle filter) throws RuntimeException {
MethodType targetType = target.type();
MethodType filterType = filter.type();
Class<?> rtype = filterType.returnType();
Class<?>[] filterArgs = filterType.ptypes();
if (pos < 0 || (rtype == void.class && pos > targetType.parameterCount()) ||
(rtype != void.class && pos >= targetType.parameterCount())) {
throw newIllegalArgumentException("position is out of range for target", target, pos);
}
if (rtype == void.class) {
return targetType.insertParameterTypes(pos, filterArgs);
}
if (rtype != targetType.parameterType(pos)) {
throw newIllegalArgumentException("target and filter types do not match", targetType, filterType);
}
return targetType.dropParameterTypes(pos, pos + 1).insertParameterTypes(pos, filterArgs);
}
/**
* Adapts a target method handle by post-processing
* its return value (if any) with a filter (another method handle).
* The result of the filter is returned from the adapter.
* <p>
* If the target returns a value, the filter must accept that value as
* its only argument.
* If the target returns void, the filter must accept no arguments.
* <p>
* The return type of the filter
* replaces the return type of the target
* in the resulting adapted method handle.
* The argument type of the filter (if any) must be identical to the
* return type of the target.
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
MethodHandle length = lookup().findVirtual(String.class,
"length", methodType(int.class));
System.out.println((String) cat.invokeExact("x", "y")); // xy
MethodHandle f0 = filterReturnValue(cat, length);
System.out.println((int) f0.invokeExact("x", "y")); // 2
* }
* <p>Here is pseudocode for the resulting adapter. In the code,
* {@code T}/{@code t} represent the result type and value of the
* {@code target}; {@code V}, the result type of the {@code filter}; and
* {@code A}/{@code a}, the types and values of the parameters and arguments
* of the {@code target} as well as the resulting adapter.
* {@snippet lang="java" :
* T target(A...);
* V filter(T);
* V adapter(A... a) {
* T t = target(a...);
* return filter(t);
* }
* // and if the target has a void return:
* void target2(A...);
* V filter2();
* V adapter2(A... a) {
* target2(a...);
* return filter2();
* }
* // and if the filter has a void return:
* T target3(A...);
* void filter3(V);
* void adapter3(A... a) {
* T t = target3(a...);
* filter3(t);
* }
* }
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
* @param target the method handle to invoke before filtering the return value
* @param filter method handle to call on the return value
* @return method handle which incorporates the specified return value filtering logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if the argument list of {@code filter}
* does not match the return type of target as described above
*/
public static MethodHandle filterReturnValue(MethodHandle target, MethodHandle filter) {
MethodType targetType = target.type();
MethodType filterType = filter.type();
filterReturnValueChecks(targetType, filterType);
BoundMethodHandle result = target.rebind();
BasicType rtype = BasicType.basicType(filterType.returnType());
LambdaForm lform = result.editor().filterReturnForm(rtype, false);
MethodType newType = targetType.changeReturnType(filterType.returnType());
result = result.copyWithExtendL(newType, lform, filter);
return result;
}
private static void filterReturnValueChecks(MethodType targetType, MethodType filterType) throws RuntimeException {
Class<?> rtype = targetType.returnType();
int filterValues = filterType.parameterCount();
if (filterValues == 0
? (rtype != void.class)
: (rtype != filterType.parameterType(0) || filterValues != 1))
throw newIllegalArgumentException("target and filter types do not match", targetType, filterType);
}
/**
* Filter the return value of a target method handle with a filter function. The filter function is
* applied to the return value of the original handle; if the filter specifies more than one parameters,
* then any remaining parameter is appended to the adapter handle. In other words, the adaptation works
* as follows:
* {@snippet lang="java" :
* T target(A...)
* V filter(B... , T)
* V adapter(A... a, B... b) {
* T t = target(a...);
* return filter(b..., t);
* }
* }
* <p>
* If the filter handle is a unary function, then this method behaves like {@link #filterReturnValue(MethodHandle, MethodHandle)}.
*
* @param target the target method handle
* @param filter the filter method handle
* @return the adapter method handle
*/
/* package */ static MethodHandle collectReturnValue(MethodHandle target, MethodHandle filter) {
MethodType targetType = target.type();
MethodType filterType = filter.type();
BoundMethodHandle result = target.rebind();
LambdaForm lform = result.editor().collectReturnValueForm(filterType.basicType());
MethodType newType = targetType.changeReturnType(filterType.returnType());
if (filterType.parameterCount() > 1) {
for (int i = 0 ; i < filterType.parameterCount() - 1 ; i++) {
newType = newType.appendParameterTypes(filterType.parameterType(i));
}
}
result = result.copyWithExtendL(newType, lform, filter);
return result;
}
/**
* Adapts a target method handle by pre-processing
* some of its arguments, and then calling the target with
* the result of the pre-processing, inserted into the original
* sequence of arguments.
* <p>
* The pre-processing is performed by {@code combiner}, a second method handle.
* Of the arguments passed to the adapter, the first {@code N} arguments
* are copied to the combiner, which is then called.
* (Here, {@code N} is defined as the parameter count of the combiner.)
* After this, control passes to the target, with any result
* from the combiner inserted before the original {@code N} incoming
* arguments.
* <p>
* If the combiner returns a value, the first parameter type of the target
* must be identical with the return type of the combiner, and the next
* {@code N} parameter types of the target must exactly match the parameters
* of the combiner.
* <p>
* If the combiner has a void return, no result will be inserted,
* and the first {@code N} parameter types of the target
* must exactly match the parameters of the combiner.
* <p>
* The resulting adapter is the same type as the target, except that the
* first parameter type is dropped,
* if it corresponds to the result of the combiner.
* <p>
* (Note that {@link #dropArguments(MethodHandle,int,List) dropArguments} can be used to remove any arguments
* that either the combiner or the target does not wish to receive.
* If some of the incoming arguments are destined only for the combiner,
* consider using {@link MethodHandle#asCollector asCollector} instead, since those
* arguments will not need to be live on the stack on entry to the
* target.)
* <p><b>Example:</b>
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle trace = publicLookup().findVirtual(java.io.PrintStream.class,
"println", methodType(void.class, String.class))
.bindTo(System.out);
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
assertEquals("boojum", (String) cat.invokeExact("boo", "jum"));
MethodHandle catTrace = foldArguments(cat, trace);
// also prints "boo":
assertEquals("boojum", (String) catTrace.invokeExact("boo", "jum"));
* }
* <p>Here is pseudocode for the resulting adapter. In the code, {@code T}
* represents the result type of the {@code target} and resulting adapter.
* {@code V}/{@code v} represent the type and value of the parameter and argument
* of {@code target} that precedes the folding position; {@code V} also is
* the result type of the {@code combiner}. {@code A}/{@code a} denote the
* types and values of the {@code N} parameters and arguments at the folding
* position. {@code B}/{@code b} represent the types and values of the
* {@code target} parameters and arguments that follow the folded parameters
* and arguments.
* {@snippet lang="java" :
* // there are N arguments in A...
* T target(V, A[N]..., B...);
* V combiner(A...);
* T adapter(A... a, B... b) {
* V v = combiner(a...);
* return target(v, a..., b...);
* }
* // and if the combiner has a void return:
* T target2(A[N]..., B...);
* void combiner2(A...);
* T adapter2(A... a, B... b) {
* combiner2(a...);
* return target2(a..., b...);
* }
* }
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
* @param target the method handle to invoke after arguments are combined
* @param combiner method handle to call initially on the incoming arguments
* @return method handle which incorporates the specified argument folding logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if {@code combiner}'s return type
* is non-void and not the same as the first argument type of
* the target, or if the initial {@code N} argument types
* of the target
* (skipping one matching the {@code combiner}'s return type)
* are not identical with the argument types of {@code combiner}
*/
public static MethodHandle foldArguments(MethodHandle target, MethodHandle combiner) {
return foldArguments(target, 0, combiner);
}
/**
* Adapts a target method handle by pre-processing some of its arguments, starting at a given position, and then
* calling the target with the result of the pre-processing, inserted into the original sequence of arguments just
* before the folded arguments.
* <p>
* This method is closely related to {@link #foldArguments(MethodHandle, MethodHandle)}, but allows to control the
* position in the parameter list at which folding takes place. The argument controlling this, {@code pos}, is a
* zero-based index. The aforementioned method {@link #foldArguments(MethodHandle, MethodHandle)} assumes position
* 0.
*
* @apiNote Example:
* {@snippet lang="java" :
import static java.lang.invoke.MethodHandles.*;
import static java.lang.invoke.MethodType.*;
...
MethodHandle trace = publicLookup().findVirtual(java.io.PrintStream.class,
"println", methodType(void.class, String.class))
.bindTo(System.out);
MethodHandle cat = lookup().findVirtual(String.class,
"concat", methodType(String.class, String.class));
assertEquals("boojum", (String) cat.invokeExact("boo", "jum"));
MethodHandle catTrace = foldArguments(cat, 1, trace);
// also prints "jum":
assertEquals("boojum", (String) catTrace.invokeExact("boo", "jum"));
* }
* <p>Here is pseudocode for the resulting adapter. In the code, {@code T}
* represents the result type of the {@code target} and resulting adapter.
* {@code V}/{@code v} represent the type and value of the parameter and argument
* of {@code target} that precedes the folding position; {@code V} also is
* the result type of the {@code combiner}. {@code A}/{@code a} denote the
* types and values of the {@code N} parameters and arguments at the folding
* position. {@code Z}/{@code z} and {@code B}/{@code b} represent the types
* and values of the {@code target} parameters and arguments that precede and
* follow the folded parameters and arguments starting at {@code pos},
* respectively.
* {@snippet lang="java" :
* // there are N arguments in A...
* T target(Z..., V, A[N]..., B...);
* V combiner(A...);
* T adapter(Z... z, A... a, B... b) {
* V v = combiner(a...);
* return target(z..., v, a..., b...);
* }
* // and if the combiner has a void return:
* T target2(Z..., A[N]..., B...);
* void combiner2(A...);
* T adapter2(Z... z, A... a, B... b) {
* combiner2(a...);
* return target2(z..., a..., b...);
* }
* }
* <p>
* <em>Note:</em> The resulting adapter is never a {@linkplain MethodHandle#asVarargsCollector
* variable-arity method handle}, even if the original target method handle was.
*
* @param target the method handle to invoke after arguments are combined
* @param pos the position at which to start folding and at which to insert the folding result; if this is {@code
* 0}, the effect is the same as for {@link #foldArguments(MethodHandle, MethodHandle)}.
* @param combiner method handle to call initially on the incoming arguments
* @return method handle which incorporates the specified argument folding logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if either of the following two conditions holds:
* (1) {@code combiner}'s return type is non-{@code void} and not the same as the argument type at position
* {@code pos} of the target signature;
* (2) the {@code N} argument types at position {@code pos} of the target signature (skipping one matching
* the {@code combiner}'s return type) are not identical with the argument types of {@code combiner}.
*
* @see #foldArguments(MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle foldArguments(MethodHandle target, int pos, MethodHandle combiner) {
MethodType targetType = target.type();
MethodType combinerType = combiner.type();
Class<?> rtype = foldArgumentChecks(pos, targetType, combinerType);
BoundMethodHandle result = target.rebind();
boolean dropResult = rtype == void.class;
LambdaForm lform = result.editor().foldArgumentsForm(1 + pos, dropResult, combinerType.basicType());
MethodType newType = targetType;
if (!dropResult) {
newType = newType.dropParameterTypes(pos, pos + 1);
}
result = result.copyWithExtendL(newType, lform, combiner);
return result;
}
private static Class<?> foldArgumentChecks(int foldPos, MethodType targetType, MethodType combinerType) {
int foldArgs = combinerType.parameterCount();
Class<?> rtype = combinerType.returnType();
int foldVals = rtype == void.class ? 0 : 1;
int afterInsertPos = foldPos + foldVals;
boolean ok = (targetType.parameterCount() >= afterInsertPos + foldArgs);
if (ok) {
for (int i = 0; i < foldArgs; i++) {
if (combinerType.parameterType(i) != targetType.parameterType(i + afterInsertPos)) {
ok = false;
break;
}
}
}
if (ok && foldVals != 0 && combinerType.returnType() != targetType.parameterType(foldPos))
ok = false;
if (!ok)
throw misMatchedTypes("target and combiner types", targetType, combinerType);
return rtype;
}
/**
* Adapts a target method handle by pre-processing some of its arguments, then calling the target with the result
* of the pre-processing replacing the argument at the given position.
*
* @param target the method handle to invoke after arguments are combined
* @param position the position at which to start folding and at which to insert the folding result; if this is {@code
* 0}, the effect is the same as for {@link #foldArguments(MethodHandle, MethodHandle)}.
* @param combiner method handle to call initially on the incoming arguments
* @param argPositions indexes of the target to pick arguments sent to the combiner from
* @return method handle which incorporates the specified argument folding logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if either of the following two conditions holds:
* (1) {@code combiner}'s return type is not the same as the argument type at position
* {@code pos} of the target signature;
* (2) the {@code N} argument types at positions {@code argPositions[1...N]} of the target signature are
* not identical with the argument types of {@code combiner}.
*/
/*non-public*/
static MethodHandle filterArgumentsWithCombiner(MethodHandle target, int position, MethodHandle combiner, int ... argPositions) {
return argumentsWithCombiner(true, target, position, combiner, argPositions);
}
/**
* Adapts a target method handle by pre-processing some of its arguments, calling the target with the result of
* the pre-processing inserted into the original sequence of arguments at the given position.
*
* @param target the method handle to invoke after arguments are combined
* @param position the position at which to start folding and at which to insert the folding result; if this is {@code
* 0}, the effect is the same as for {@link #foldArguments(MethodHandle, MethodHandle)}.
* @param combiner method handle to call initially on the incoming arguments
* @param argPositions indexes of the target to pick arguments sent to the combiner from
* @return method handle which incorporates the specified argument folding logic
* @throws NullPointerException if either argument is null
* @throws IllegalArgumentException if either of the following two conditions holds:
* (1) {@code combiner}'s return type is non-{@code void} and not the same as the argument type at position
* {@code pos} of the target signature;
* (2) the {@code N} argument types at positions {@code argPositions[1...N]} of the target signature
* (skipping {@code position} where the {@code combiner}'s return will be folded in) are not identical
* with the argument types of {@code combiner}.
*/
/*non-public*/
static MethodHandle foldArgumentsWithCombiner(MethodHandle target, int position, MethodHandle combiner, int ... argPositions) {
return argumentsWithCombiner(false, target, position, combiner, argPositions);
}
private static MethodHandle argumentsWithCombiner(boolean filter, MethodHandle target, int position, MethodHandle combiner, int ... argPositions) {
MethodType targetType = target.type();
MethodType combinerType = combiner.type();
Class<?> rtype = argumentsWithCombinerChecks(position, filter, targetType, combinerType, argPositions);
BoundMethodHandle result = target.rebind();
MethodType newType = targetType;
LambdaForm lform;
if (filter) {
lform = result.editor().filterArgumentsForm(1 + position, combinerType.basicType(), argPositions);
} else {
boolean dropResult = rtype == void.class;
lform = result.editor().foldArgumentsForm(1 + position, dropResult, combinerType.basicType(), argPositions);
if (!dropResult) {
newType = newType.dropParameterTypes(position, position + 1);
}
}
result = result.copyWithExtendL(newType, lform, combiner);
return result;
}
private static Class<?> argumentsWithCombinerChecks(int position, boolean filter, MethodType targetType, MethodType combinerType, int ... argPos) {
int combinerArgs = combinerType.parameterCount();
if (argPos.length != combinerArgs) {
throw newIllegalArgumentException("combiner and argument map must be equal size", combinerType, argPos.length);
}
Class<?> rtype = combinerType.returnType();
for (int i = 0; i < combinerArgs; i++) {
int arg = argPos[i];
if (arg < 0 || arg > targetType.parameterCount()) {
throw newIllegalArgumentException("arg outside of target parameterRange", targetType, arg);
}
if (combinerType.parameterType(i) != targetType.parameterType(arg)) {
throw newIllegalArgumentException("target argument type at position " + arg
+ " must match combiner argument type at index " + i + ": " + targetType
+ " -> " + combinerType + ", map: " + Arrays.toString(argPos));
}
}
if (filter && combinerType.returnType() != targetType.parameterType(position)) {
throw misMatchedTypes("target and combiner types", targetType, combinerType);
}
return rtype;
}
/**
* Makes a method handle which adapts a target method handle,
* by guarding it with a test, a boolean-valued method handle.
* If the guard fails, a fallback handle is called instead.
* All three method handles must have the same corresponding
* argument and return types, except that the return type
* of the test must be boolean, and the test is allowed
* to have fewer arguments than the other two method handles.
* <p>
* Here is pseudocode for the resulting adapter. In the code, {@code T}
* represents the uniform result type of the three involved handles;
* {@code A}/{@code a}, the types and values of the {@code target}
* parameters and arguments that are consumed by the {@code test}; and
* {@code B}/{@code b}, those types and values of the {@code target}
* parameters and arguments that are not consumed by the {@code test}.
* {@snippet lang="java" :
* boolean test(A...);
* T target(A...,B...);
* T fallback(A...,B...);
* T adapter(A... a,B... b) {
* if (test(a...))
* return target(a..., b...);
* else
* return fallback(a..., b...);
* }
* }
* Note that the test arguments ({@code a...} in the pseudocode) cannot
* be modified by execution of the test, and so are passed unchanged
* from the caller to the target or fallback as appropriate.
* @param test method handle used for test, must return boolean
* @param target method handle to call if test passes
* @param fallback method handle to call if test fails
* @return method handle which incorporates the specified if/then/else logic
* @throws NullPointerException if any argument is null
* @throws IllegalArgumentException if {@code test} does not return boolean,
* or if all three method types do not match (with the return
* type of {@code test} changed to match that of the target).
*/
public static MethodHandle guardWithTest(MethodHandle test,
MethodHandle target,
MethodHandle fallback) {
MethodType gtype = test.type();
MethodType ttype = target.type();
MethodType ftype = fallback.type();
if (!ttype.equals(ftype))
throw misMatchedTypes("target and fallback types", ttype, ftype);
if (gtype.returnType() != boolean.class)
throw newIllegalArgumentException("guard type is not a predicate "+gtype);
test = dropArgumentsToMatch(test, 0, ttype.ptypes(), 0, true);
if (test == null) {
throw misMatchedTypes("target and test types", ttype, gtype);
}
return MethodHandleImpl.makeGuardWithTest(test, target, fallback);
}
static <T> RuntimeException misMatchedTypes(String what, T t1, T t2) {
return newIllegalArgumentException(what + " must match: " + t1 + " != " + t2);
}
/**
* Makes a method handle which adapts a target method handle,
* by running it inside an exception handler.
* If the target returns normally, the adapter returns that value.
* If an exception matching the specified type is thrown, the fallback
* handle is called instead on the exception, plus the original arguments.
* <p>
* The target and handler must have the same corresponding
* argument and return types, except that handler may omit trailing arguments
* (similarly to the predicate in {@link #guardWithTest guardWithTest}).
* Also, the handler must have an extra leading parameter of {@code exType} or a supertype.
* <p>
* Here is pseudocode for the resulting adapter. In the code, {@code T}
* represents the return type of the {@code target} and {@code handler},
* and correspondingly that of the resulting adapter; {@code A}/{@code a},
* the types and values of arguments to the resulting handle consumed by
* {@code handler}; and {@code B}/{@code b}, those of arguments to the
* resulting handle discarded by {@code handler}.
* {@snippet lang="java" :
* T target(A..., B...);
* T handler(ExType, A...);
* T adapter(A... a, B... b) {
* try {
* return target(a..., b...);
* } catch (ExType ex) {
* return handler(ex, a...);
* }
* }
* }
* Note that the saved arguments ({@code a...} in the pseudocode) cannot
* be modified by execution of the target, and so are passed unchanged
* from the caller to the handler, if the handler is invoked.
* <p>
* The target and handler must return the same type, even if the handler
* always throws. (This might happen, for instance, because the handler
* is simulating a {@code finally} clause).
* To create such a throwing handler, compose the handler creation logic
* with {@link #throwException throwException},
* in order to create a method handle of the correct return type.
* @param target method handle to call
* @param exType the type of exception which the handler will catch
* @param handler method handle to call if a matching exception is thrown
* @return method handle which incorporates the specified try/catch logic
* @throws NullPointerException if any argument is null
* @throws IllegalArgumentException if {@code handler} does not accept
* the given exception type, or if the method handle types do
* not match in their return types and their
* corresponding parameters
* @see MethodHandles#tryFinally(MethodHandle, MethodHandle)
*/
public static MethodHandle catchException(MethodHandle target,
Class<? extends Throwable> exType,
MethodHandle handler) {
MethodType ttype = target.type();
MethodType htype = handler.type();
if (!Throwable.class.isAssignableFrom(exType))
throw new ClassCastException(exType.getName());
if (htype.parameterCount() < 1 ||
!htype.parameterType(0).isAssignableFrom(exType))
throw newIllegalArgumentException("handler does not accept exception type "+exType);
if (htype.returnType() != ttype.returnType())
throw misMatchedTypes("target and handler return types", ttype, htype);
handler = dropArgumentsToMatch(handler, 1, ttype.ptypes(), 0, true);
if (handler == null) {
throw misMatchedTypes("target and handler types", ttype, htype);
}
return MethodHandleImpl.makeGuardWithCatch(target, exType, handler);
}
/**
* Produces a method handle which will throw exceptions of the given {@code exType}.
* The method handle will accept a single argument of {@code exType},
* and immediately throw it as an exception.
* The method type will nominally specify a return of {@code returnType}.
* The return type may be anything convenient: It doesn't matter to the
* method handle's behavior, since it will never return normally.
* @param returnType the return type of the desired method handle
* @param exType the parameter type of the desired method handle
* @return method handle which can throw the given exceptions
* @throws NullPointerException if either argument is null
*/
public static MethodHandle throwException(Class<?> returnType, Class<? extends Throwable> exType) {
if (!Throwable.class.isAssignableFrom(exType))
throw new ClassCastException(exType.getName());
return MethodHandleImpl.throwException(methodType(returnType, exType));
}
/**
* Constructs a method handle representing a loop with several loop variables that are updated and checked upon each
* iteration. Upon termination of the loop due to one of the predicates, a corresponding finalizer is run and
* delivers the loop's result, which is the return value of the resulting handle.
* <p>
* Intuitively, every loop is formed by one or more "clauses", each specifying a local <em>iteration variable</em> and/or a loop
* exit. Each iteration of the loop executes each clause in order. A clause can optionally update its iteration
* variable; it can also optionally perform a test and conditional loop exit. In order to express this logic in
* terms of method handles, each clause will specify up to four independent actions:<ul>
* <li><em>init:</em> Before the loop executes, the initialization of an iteration variable {@code v} of type {@code V}.
* <li><em>step:</em> When a clause executes, an update step for the iteration variable {@code v}.
* <li><em>pred:</em> When a clause executes, a predicate execution to test for loop exit.
* <li><em>fini:</em> If a clause causes a loop exit, a finalizer execution to compute the loop's return value.
* </ul>
* The full sequence of all iteration variable types, in clause order, will be notated as {@code (V...)}.
* The values themselves will be {@code (v...)}. When we speak of "parameter lists", we will usually
* be referring to types, but in some contexts (describing execution) the lists will be of actual values.
* <p>
* Some of these clause parts may be omitted according to certain rules, and useful default behavior is provided in
* this case. See below for a detailed description.
* <p>
* <em>Parameters optional everywhere:</em>
* Each clause function is allowed but not required to accept a parameter for each iteration variable {@code v}.
* As an exception, the init functions cannot take any {@code v} parameters,
* because those values are not yet computed when the init functions are executed.
* Any clause function may neglect to take any trailing subsequence of parameters it is entitled to take.
* In fact, any clause function may take no arguments at all.
* <p>
* <em>Loop parameters:</em>
* A clause function may take all the iteration variable values it is entitled to, in which case
* it may also take more trailing parameters. Such extra values are called <em>loop parameters</em>,
* with their types and values notated as {@code (A...)} and {@code (a...)}.
* These become the parameters of the resulting loop handle, to be supplied whenever the loop is executed.
* (Since init functions do not accept iteration variables {@code v}, any parameter to an
* init function is automatically a loop parameter {@code a}.)
* As with iteration variables, clause functions are allowed but not required to accept loop parameters.
* These loop parameters act as loop-invariant values visible across the whole loop.
* <p>
* <em>Parameters visible everywhere:</em>
* Each non-init clause function is permitted to observe the entire loop state, because it can be passed the full
* list {@code (v... a...)} of current iteration variable values and incoming loop parameters.
* The init functions can observe initial pre-loop state, in the form {@code (a...)}.
* Most clause functions will not need all of this information, but they will be formally connected to it
* as if by {@link #dropArguments}.
* <a id="astar"></a>
* More specifically, we shall use the notation {@code (V*)} to express an arbitrary prefix of a full
* sequence {@code (V...)} (and likewise for {@code (v*)}, {@code (A*)}, {@code (a*)}).
* In that notation, the general form of an init function parameter list
* is {@code (A*)}, and the general form of a non-init function parameter list is {@code (V*)} or {@code (V... A*)}.
* <p>
* <em>Checking clause structure:</em>
* Given a set of clauses, there is a number of checks and adjustments performed to connect all the parts of the
* loop. They are spelled out in detail in the steps below. In these steps, every occurrence of the word "must"
* corresponds to a place where {@link IllegalArgumentException} will be thrown if the required constraint is not
* met by the inputs to the loop combinator.
* <p>
* <em>Effectively identical sequences:</em>
* <a id="effid"></a>
* A parameter list {@code A} is defined to be <em>effectively identical</em> to another parameter list {@code B}
* if {@code A} and {@code B} are identical, or if {@code A} is shorter and is identical with a proper prefix of {@code B}.
* When speaking of an unordered set of parameter lists, we say they the set is "effectively identical"
* as a whole if the set contains a longest list, and all members of the set are effectively identical to
* that longest list.
* For example, any set of type sequences of the form {@code (V*)} is effectively identical,
* and the same is true if more sequences of the form {@code (V... A*)} are added.
* <p>
* <em>Step 0: Determine clause structure.</em><ol type="a">
* <li>The clause array (of type {@code MethodHandle[][]}) must be non-{@code null} and contain at least one element.
* <li>The clause array may not contain {@code null}s or sub-arrays longer than four elements.
* <li>Clauses shorter than four elements are treated as if they were padded by {@code null} elements to length
* four. Padding takes place by appending elements to the array.
* <li>Clauses with all {@code null}s are disregarded.
* <li>Each clause is treated as a four-tuple of functions, called "init", "step", "pred", and "fini".
* </ol>
* <p>
* <em>Step 1A: Determine iteration variable types {@code (V...)}.</em><ol type="a">
* <li>The iteration variable type for each clause is determined using the clause's init and step return types.
* <li>If both functions are omitted, there is no iteration variable for the corresponding clause ({@code void} is
* used as the type to indicate that). If one of them is omitted, the other's return type defines the clause's
* iteration variable type. If both are given, the common return type (they must be identical) defines the clause's
* iteration variable type.
* <li>Form the list of return types (in clause order), omitting all occurrences of {@code void}.
* <li>This list of types is called the "iteration variable types" ({@code (V...)}).
* </ol>
* <p>
* <em>Step 1B: Determine loop parameters {@code (A...)}.</em><ul>
* <li>Examine and collect init function parameter lists (which are of the form {@code (A*)}).
* <li>Examine and collect the suffixes of the step, pred, and fini parameter lists, after removing the iteration variable types.
* (They must have the form {@code (V... A*)}; collect the {@code (A*)} parts only.)
* <li>Do not collect suffixes from step, pred, and fini parameter lists that do not begin with all the iteration variable types.
* (These types will be checked in step 2, along with all the clause function types.)
* <li>Omitted clause functions are ignored. (Equivalently, they are deemed to have empty parameter lists.)
* <li>All of the collected parameter lists must be effectively identical.
* <li>The longest parameter list (which is necessarily unique) is called the "external parameter list" ({@code (A...)}).
* <li>If there is no such parameter list, the external parameter list is taken to be the empty sequence.
* <li>The combined list consisting of iteration variable types followed by the external parameter types is called
* the "internal parameter list".
* </ul>
* <p>
* <em>Step 1C: Determine loop return type.</em><ol type="a">
* <li>Examine fini function return types, disregarding omitted fini functions.
* <li>If there are no fini functions, the loop return type is {@code void}.
* <li>Otherwise, the common return type {@code R} of the fini functions (their return types must be identical) defines the loop return
* type.
* </ol>
* <p>
* <em>Step 1D: Check other types.</em><ol type="a">
* <li>There must be at least one non-omitted pred function.
* <li>Every non-omitted pred function must have a {@code boolean} return type.
* </ol>
* <p>
* <em>Step 2: Determine parameter lists.</em><ol type="a">
* <li>The parameter list for the resulting loop handle will be the external parameter list {@code (A...)}.
* <li>The parameter list for init functions will be adjusted to the external parameter list.
* (Note that their parameter lists are already effectively identical to this list.)
* <li>The parameter list for every non-omitted, non-init (step, pred, and fini) function must be
* effectively identical to the internal parameter list {@code (V... A...)}.
* </ol>
* <p>
* <em>Step 3: Fill in omitted functions.</em><ol type="a">
* <li>If an init function is omitted, use a {@linkplain #empty default value} for the clause's iteration variable
* type.
* <li>If a step function is omitted, use an {@linkplain #identity identity function} of the clause's iteration
* variable type; insert dropped argument parameters before the identity function parameter for the non-{@code void}
* iteration variables of preceding clauses. (This will turn the loop variable into a local loop invariant.)
* <li>If a pred function is omitted, use a constant {@code true} function. (This will keep the loop going, as far
* as this clause is concerned. Note that in such cases the corresponding fini function is unreachable.)
* <li>If a fini function is omitted, use a {@linkplain #empty default value} for the
* loop return type.
* </ol>
* <p>
* <em>Step 4: Fill in missing parameter types.</em><ol type="a">
* <li>At this point, every init function parameter list is effectively identical to the external parameter list {@code (A...)},
* but some lists may be shorter. For every init function with a short parameter list, pad out the end of the list.
* <li>At this point, every non-init function parameter list is effectively identical to the internal parameter
* list {@code (V... A...)}, but some lists may be shorter. For every non-init function with a short parameter list,
* pad out the end of the list.
* <li>Argument lists are padded out by {@linkplain #dropArgumentsToMatch(MethodHandle, int, List, int) dropping unused trailing arguments}.
* </ol>
* <p>
* <em>Final observations.</em><ol type="a">
* <li>After these steps, all clauses have been adjusted by supplying omitted functions and arguments.
* <li>All init functions have a common parameter type list {@code (A...)}, which the final loop handle will also have.
* <li>All fini functions have a common return type {@code R}, which the final loop handle will also have.
* <li>All non-init functions have a common parameter type list {@code (V... A...)}, of
* (non-{@code void}) iteration variables {@code V} followed by loop parameters.
* <li>Each pair of init and step functions agrees in their return type {@code V}.
* <li>Each non-init function will be able to observe the current values {@code (v...)} of all iteration variables.
* <li>Every function will be able to observe the incoming values {@code (a...)} of all loop parameters.
* </ol>
* <p>
* <em>Example.</em> As a consequence of step 1A above, the {@code loop} combinator has the following property:
* <ul>
* <li>Given {@code N} clauses {@code Cn = {null, Sn, Pn}} with {@code n = 1..N}.
* <li>Suppose predicate handles {@code Pn} are either {@code null} or have no parameters.
* (Only one {@code Pn} has to be non-{@code null}.)
* <li>Suppose step handles {@code Sn} have signatures {@code (B1..BX)Rn}, for some constant {@code X>=N}.
* <li>Suppose {@code Q} is the count of non-void types {@code Rn}, and {@code (V1...VQ)} is the sequence of those types.
* <li>It must be that {@code Vn == Bn} for {@code n = 1..min(X,Q)}.
* <li>The parameter types {@code Vn} will be interpreted as loop-local state elements {@code (V...)}.
* <li>Any remaining types {@code BQ+1..BX} (if {@code Q<X}) will determine
* the resulting loop handle's parameter types {@code (A...)}.
* </ul>
* In this example, the loop handle parameters {@code (A...)} were derived from the step functions,
* which is natural if most of the loop computation happens in the steps. For some loops,
* the burden of computation might be heaviest in the pred functions, and so the pred functions
* might need to accept the loop parameter values. For loops with complex exit logic, the fini
* functions might need to accept loop parameters, and likewise for loops with complex entry logic,
* where the init functions will need the extra parameters. For such reasons, the rules for
* determining these parameters are as symmetric as possible, across all clause parts.
* In general, the loop parameters function as common invariant values across the whole
* loop, while the iteration variables function as common variant values, or (if there is
* no step function) as internal loop invariant temporaries.
* <p>
* <em>Loop execution.</em><ol type="a">
* <li>When the loop is called, the loop input values are saved in locals, to be passed to
* every clause function. These locals are loop invariant.
* <li>Each init function is executed in clause order (passing the external arguments {@code (a...)})
* and the non-{@code void} values are saved (as the iteration variables {@code (v...)}) into locals.
* These locals will be loop varying (unless their steps behave as identity functions, as noted above).
* <li>All function executions (except init functions) will be passed the internal parameter list, consisting of
* the non-{@code void} iteration values {@code (v...)} (in clause order) and then the loop inputs {@code (a...)}
* (in argument order).
* <li>The step and pred functions are then executed, in clause order (step before pred), until a pred function
* returns {@code false}.
* <li>The non-{@code void} result from a step function call is used to update the corresponding value in the
* sequence {@code (v...)} of loop variables.
* The updated value is immediately visible to all subsequent function calls.
* <li>If a pred function returns {@code false}, the corresponding fini function is called, and the resulting value
* (of type {@code R}) is returned from the loop as a whole.
* <li>If all the pred functions always return true, no fini function is ever invoked, and the loop cannot exit
* except by throwing an exception.
* </ol>
* <p>
* <em>Usage tips.</em>
* <ul>
* <li>Although each step function will receive the current values of <em>all</em> the loop variables,
* sometimes a step function only needs to observe the current value of its own variable.
* In that case, the step function may need to explicitly {@linkplain #dropArguments drop all preceding loop variables}.
* This will require mentioning their types, in an expression like {@code dropArguments(step, 0, V0.class, ...)}.
* <li>Loop variables are not required to vary; they can be loop invariant. A clause can create
* a loop invariant by a suitable init function with no step, pred, or fini function. This may be
* useful to "wire" an incoming loop argument into the step or pred function of an adjacent loop variable.
* <li>If some of the clause functions are virtual methods on an instance, the instance
* itself can be conveniently placed in an initial invariant loop "variable", using an initial clause
* like {@code new MethodHandle[]{identity(ObjType.class)}}. In that case, the instance reference
* will be the first iteration variable value, and it will be easy to use virtual
* methods as clause parts, since all of them will take a leading instance reference matching that value.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. As above, {@code V} and {@code v} represent the types
* and values of loop variables; {@code A} and {@code a} represent arguments passed to the whole loop;
* and {@code R} is the common result type of all finalizers as well as of the resulting loop.
* {@snippet lang="java" :
* V... init...(A...);
* boolean pred...(V..., A...);
* V... step...(V..., A...);
* R fini...(V..., A...);
* R loop(A... a) {
* V... v... = init...(a...);
* for (;;) {
* for ((v, p, s, f) in (v..., pred..., step..., fini...)) {
* v = s(v..., a...);
* if (!p(v..., a...)) {
* return f(v..., a...);
* }
* }
* }
* }
* }
* Note that the parameter type lists {@code (V...)} and {@code (A...)} have been expanded
* to their full length, even though individual clause functions may neglect to take them all.
* As noted above, missing parameters are filled in as if by {@link #dropArgumentsToMatch(MethodHandle, int, List, int)}.
*
* @apiNote Example:
* {@snippet lang="java" :
* // iterative implementation of the factorial function as a loop handle
* static int one(int k) { return 1; }
* static int inc(int i, int acc, int k) { return i + 1; }
* static int mult(int i, int acc, int k) { return i * acc; }
* static boolean pred(int i, int acc, int k) { return i < k; }
* static int fin(int i, int acc, int k) { return acc; }
* // assume MH_one, MH_inc, MH_mult, MH_pred, and MH_fin are handles to the above methods
* // null initializer for counter, should initialize to 0
* MethodHandle[] counterClause = new MethodHandle[]{null, MH_inc};
* MethodHandle[] accumulatorClause = new MethodHandle[]{MH_one, MH_mult, MH_pred, MH_fin};
* MethodHandle loop = MethodHandles.loop(counterClause, accumulatorClause);
* assertEquals(120, loop.invoke(5));
* }
* The same example, dropping arguments and using combinators:
* {@snippet lang="java" :
* // simplified implementation of the factorial function as a loop handle
* static int inc(int i) { return i + 1; } // drop acc, k
* static int mult(int i, int acc) { return i * acc; } //drop k
* static boolean cmp(int i, int k) { return i < k; }
* // assume MH_inc, MH_mult, and MH_cmp are handles to the above methods
* // null initializer for counter, should initialize to 0
* MethodHandle MH_one = MethodHandles.constant(int.class, 1);
* MethodHandle MH_pred = MethodHandles.dropArguments(MH_cmp, 1, int.class); // drop acc
* MethodHandle MH_fin = MethodHandles.dropArguments(MethodHandles.identity(int.class), 0, int.class); // drop i
* MethodHandle[] counterClause = new MethodHandle[]{null, MH_inc};
* MethodHandle[] accumulatorClause = new MethodHandle[]{MH_one, MH_mult, MH_pred, MH_fin};
* MethodHandle loop = MethodHandles.loop(counterClause, accumulatorClause);
* assertEquals(720, loop.invoke(6));
* }
* A similar example, using a helper object to hold a loop parameter:
* {@snippet lang="java" :
* // instance-based implementation of the factorial function as a loop handle
* static class FacLoop {
* final int k;
* FacLoop(int k) { this.k = k; }
* int inc(int i) { return i + 1; }
* int mult(int i, int acc) { return i * acc; }
* boolean pred(int i) { return i < k; }
* int fin(int i, int acc) { return acc; }
* }
* // assume MH_FacLoop is a handle to the constructor
* // assume MH_inc, MH_mult, MH_pred, and MH_fin are handles to the above methods
* // null initializer for counter, should initialize to 0
* MethodHandle MH_one = MethodHandles.constant(int.class, 1);
* MethodHandle[] instanceClause = new MethodHandle[]{MH_FacLoop};
* MethodHandle[] counterClause = new MethodHandle[]{null, MH_inc};
* MethodHandle[] accumulatorClause = new MethodHandle[]{MH_one, MH_mult, MH_pred, MH_fin};
* MethodHandle loop = MethodHandles.loop(instanceClause, counterClause, accumulatorClause);
* assertEquals(5040, loop.invoke(7));
* }
*
* @param clauses an array of arrays (4-tuples) of {@link MethodHandle}s adhering to the rules described above.
*
* @return a method handle embodying the looping behavior as defined by the arguments.
*
* @throws IllegalArgumentException in case any of the constraints described above is violated.
*
* @see MethodHandles#whileLoop(MethodHandle, MethodHandle, MethodHandle)
* @see MethodHandles#doWhileLoop(MethodHandle, MethodHandle, MethodHandle)
* @see MethodHandles#countedLoop(MethodHandle, MethodHandle, MethodHandle)
* @see MethodHandles#iteratedLoop(MethodHandle, MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle loop(MethodHandle[]... clauses) {
// Step 0: determine clause structure.
loopChecks0(clauses);
List<MethodHandle> init = new ArrayList<>();
List<MethodHandle> step = new ArrayList<>();
List<MethodHandle> pred = new ArrayList<>();
List<MethodHandle> fini = new ArrayList<>();
Stream.of(clauses).filter(c -> Stream.of(c).anyMatch(Objects::nonNull)).forEach(clause -> {
init.add(clause[0]); // all clauses have at least length 1
step.add(clause.length <= 1 ? null : clause[1]);
pred.add(clause.length <= 2 ? null : clause[2]);
fini.add(clause.length <= 3 ? null : clause[3]);
});
assert Stream.of(init, step, pred, fini).map(List::size).distinct().count() == 1;
final int nclauses = init.size();
// Step 1A: determine iteration variables (V...).
final List<Class<?>> iterationVariableTypes = new ArrayList<>();
for (int i = 0; i < nclauses; ++i) {
MethodHandle in = init.get(i);
MethodHandle st = step.get(i);
if (in == null && st == null) {
iterationVariableTypes.add(void.class);
} else if (in != null && st != null) {
loopChecks1a(i, in, st);
iterationVariableTypes.add(in.type().returnType());
} else {
iterationVariableTypes.add(in == null ? st.type().returnType() : in.type().returnType());
}
}
final List<Class<?>> commonPrefix = iterationVariableTypes.stream().filter(t -> t != void.class).toList();
// Step 1B: determine loop parameters (A...).
final List<Class<?>> commonSuffix = buildCommonSuffix(init, step, pred, fini, commonPrefix.size());
loopChecks1b(init, commonSuffix);
// Step 1C: determine loop return type.
// Step 1D: check other types.
// local variable required here; see JDK-8223553
Stream<Class<?>> cstream = fini.stream().filter(Objects::nonNull).map(MethodHandle::type)
.map(MethodType::returnType);
final Class<?> loopReturnType = cstream.findFirst().orElse(void.class);
loopChecks1cd(pred, fini, loopReturnType);
// Step 2: determine parameter lists.
final List<Class<?>> commonParameterSequence = new ArrayList<>(commonPrefix);
commonParameterSequence.addAll(commonSuffix);
loopChecks2(step, pred, fini, commonParameterSequence);
// Step 3: fill in omitted functions.
for (int i = 0; i < nclauses; ++i) {
Class<?> t = iterationVariableTypes.get(i);
if (init.get(i) == null) {
init.set(i, empty(methodType(t, commonSuffix)));
}
if (step.get(i) == null) {
step.set(i, dropArgumentsToMatch(identityOrVoid(t), 0, commonParameterSequence, i));
}
if (pred.get(i) == null) {
pred.set(i, dropArguments(constant(boolean.class, true), 0, commonParameterSequence));
}
if (fini.get(i) == null) {
fini.set(i, empty(methodType(t, commonParameterSequence)));
}
}
// Step 4: fill in missing parameter types.
// Also convert all handles to fixed-arity handles.
List<MethodHandle> finit = fixArities(fillParameterTypes(init, commonSuffix));
List<MethodHandle> fstep = fixArities(fillParameterTypes(step, commonParameterSequence));
List<MethodHandle> fpred = fixArities(fillParameterTypes(pred, commonParameterSequence));
List<MethodHandle> ffini = fixArities(fillParameterTypes(fini, commonParameterSequence));
assert finit.stream().map(MethodHandle::type).map(MethodType::parameterList).
allMatch(pl -> pl.equals(commonSuffix));
assert Stream.of(fstep, fpred, ffini).flatMap(List::stream).map(MethodHandle::type).map(MethodType::parameterList).
allMatch(pl -> pl.equals(commonParameterSequence));
return MethodHandleImpl.makeLoop(loopReturnType, commonSuffix, finit, fstep, fpred, ffini);
}
private static void loopChecks0(MethodHandle[][] clauses) {
if (clauses == null || clauses.length == 0) {
throw newIllegalArgumentException("null or no clauses passed");
}
if (Stream.of(clauses).anyMatch(Objects::isNull)) {
throw newIllegalArgumentException("null clauses are not allowed");
}
if (Stream.of(clauses).anyMatch(c -> c.length > 4)) {
throw newIllegalArgumentException("All loop clauses must be represented as MethodHandle arrays with at most 4 elements.");
}
}
private static void loopChecks1a(int i, MethodHandle in, MethodHandle st) {
if (in.type().returnType() != st.type().returnType()) {
throw misMatchedTypes("clause " + i + ": init and step return types", in.type().returnType(),
st.type().returnType());
}
}
private static List<Class<?>> longestParameterList(Stream<MethodHandle> mhs, int skipSize) {
return mhs.filter(Objects::nonNull)
// take only those that can contribute to a common suffix because they are longer than the prefix
.map(MethodHandle::type)
.filter(t -> t.parameterCount() > skipSize)
.max(Comparator.comparingInt(MethodType::parameterCount))
.map(methodType -> List.of(Arrays.copyOfRange(methodType.ptypes(), skipSize, methodType.parameterCount())))
.orElse(List.of());
}
private static List<Class<?>> buildCommonSuffix(List<MethodHandle> init, List<MethodHandle> step, List<MethodHandle> pred, List<MethodHandle> fini, int cpSize) {
final List<Class<?>> longest1 = longestParameterList(Stream.of(step, pred, fini).flatMap(List::stream), cpSize);
final List<Class<?>> longest2 = longestParameterList(init.stream(), 0);
return longest1.size() >= longest2.size() ? longest1 : longest2;
}
private static void loopChecks1b(List<MethodHandle> init, List<Class<?>> commonSuffix) {
if (init.stream().filter(Objects::nonNull).map(MethodHandle::type).
anyMatch(t -> !t.effectivelyIdenticalParameters(0, commonSuffix))) {
throw newIllegalArgumentException("found non-effectively identical init parameter type lists: " + init +
" (common suffix: " + commonSuffix + ")");
}
}
private static void loopChecks1cd(List<MethodHandle> pred, List<MethodHandle> fini, Class<?> loopReturnType) {
if (fini.stream().filter(Objects::nonNull).map(MethodHandle::type).map(MethodType::returnType).
anyMatch(t -> t != loopReturnType)) {
throw newIllegalArgumentException("found non-identical finalizer return types: " + fini + " (return type: " +
loopReturnType + ")");
}
if (pred.stream().noneMatch(Objects::nonNull)) {
throw newIllegalArgumentException("no predicate found", pred);
}
if (pred.stream().filter(Objects::nonNull).map(MethodHandle::type).map(MethodType::returnType).
anyMatch(t -> t != boolean.class)) {
throw newIllegalArgumentException("predicates must have boolean return type", pred);
}
}
private static void loopChecks2(List<MethodHandle> step, List<MethodHandle> pred, List<MethodHandle> fini, List<Class<?>> commonParameterSequence) {
if (Stream.of(step, pred, fini).flatMap(List::stream).filter(Objects::nonNull).map(MethodHandle::type).
anyMatch(t -> !t.effectivelyIdenticalParameters(0, commonParameterSequence))) {
throw newIllegalArgumentException("found non-effectively identical parameter type lists:\nstep: " + step +
"\npred: " + pred + "\nfini: " + fini + " (common parameter sequence: " + commonParameterSequence + ")");
}
}
private static List<MethodHandle> fillParameterTypes(List<MethodHandle> hs, final List<Class<?>> targetParams) {
return hs.stream().map(h -> {
int pc = h.type().parameterCount();
int tpsize = targetParams.size();
return pc < tpsize ? dropArguments(h, pc, targetParams.subList(pc, tpsize)) : h;
}).toList();
}
private static List<MethodHandle> fixArities(List<MethodHandle> hs) {
return hs.stream().map(MethodHandle::asFixedArity).toList();
}
/**
* Constructs a {@code while} loop from an initializer, a body, and a predicate.
* This is a convenience wrapper for the {@linkplain #loop(MethodHandle[][]) generic loop combinator}.
* <p>
* The {@code pred} handle describes the loop condition; and {@code body}, its body. The loop resulting from this
* method will, in each iteration, first evaluate the predicate and then execute its body (if the predicate
* evaluates to {@code true}).
* The loop will terminate once the predicate evaluates to {@code false} (the body will not be executed in this case).
* <p>
* The {@code init} handle describes the initial value of an additional optional loop-local variable.
* In each iteration, this loop-local variable, if present, will be passed to the {@code body}
* and updated with the value returned from its invocation. The result of loop execution will be
* the final value of the additional loop-local variable (if present).
* <p>
* The following rules hold for these argument handles:<ul>
* <li>The {@code body} handle must not be {@code null}; its type must be of the form
* {@code (V A...)V}, where {@code V} is non-{@code void}, or else {@code (A...)void}.
* (In the {@code void} case, we assign the type {@code void} to the name {@code V},
* and we will write {@code (V A...)V} with the understanding that a {@code void} type {@code V}
* is quietly dropped from the parameter list, leaving {@code (A...)V}.)
* <li>The parameter list {@code (V A...)} of the body is called the <em>internal parameter list</em>.
* It will constrain the parameter lists of the other loop parts.
* <li>If the iteration variable type {@code V} is dropped from the internal parameter list, the resulting shorter
* list {@code (A...)} is called the <em>external parameter list</em>.
* <li>The body return type {@code V}, if non-{@code void}, determines the type of an
* additional state variable of the loop.
* The body must both accept and return a value of this type {@code V}.
* <li>If {@code init} is non-{@code null}, it must have return type {@code V}.
* Its parameter list (of some <a href="MethodHandles.html#astar">form {@code (A*)}</a>) must be
* <a href="MethodHandles.html#effid">effectively identical</a>
* to the external parameter list {@code (A...)}.
* <li>If {@code init} is {@code null}, the loop variable will be initialized to its
* {@linkplain #empty default value}.
* <li>The {@code pred} handle must not be {@code null}. It must have {@code boolean} as its return type.
* Its parameter list (either empty or of the form {@code (V A*)}) must be
* effectively identical to the internal parameter list.
* </ul>
* <p>
* The resulting loop handle's result type and parameter signature are determined as follows:<ul>
* <li>The loop handle's result type is the result type {@code V} of the body.
* <li>The loop handle's parameter types are the types {@code (A...)},
* from the external parameter list.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. In the code, {@code V}/{@code v} represent the type / value of
* the sole loop variable as well as the result type of the loop; and {@code A}/{@code a}, that of the argument
* passed to the loop.
* {@snippet lang="java" :
* V init(A...);
* boolean pred(V, A...);
* V body(V, A...);
* V whileLoop(A... a...) {
* V v = init(a...);
* while (pred(v, a...)) {
* v = body(v, a...);
* }
* return v;
* }
* }
*
* @apiNote Example:
* {@snippet lang="java" :
* // implement the zip function for lists as a loop handle
* static List<String> initZip(Iterator<String> a, Iterator<String> b) { return new ArrayList<>(); }
* static boolean zipPred(List<String> zip, Iterator<String> a, Iterator<String> b) { return a.hasNext() && b.hasNext(); }
* static List<String> zipStep(List<String> zip, Iterator<String> a, Iterator<String> b) {
* zip.add(a.next());
* zip.add(b.next());
* return zip;
* }
* // assume MH_initZip, MH_zipPred, and MH_zipStep are handles to the above methods
* MethodHandle loop = MethodHandles.whileLoop(MH_initZip, MH_zipPred, MH_zipStep);
* List<String> a = Arrays.asList("a", "b", "c", "d");
* List<String> b = Arrays.asList("e", "f", "g", "h");
* List<String> zipped = Arrays.asList("a", "e", "b", "f", "c", "g", "d", "h");
* assertEquals(zipped, (List<String>) loop.invoke(a.iterator(), b.iterator()));
* }
*
*
* @apiNote The implementation of this method can be expressed as follows:
* {@snippet lang="java" :
* MethodHandle whileLoop(MethodHandle init, MethodHandle pred, MethodHandle body) {
* MethodHandle fini = (body.type().returnType() == void.class
* ? null : identity(body.type().returnType()));
* MethodHandle[]
* checkExit = { null, null, pred, fini },
* varBody = { init, body };
* return loop(checkExit, varBody);
* }
* }
*
* @param init optional initializer, providing the initial value of the loop variable.
* May be {@code null}, implying a default initial value. See above for other constraints.
* @param pred condition for the loop, which may not be {@code null}. Its result type must be {@code boolean}. See
* above for other constraints.
* @param body body of the loop, which may not be {@code null}. It controls the loop parameters and result type.
* See above for other constraints.
*
* @return a method handle implementing the {@code while} loop as described by the arguments.
* @throws IllegalArgumentException if the rules for the arguments are violated.
* @throws NullPointerException if {@code pred} or {@code body} are {@code null}.
*
* @see #loop(MethodHandle[][])
* @see #doWhileLoop(MethodHandle, MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle whileLoop(MethodHandle init, MethodHandle pred, MethodHandle body) {
whileLoopChecks(init, pred, body);
MethodHandle fini = identityOrVoid(body.type().returnType());
MethodHandle[] checkExit = { null, null, pred, fini };
MethodHandle[] varBody = { init, body };
return loop(checkExit, varBody);
}
/**
* Constructs a {@code do-while} loop from an initializer, a body, and a predicate.
* This is a convenience wrapper for the {@linkplain #loop(MethodHandle[][]) generic loop combinator}.
* <p>
* The {@code pred} handle describes the loop condition; and {@code body}, its body. The loop resulting from this
* method will, in each iteration, first execute its body and then evaluate the predicate.
* The loop will terminate once the predicate evaluates to {@code false} after an execution of the body.
* <p>
* The {@code init} handle describes the initial value of an additional optional loop-local variable.
* In each iteration, this loop-local variable, if present, will be passed to the {@code body}
* and updated with the value returned from its invocation. The result of loop execution will be
* the final value of the additional loop-local variable (if present).
* <p>
* The following rules hold for these argument handles:<ul>
* <li>The {@code body} handle must not be {@code null}; its type must be of the form
* {@code (V A...)V}, where {@code V} is non-{@code void}, or else {@code (A...)void}.
* (In the {@code void} case, we assign the type {@code void} to the name {@code V},
* and we will write {@code (V A...)V} with the understanding that a {@code void} type {@code V}
* is quietly dropped from the parameter list, leaving {@code (A...)V}.)
* <li>The parameter list {@code (V A...)} of the body is called the <em>internal parameter list</em>.
* It will constrain the parameter lists of the other loop parts.
* <li>If the iteration variable type {@code V} is dropped from the internal parameter list, the resulting shorter
* list {@code (A...)} is called the <em>external parameter list</em>.
* <li>The body return type {@code V}, if non-{@code void}, determines the type of an
* additional state variable of the loop.
* The body must both accept and return a value of this type {@code V}.
* <li>If {@code init} is non-{@code null}, it must have return type {@code V}.
* Its parameter list (of some <a href="MethodHandles.html#astar">form {@code (A*)}</a>) must be
* <a href="MethodHandles.html#effid">effectively identical</a>
* to the external parameter list {@code (A...)}.
* <li>If {@code init} is {@code null}, the loop variable will be initialized to its
* {@linkplain #empty default value}.
* <li>The {@code pred} handle must not be {@code null}. It must have {@code boolean} as its return type.
* Its parameter list (either empty or of the form {@code (V A*)}) must be
* effectively identical to the internal parameter list.
* </ul>
* <p>
* The resulting loop handle's result type and parameter signature are determined as follows:<ul>
* <li>The loop handle's result type is the result type {@code V} of the body.
* <li>The loop handle's parameter types are the types {@code (A...)},
* from the external parameter list.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. In the code, {@code V}/{@code v} represent the type / value of
* the sole loop variable as well as the result type of the loop; and {@code A}/{@code a}, that of the argument
* passed to the loop.
* {@snippet lang="java" :
* V init(A...);
* boolean pred(V, A...);
* V body(V, A...);
* V doWhileLoop(A... a...) {
* V v = init(a...);
* do {
* v = body(v, a...);
* } while (pred(v, a...));
* return v;
* }
* }
*
* @apiNote Example:
* {@snippet lang="java" :
* // int i = 0; while (i < limit) { ++i; } return i; => limit
* static int zero(int limit) { return 0; }
* static int step(int i, int limit) { return i + 1; }
* static boolean pred(int i, int limit) { return i < limit; }
* // assume MH_zero, MH_step, and MH_pred are handles to the above methods
* MethodHandle loop = MethodHandles.doWhileLoop(MH_zero, MH_step, MH_pred);
* assertEquals(23, loop.invoke(23));
* }
*
*
* @apiNote The implementation of this method can be expressed as follows:
* {@snippet lang="java" :
* MethodHandle doWhileLoop(MethodHandle init, MethodHandle body, MethodHandle pred) {
* MethodHandle fini = (body.type().returnType() == void.class
* ? null : identity(body.type().returnType()));
* MethodHandle[] clause = { init, body, pred, fini };
* return loop(clause);
* }
* }
*
* @param init optional initializer, providing the initial value of the loop variable.
* May be {@code null}, implying a default initial value. See above for other constraints.
* @param body body of the loop, which may not be {@code null}. It controls the loop parameters and result type.
* See above for other constraints.
* @param pred condition for the loop, which may not be {@code null}. Its result type must be {@code boolean}. See
* above for other constraints.
*
* @return a method handle implementing the {@code while} loop as described by the arguments.
* @throws IllegalArgumentException if the rules for the arguments are violated.
* @throws NullPointerException if {@code pred} or {@code body} are {@code null}.
*
* @see #loop(MethodHandle[][])
* @see #whileLoop(MethodHandle, MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle doWhileLoop(MethodHandle init, MethodHandle body, MethodHandle pred) {
whileLoopChecks(init, pred, body);
MethodHandle fini = identityOrVoid(body.type().returnType());
MethodHandle[] clause = {init, body, pred, fini };
return loop(clause);
}
private static void whileLoopChecks(MethodHandle init, MethodHandle pred, MethodHandle body) {
Objects.requireNonNull(pred);
Objects.requireNonNull(body);
MethodType bodyType = body.type();
Class<?> returnType = bodyType.returnType();
List<Class<?>> innerList = bodyType.parameterList();
List<Class<?>> outerList = innerList;
if (returnType == void.class) {
// OK
} else if (innerList.isEmpty() || innerList.get(0) != returnType) {
// leading V argument missing => error
MethodType expected = bodyType.insertParameterTypes(0, returnType);
throw misMatchedTypes("body function", bodyType, expected);
} else {
outerList = innerList.subList(1, innerList.size());
}
MethodType predType = pred.type();
if (predType.returnType() != boolean.class ||
!predType.effectivelyIdenticalParameters(0, innerList)) {
throw misMatchedTypes("loop predicate", predType, methodType(boolean.class, innerList));
}
if (init != null) {
MethodType initType = init.type();
if (initType.returnType() != returnType ||
!initType.effectivelyIdenticalParameters(0, outerList)) {
throw misMatchedTypes("loop initializer", initType, methodType(returnType, outerList));
}
}
}
/**
* Constructs a loop that runs a given number of iterations.
* This is a convenience wrapper for the {@linkplain #loop(MethodHandle[][]) generic loop combinator}.
* <p>
* The number of iterations is determined by the {@code iterations} handle evaluation result.
* The loop counter {@code i} is an extra loop iteration variable of type {@code int}.
* It will be initialized to 0 and incremented by 1 in each iteration.
* <p>
* If the {@code body} handle returns a non-{@code void} type {@code V}, a leading loop iteration variable
* of that type is also present. This variable is initialized using the optional {@code init} handle,
* or to the {@linkplain #empty default value} of type {@code V} if that handle is {@code null}.
* <p>
* In each iteration, the iteration variables are passed to an invocation of the {@code body} handle.
* A non-{@code void} value returned from the body (of type {@code V}) updates the leading
* iteration variable.
* The result of the loop handle execution will be the final {@code V} value of that variable
* (or {@code void} if there is no {@code V} variable).
* <p>
* The following rules hold for the argument handles:<ul>
* <li>The {@code iterations} handle must not be {@code null}, and must return
* the type {@code int}, referred to here as {@code I} in parameter type lists.
* <li>The {@code body} handle must not be {@code null}; its type must be of the form
* {@code (V I A...)V}, where {@code V} is non-{@code void}, or else {@code (I A...)void}.
* (In the {@code void} case, we assign the type {@code void} to the name {@code V},
* and we will write {@code (V I A...)V} with the understanding that a {@code void} type {@code V}
* is quietly dropped from the parameter list, leaving {@code (I A...)V}.)
* <li>The parameter list {@code (V I A...)} of the body contributes to a list
* of types called the <em>internal parameter list</em>.
* It will constrain the parameter lists of the other loop parts.
* <li>As a special case, if the body contributes only {@code V} and {@code I} types,
* with no additional {@code A} types, then the internal parameter list is extended by
* the argument types {@code A...} of the {@code iterations} handle.
* <li>If the iteration variable types {@code (V I)} are dropped from the internal parameter list, the resulting shorter
* list {@code (A...)} is called the <em>external parameter list</em>.
* <li>The body return type {@code V}, if non-{@code void}, determines the type of an
* additional state variable of the loop.
* The body must both accept a leading parameter and return a value of this type {@code V}.
* <li>If {@code init} is non-{@code null}, it must have return type {@code V}.
* Its parameter list (of some <a href="MethodHandles.html#astar">form {@code (A*)}</a>) must be
* <a href="MethodHandles.html#effid">effectively identical</a>
* to the external parameter list {@code (A...)}.
* <li>If {@code init} is {@code null}, the loop variable will be initialized to its
* {@linkplain #empty default value}.
* <li>The parameter list of {@code iterations} (of some form {@code (A*)}) must be
* effectively identical to the external parameter list {@code (A...)}.
* </ul>
* <p>
* The resulting loop handle's result type and parameter signature are determined as follows:<ul>
* <li>The loop handle's result type is the result type {@code V} of the body.
* <li>The loop handle's parameter types are the types {@code (A...)},
* from the external parameter list.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. In the code, {@code V}/{@code v} represent the type / value of
* the second loop variable as well as the result type of the loop; and {@code A...}/{@code a...} represent
* arguments passed to the loop.
* {@snippet lang="java" :
* int iterations(A...);
* V init(A...);
* V body(V, int, A...);
* V countedLoop(A... a...) {
* int end = iterations(a...);
* V v = init(a...);
* for (int i = 0; i < end; ++i) {
* v = body(v, i, a...);
* }
* return v;
* }
* }
*
* @apiNote Example with a fully conformant body method:
* {@snippet lang="java" :
* // String s = "Lambdaman!"; for (int i = 0; i < 13; ++i) { s = "na " + s; } return s;
* // => a variation on a well known theme
* static String step(String v, int counter, String init) { return "na " + v; }
* // assume MH_step is a handle to the method above
* MethodHandle fit13 = MethodHandles.constant(int.class, 13);
* MethodHandle start = MethodHandles.identity(String.class);
* MethodHandle loop = MethodHandles.countedLoop(fit13, start, MH_step);
* assertEquals("na na na na na na na na na na na na na Lambdaman!", loop.invoke("Lambdaman!"));
* }
*
* @apiNote Example with the simplest possible body method type,
* and passing the number of iterations to the loop invocation:
* {@snippet lang="java" :
* // String s = "Lambdaman!"; for (int i = 0; i < 13; ++i) { s = "na " + s; } return s;
* // => a variation on a well known theme
* static String step(String v, int counter ) { return "na " + v; }
* // assume MH_step is a handle to the method above
* MethodHandle count = MethodHandles.dropArguments(MethodHandles.identity(int.class), 1, String.class);
* MethodHandle start = MethodHandles.dropArguments(MethodHandles.identity(String.class), 0, int.class);
* MethodHandle loop = MethodHandles.countedLoop(count, start, MH_step); // (v, i) -> "na " + v
* assertEquals("na na na na na na na na na na na na na Lambdaman!", loop.invoke(13, "Lambdaman!"));
* }
*
* @apiNote Example that treats the number of iterations, string to append to, and string to append
* as loop parameters:
* {@snippet lang="java" :
* // String s = "Lambdaman!", t = "na"; for (int i = 0; i < 13; ++i) { s = t + " " + s; } return s;
* // => a variation on a well known theme
* static String step(String v, int counter, int iterations_, String pre, String start_) { return pre + " " + v; }
* // assume MH_step is a handle to the method above
* MethodHandle count = MethodHandles.identity(int.class);
* MethodHandle start = MethodHandles.dropArguments(MethodHandles.identity(String.class), 0, int.class, String.class);
* MethodHandle loop = MethodHandles.countedLoop(count, start, MH_step); // (v, i, _, pre, _) -> pre + " " + v
* assertEquals("na na na na na na na na na na na na na Lambdaman!", loop.invoke(13, "na", "Lambdaman!"));
* }
*
* @apiNote Example that illustrates the usage of {@link #dropArgumentsToMatch(MethodHandle, int, List, int)}
* to enforce a loop type:
* {@snippet lang="java" :
* // String s = "Lambdaman!", t = "na"; for (int i = 0; i < 13; ++i) { s = t + " " + s; } return s;
* // => a variation on a well known theme
* static String step(String v, int counter, String pre) { return pre + " " + v; }
* // assume MH_step is a handle to the method above
* MethodType loopType = methodType(String.class, String.class, int.class, String.class);
* MethodHandle count = MethodHandles.dropArgumentsToMatch(MethodHandles.identity(int.class), 0, loopType.parameterList(), 1);
* MethodHandle start = MethodHandles.dropArgumentsToMatch(MethodHandles.identity(String.class), 0, loopType.parameterList(), 2);
* MethodHandle body = MethodHandles.dropArgumentsToMatch(MH_step, 2, loopType.parameterList(), 0);
* MethodHandle loop = MethodHandles.countedLoop(count, start, body); // (v, i, pre, _, _) -> pre + " " + v
* assertEquals("na na na na na na na na na na na na na Lambdaman!", loop.invoke("na", 13, "Lambdaman!"));
* }
*
* @apiNote The implementation of this method can be expressed as follows:
* {@snippet lang="java" :
* MethodHandle countedLoop(MethodHandle iterations, MethodHandle init, MethodHandle body) {
* return countedLoop(empty(iterations.type()), iterations, init, body);
* }
* }
*
* @param iterations a non-{@code null} handle to return the number of iterations this loop should run. The handle's
* result type must be {@code int}. See above for other constraints.
* @param init optional initializer, providing the initial value of the loop variable.
* May be {@code null}, implying a default initial value. See above for other constraints.
* @param body body of the loop, which may not be {@code null}.
* It controls the loop parameters and result type in the standard case (see above for details).
* It must accept its own return type (if non-void) plus an {@code int} parameter (for the counter),
* and may accept any number of additional types.
* See above for other constraints.
*
* @return a method handle representing the loop.
* @throws NullPointerException if either of the {@code iterations} or {@code body} handles is {@code null}.
* @throws IllegalArgumentException if any argument violates the rules formulated above.
*
* @see #countedLoop(MethodHandle, MethodHandle, MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle countedLoop(MethodHandle iterations, MethodHandle init, MethodHandle body) {
return countedLoop(empty(iterations.type()), iterations, init, body);
}
/**
* Constructs a loop that counts over a range of numbers.
* This is a convenience wrapper for the {@linkplain #loop(MethodHandle[][]) generic loop combinator}.
* <p>
* The loop counter {@code i} is a loop iteration variable of type {@code int}.
* The {@code start} and {@code end} handles determine the start (inclusive) and end (exclusive)
* values of the loop counter.
* The loop counter will be initialized to the {@code int} value returned from the evaluation of the
* {@code start} handle and run to the value returned from {@code end} (exclusively) with a step width of 1.
* <p>
* If the {@code body} handle returns a non-{@code void} type {@code V}, a leading loop iteration variable
* of that type is also present. This variable is initialized using the optional {@code init} handle,
* or to the {@linkplain #empty default value} of type {@code V} if that handle is {@code null}.
* <p>
* In each iteration, the iteration variables are passed to an invocation of the {@code body} handle.
* A non-{@code void} value returned from the body (of type {@code V}) updates the leading
* iteration variable.
* The result of the loop handle execution will be the final {@code V} value of that variable
* (or {@code void} if there is no {@code V} variable).
* <p>
* The following rules hold for the argument handles:<ul>
* <li>The {@code start} and {@code end} handles must not be {@code null}, and must both return
* the common type {@code int}, referred to here as {@code I} in parameter type lists.
* <li>The {@code body} handle must not be {@code null}; its type must be of the form
* {@code (V I A...)V}, where {@code V} is non-{@code void}, or else {@code (I A...)void}.
* (In the {@code void} case, we assign the type {@code void} to the name {@code V},
* and we will write {@code (V I A...)V} with the understanding that a {@code void} type {@code V}
* is quietly dropped from the parameter list, leaving {@code (I A...)V}.)
* <li>The parameter list {@code (V I A...)} of the body contributes to a list
* of types called the <em>internal parameter list</em>.
* It will constrain the parameter lists of the other loop parts.
* <li>As a special case, if the body contributes only {@code V} and {@code I} types,
* with no additional {@code A} types, then the internal parameter list is extended by
* the argument types {@code A...} of the {@code end} handle.
* <li>If the iteration variable types {@code (V I)} are dropped from the internal parameter list, the resulting shorter
* list {@code (A...)} is called the <em>external parameter list</em>.
* <li>The body return type {@code V}, if non-{@code void}, determines the type of an
* additional state variable of the loop.
* The body must both accept a leading parameter and return a value of this type {@code V}.
* <li>If {@code init} is non-{@code null}, it must have return type {@code V}.
* Its parameter list (of some <a href="MethodHandles.html#astar">form {@code (A*)}</a>) must be
* <a href="MethodHandles.html#effid">effectively identical</a>
* to the external parameter list {@code (A...)}.
* <li>If {@code init} is {@code null}, the loop variable will be initialized to its
* {@linkplain #empty default value}.
* <li>The parameter list of {@code start} (of some form {@code (A*)}) must be
* effectively identical to the external parameter list {@code (A...)}.
* <li>Likewise, the parameter list of {@code end} must be effectively identical
* to the external parameter list.
* </ul>
* <p>
* The resulting loop handle's result type and parameter signature are determined as follows:<ul>
* <li>The loop handle's result type is the result type {@code V} of the body.
* <li>The loop handle's parameter types are the types {@code (A...)},
* from the external parameter list.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. In the code, {@code V}/{@code v} represent the type / value of
* the second loop variable as well as the result type of the loop; and {@code A...}/{@code a...} represent
* arguments passed to the loop.
* {@snippet lang="java" :
* int start(A...);
* int end(A...);
* V init(A...);
* V body(V, int, A...);
* V countedLoop(A... a...) {
* int e = end(a...);
* int s = start(a...);
* V v = init(a...);
* for (int i = s; i < e; ++i) {
* v = body(v, i, a...);
* }
* return v;
* }
* }
*
* @apiNote The implementation of this method can be expressed as follows:
* {@snippet lang="java" :
* MethodHandle countedLoop(MethodHandle start, MethodHandle end, MethodHandle init, MethodHandle body) {
* MethodHandle returnVar = dropArguments(identity(init.type().returnType()), 0, int.class, int.class);
* // assume MH_increment and MH_predicate are handles to implementation-internal methods with
* // the following semantics:
* // MH_increment: (int limit, int counter) -> counter + 1
* // MH_predicate: (int limit, int counter) -> counter < limit
* Class<?> counterType = start.type().returnType(); // int
* Class<?> returnType = body.type().returnType();
* MethodHandle incr = MH_increment, pred = MH_predicate, retv = null;
* if (returnType != void.class) { // ignore the V variable
* incr = dropArguments(incr, 1, returnType); // (limit, v, i) => (limit, i)
* pred = dropArguments(pred, 1, returnType); // ditto
* retv = dropArguments(identity(returnType), 0, counterType); // ignore limit
* }
* body = dropArguments(body, 0, counterType); // ignore the limit variable
* MethodHandle[]
* loopLimit = { end, null, pred, retv }, // limit = end(); i < limit || return v
* bodyClause = { init, body }, // v = init(); v = body(v, i)
* indexVar = { start, incr }; // i = start(); i = i + 1
* return loop(loopLimit, bodyClause, indexVar);
* }
* }
*
* @param start a non-{@code null} handle to return the start value of the loop counter, which must be {@code int}.
* See above for other constraints.
* @param end a non-{@code null} handle to return the end value of the loop counter (the loop will run to
* {@code end-1}). The result type must be {@code int}. See above for other constraints.
* @param init optional initializer, providing the initial value of the loop variable.
* May be {@code null}, implying a default initial value. See above for other constraints.
* @param body body of the loop, which may not be {@code null}.
* It controls the loop parameters and result type in the standard case (see above for details).
* It must accept its own return type (if non-void) plus an {@code int} parameter (for the counter),
* and may accept any number of additional types.
* See above for other constraints.
*
* @return a method handle representing the loop.
* @throws NullPointerException if any of the {@code start}, {@code end}, or {@code body} handles is {@code null}.
* @throws IllegalArgumentException if any argument violates the rules formulated above.
*
* @see #countedLoop(MethodHandle, MethodHandle, MethodHandle)
* @since 9
*/
public static MethodHandle countedLoop(MethodHandle start, MethodHandle end, MethodHandle init, MethodHandle body) {
countedLoopChecks(start, end, init, body);
Class<?> counterType = start.type().returnType(); // int, but who's counting?
Class<?> limitType = end.type().returnType(); // yes, int again
Class<?> returnType = body.type().returnType();
MethodHandle incr = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_countedLoopStep);
MethodHandle pred = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_countedLoopPred);
MethodHandle retv = null;
if (returnType != void.class) {
incr = dropArguments(incr, 1, returnType); // (limit, v, i) => (limit, i)
pred = dropArguments(pred, 1, returnType); // ditto
retv = dropArguments(identity(returnType), 0, counterType);
}
body = dropArguments(body, 0, counterType); // ignore the limit variable
MethodHandle[]
loopLimit = { end, null, pred, retv }, // limit = end(); i < limit || return v
bodyClause = { init, body }, // v = init(); v = body(v, i)
indexVar = { start, incr }; // i = start(); i = i + 1
return loop(loopLimit, bodyClause, indexVar);
}
private static void countedLoopChecks(MethodHandle start, MethodHandle end, MethodHandle init, MethodHandle body) {
Objects.requireNonNull(start);
Objects.requireNonNull(end);
Objects.requireNonNull(body);
Class<?> counterType = start.type().returnType();
if (counterType != int.class) {
MethodType expected = start.type().changeReturnType(int.class);
throw misMatchedTypes("start function", start.type(), expected);
} else if (end.type().returnType() != counterType) {
MethodType expected = end.type().changeReturnType(counterType);
throw misMatchedTypes("end function", end.type(), expected);
}
MethodType bodyType = body.type();
Class<?> returnType = bodyType.returnType();
List<Class<?>> innerList = bodyType.parameterList();
// strip leading V value if present
int vsize = (returnType == void.class ? 0 : 1);
if (vsize != 0 && (innerList.isEmpty() || innerList.get(0) != returnType)) {
// argument list has no "V" => error
MethodType expected = bodyType.insertParameterTypes(0, returnType);
throw misMatchedTypes("body function", bodyType, expected);
} else if (innerList.size() <= vsize || innerList.get(vsize) != counterType) {
// missing I type => error
MethodType expected = bodyType.insertParameterTypes(vsize, counterType);
throw misMatchedTypes("body function", bodyType, expected);
}
List<Class<?>> outerList = innerList.subList(vsize + 1, innerList.size());
if (outerList.isEmpty()) {
// special case; take lists from end handle
outerList = end.type().parameterList();
innerList = bodyType.insertParameterTypes(vsize + 1, outerList).parameterList();
}
MethodType expected = methodType(counterType, outerList);
if (!start.type().effectivelyIdenticalParameters(0, outerList)) {
throw misMatchedTypes("start parameter types", start.type(), expected);
}
if (end.type() != start.type() &&
!end.type().effectivelyIdenticalParameters(0, outerList)) {
throw misMatchedTypes("end parameter types", end.type(), expected);
}
if (init != null) {
MethodType initType = init.type();
if (initType.returnType() != returnType ||
!initType.effectivelyIdenticalParameters(0, outerList)) {
throw misMatchedTypes("loop initializer", initType, methodType(returnType, outerList));
}
}
}
/**
* Constructs a loop that ranges over the values produced by an {@code Iterator<T>}.
* This is a convenience wrapper for the {@linkplain #loop(MethodHandle[][]) generic loop combinator}.
* <p>
* The iterator itself will be determined by the evaluation of the {@code iterator} handle.
* Each value it produces will be stored in a loop iteration variable of type {@code T}.
* <p>
* If the {@code body} handle returns a non-{@code void} type {@code V}, a leading loop iteration variable
* of that type is also present. This variable is initialized using the optional {@code init} handle,
* or to the {@linkplain #empty default value} of type {@code V} if that handle is {@code null}.
* <p>
* In each iteration, the iteration variables are passed to an invocation of the {@code body} handle.
* A non-{@code void} value returned from the body (of type {@code V}) updates the leading
* iteration variable.
* The result of the loop handle execution will be the final {@code V} value of that variable
* (or {@code void} if there is no {@code V} variable).
* <p>
* The following rules hold for the argument handles:<ul>
* <li>The {@code body} handle must not be {@code null}; its type must be of the form
* {@code (V T A...)V}, where {@code V} is non-{@code void}, or else {@code (T A...)void}.
* (In the {@code void} case, we assign the type {@code void} to the name {@code V},
* and we will write {@code (V T A...)V} with the understanding that a {@code void} type {@code V}
* is quietly dropped from the parameter list, leaving {@code (T A...)V}.)
* <li>The parameter list {@code (V T A...)} of the body contributes to a list
* of types called the <em>internal parameter list</em>.
* It will constrain the parameter lists of the other loop parts.
* <li>As a special case, if the body contributes only {@code V} and {@code T} types,
* with no additional {@code A} types, then the internal parameter list is extended by
* the argument types {@code A...} of the {@code iterator} handle; if it is {@code null} the
* single type {@code Iterable} is added and constitutes the {@code A...} list.
* <li>If the iteration variable types {@code (V T)} are dropped from the internal parameter list, the resulting shorter
* list {@code (A...)} is called the <em>external parameter list</em>.
* <li>The body return type {@code V}, if non-{@code void}, determines the type of an
* additional state variable of the loop.
* The body must both accept a leading parameter and return a value of this type {@code V}.
* <li>If {@code init} is non-{@code null}, it must have return type {@code V}.
* Its parameter list (of some <a href="MethodHandles.html#astar">form {@code (A*)}</a>) must be
* <a href="MethodHandles.html#effid">effectively identical</a>
* to the external parameter list {@code (A...)}.
* <li>If {@code init} is {@code null}, the loop variable will be initialized to its
* {@linkplain #empty default value}.
* <li>If the {@code iterator} handle is non-{@code null}, it must have the return
* type {@code java.util.Iterator} or a subtype thereof.
* The iterator it produces when the loop is executed will be assumed
* to yield values which can be converted to type {@code T}.
* <li>The parameter list of an {@code iterator} that is non-{@code null} (of some form {@code (A*)}) must be
* effectively identical to the external parameter list {@code (A...)}.
* <li>If {@code iterator} is {@code null} it defaults to a method handle which behaves
* like {@link java.lang.Iterable#iterator()}. In that case, the internal parameter list
* {@code (V T A...)} must have at least one {@code A} type, and the default iterator
* handle parameter is adjusted to accept the leading {@code A} type, as if by
* the {@link MethodHandle#asType asType} conversion method.
* The leading {@code A} type must be {@code Iterable} or a subtype thereof.
* This conversion step, done at loop construction time, must not throw a {@code WrongMethodTypeException}.
* </ul>
* <p>
* The type {@code T} may be either a primitive or reference.
* Since type {@code Iterator<T>} is erased in the method handle representation to the raw type {@code Iterator},
* the {@code iteratedLoop} combinator adjusts the leading argument type for {@code body} to {@code Object}
* as if by the {@link MethodHandle#asType asType} conversion method.
* Therefore, if an iterator of the wrong type appears as the loop is executed, runtime exceptions may occur
* as the result of dynamic conversions performed by {@link MethodHandle#asType(MethodType)}.
* <p>
* The resulting loop handle's result type and parameter signature are determined as follows:<ul>
* <li>The loop handle's result type is the result type {@code V} of the body.
* <li>The loop handle's parameter types are the types {@code (A...)},
* from the external parameter list.
* </ul>
* <p>
* Here is pseudocode for the resulting loop handle. In the code, {@code V}/{@code v} represent the type / value of
* the loop variable as well as the result type of the loop; {@code T}/{@code t}, that of the elements of the
* structure the loop iterates over, and {@code A...}/{@code a...} represent arguments passed to the loop.
* {@snippet lang="java" :
* Iterator<T> iterator(A...); // defaults to Iterable::iterator
* V init(A...);
* V body(V,T,A...);
* V iteratedLoop(A... a...) {
* Iterator<T> it = iterator(a...);
* V v = init(a...);
* while (it.hasNext()) {
* T t = it.next();
* v = body(v, t, a...);
* }
* return v;
* }
* }
*
* @apiNote Example:
* {@snippet lang="java" :
* // get an iterator from a list
* static List<String> reverseStep(List<String> r, String e) {
* r.add(0, e);
* return r;
* }
* static List<String> newArrayList() { return new ArrayList<>(); }
* // assume MH_reverseStep and MH_newArrayList are handles to the above methods
* MethodHandle loop = MethodHandles.iteratedLoop(null, MH_newArrayList, MH_reverseStep);
* List<String> list = Arrays.asList("a", "b", "c", "d", "e");
* List<String> reversedList = Arrays.asList("e", "d", "c", "b", "a");
* assertEquals(reversedList, (List<String>) loop.invoke(list));
* }
*
* @apiNote The implementation of this method can be expressed approximately as follows:
* {@snippet lang="java" :
* MethodHandle iteratedLoop(MethodHandle iterator, MethodHandle init, MethodHandle body) {
* // assume MH_next, MH_hasNext, MH_startIter are handles to methods of Iterator/Iterable
* Class<?> returnType = body.type().returnType();
* Class<?> ttype = body.type().parameterType(returnType == void.class ? 0 : 1);
* MethodHandle nextVal = MH_next.asType(MH_next.type().changeReturnType(ttype));
* MethodHandle retv = null, step = body, startIter = iterator;
* if (returnType != void.class) {
* // the simple thing first: in (I V A...), drop the I to get V
* retv = dropArguments(identity(returnType), 0, Iterator.class);
* // body type signature (V T A...), internal loop types (I V A...)
* step = swapArguments(body, 0, 1); // swap V <-> T
* }
* if (startIter == null) startIter = MH_getIter;
* MethodHandle[]
* iterVar = { startIter, null, MH_hasNext, retv }, // it = iterator; while (it.hasNext())
* bodyClause = { init, filterArguments(step, 0, nextVal) }; // v = body(v, t, a)
* return loop(iterVar, bodyClause);
* }
* }
*
* @param iterator an optional handle to return the iterator to start the loop.
* If non-{@code null}, the handle must return {@link java.util.Iterator} or a subtype.
* See above for other constraints.
* @param init optional initializer, providing the initial value of the loop variable.
* May be {@code null}, implying a default initial value. See above for other constraints.
* @param body body of the loop, which may not be {@code null}.
* It controls the loop parameters and result type in the standard case (see above for details).
* It must accept its own return type (if non-void) plus a {@code T} parameter (for the iterated values),
* and may accept any number of additional types.
* See above for other constraints.
*
* @return a method handle embodying the iteration loop functionality.
* @throws NullPointerException if the {@code body} handle is {@code null}.
* @throws IllegalArgumentException if any argument violates the above requirements.
*
* @since 9
*/
public static MethodHandle iteratedLoop(MethodHandle iterator, MethodHandle init, MethodHandle body) {
Class<?> iterableType = iteratedLoopChecks(iterator, init, body);
Class<?> returnType = body.type().returnType();
MethodHandle hasNext = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_iteratePred);
MethodHandle nextRaw = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_iterateNext);
MethodHandle startIter;
MethodHandle nextVal;
{
MethodType iteratorType;
if (iterator == null) {
// derive argument type from body, if available, else use Iterable
startIter = MethodHandleImpl.getConstantHandle(MethodHandleImpl.MH_initIterator);
iteratorType = startIter.type().changeParameterType(0, iterableType);
} else {
// force return type to the internal iterator class
iteratorType = iterator.type().changeReturnType(Iterator.class);
startIter = iterator;
}
Class<?> ttype = body.type().parameterType(returnType == void.class ? 0 : 1);
MethodType nextValType = nextRaw.type().changeReturnType(ttype);
// perform the asType transforms under an exception transformer, as per spec.:
try {
startIter = startIter.asType(iteratorType);
nextVal = nextRaw.asType(nextValType);
} catch (WrongMethodTypeException ex) {
throw new IllegalArgumentException(ex);
}
}
MethodHandle retv = null, step = body;
if (returnType != void.class) {
// the simple thing first: in (I V A...), drop the I to get V
retv = dropArguments(identity(returnType), 0, Iterator.class);
// body type signature (V T A...), internal loop types (I V A...)
step = swapArguments(body, 0, 1); // swap V <-> T
}
MethodHandle[]
iterVar = { startIter, null, hasNext, retv },
bodyClause = { init, filterArgument(step, 0, nextVal) };
return loop(iterVar, bodyClause);
}
private static Class<?> iteratedLoopChecks(MethodHandle iterator, MethodHandle init, MethodHandle body) {
Objects.requireNonNull(body);
MethodType bodyType = body.type();
Class<?> returnType = bodyType.returnType();
List<Class<?>> internalParamList = bodyType.parameterList();
// strip leading V value if present
int vsize = (returnType == void.class ? 0 : 1);
if (vsize != 0 && (internalParamList.isEmpty() || internalParamList.get(0) != returnType)) {
// argument list has no "V" => error
MethodType expected = bodyType.insertParameterTypes(0, returnType);
throw misMatchedTypes("body function", bodyType, expected);
} else if (internalParamList.size() <= vsize) {
// missing T type => error
MethodType expected = bodyType.insertParameterTypes(vsize, Object.class);
throw misMatchedTypes("body function", bodyType, expected);
}
List<Class<?>> externalParamList = internalParamList.subList(vsize + 1, internalParamList.size());
Class<?> iterableType = null;
if (iterator != null) {
// special case; if the body handle only declares V and T then
// the external parameter list is obtained from iterator handle
if (externalParamList.isEmpty()) {
externalParamList = iterator.type().parameterList();
}
MethodType itype = iterator.type();
if (!Iterator.class.isAssignableFrom(itype.returnType())) {
throw newIllegalArgumentException("iteratedLoop first argument must have Iterator return type");
}
if (!itype.effectivelyIdenticalParameters(0, externalParamList)) {
MethodType expected = methodType(itype.returnType(), externalParamList);
throw misMatchedTypes("iterator parameters", itype, expected);
}
} else {
if (externalParamList.isEmpty()) {
// special case; if the iterator handle is null and the body handle
// only declares V and T then the external parameter list consists
// of Iterable
externalParamList = List.of(Iterable.class);
iterableType = Iterable.class;
} else {
// special case; if the iterator handle is null and the external
// parameter list is not empty then the first parameter must be
// assignable to Iterable
iterableType = externalParamList.get(0);
if (!Iterable.class.isAssignableFrom(iterableType)) {
throw newIllegalArgumentException(
"inferred first loop argument must inherit from Iterable: " + iterableType);
}
}
}
if (init != null) {
MethodType initType = init.type();
if (initType.returnType() != returnType ||
!initType.effectivelyIdenticalParameters(0, externalParamList)) {
throw misMatchedTypes("loop initializer", initType, methodType(returnType, externalParamList));
}
}
return iterableType; // help the caller a bit
}
/*non-public*/
static MethodHandle swapArguments(MethodHandle mh, int i, int j) {
// there should be a better way to uncross my wires
int arity = mh.type().parameterCount();
int[] order = new int[arity];
for (int k = 0; k < arity; k++) order[k] = k;
order[i] = j; order[j] = i;
Class<?>[] types = mh.type().parameterArray();
Class<?> ti = types[i]; types[i] = types[j]; types[j] = ti;
MethodType swapType = methodType(mh.type().returnType(), types);
return permuteArguments(mh, swapType, order);
}
/**
* Makes a method handle that adapts a {@code target} method handle by wrapping it in a {@code try-finally} block.
* Another method handle, {@code cleanup}, represents the functionality of the {@code finally} block. Any exception
* thrown during the execution of the {@code target} handle will be passed to the {@code cleanup} handle. The
* exception will be rethrown, unless {@code cleanup} handle throws an exception first. The
* value returned from the {@code cleanup} handle's execution will be the result of the execution of the
* {@code try-finally} handle.
* <p>
* The {@code cleanup} handle will be passed one or two additional leading arguments.
* The first is the exception thrown during the
* execution of the {@code target} handle, or {@code null} if no exception was thrown.
* The second is the result of the execution of the {@code target} handle, or, if it throws an exception,
* a {@code null}, zero, or {@code false} value of the required type is supplied as a placeholder.
* The second argument is not present if the {@code target} handle has a {@code void} return type.
* (Note that, except for argument type conversions, combinators represent {@code void} values in parameter lists
* by omitting the corresponding paradoxical arguments, not by inserting {@code null} or zero values.)
* <p>
* The {@code target} and {@code cleanup} handles must have the same corresponding argument and return types, except
* that the {@code cleanup} handle may omit trailing arguments. Also, the {@code cleanup} handle must have one or
* two extra leading parameters:<ul>
* <li>a {@code Throwable}, which will carry the exception thrown by the {@code target} handle (if any); and
* <li>a parameter of the same type as the return type of both {@code target} and {@code cleanup}, which will carry
* the result from the execution of the {@code target} handle.
* This parameter is not present if the {@code target} returns {@code void}.
* </ul>
* <p>
* The pseudocode for the resulting adapter looks as follows. In the code, {@code V} represents the result type of
* the {@code try/finally} construct; {@code A}/{@code a}, the types and values of arguments to the resulting
* handle consumed by the cleanup; and {@code B}/{@code b}, those of arguments to the resulting handle discarded by
* the cleanup.
* {@snippet lang="java" :
* V target(A..., B...);
* V cleanup(Throwable, V, A...);
* V adapter(A... a, B... b) {
* V result = (zero value for V);
* Throwable throwable = null;
* try {
* result = target(a..., b...);
* } catch (Throwable t) {
* throwable = t;
* throw t;
* } finally {
* result = cleanup(throwable, result, a...);
* }
* return result;
* }
* }
* <p>
* Note that the saved arguments ({@code a...} in the pseudocode) cannot
* be modified by execution of the target, and so are passed unchanged
* from the caller to the cleanup, if it is invoked.
* <p>
* The target and cleanup must return the same type, even if the cleanup
* always throws.
* To create such a throwing cleanup, compose the cleanup logic
* with {@link #throwException throwException},
* in order to create a method handle of the correct return type.
* <p>
* Note that {@code tryFinally} never converts exceptions into normal returns.
* In rare cases where exceptions must be converted in that way, first wrap
* the target with {@link #catchException(MethodHandle, Class, MethodHandle)}
* to capture an outgoing exception, and then wrap with {@code tryFinally}.
* <p>
* It is recommended that the first parameter type of {@code cleanup} be
* declared {@code Throwable} rather than a narrower subtype. This ensures
* {@code cleanup} will always be invoked with whatever exception that
* {@code target} throws. Declaring a narrower type may result in a
* {@code ClassCastException} being thrown by the {@code try-finally}
* handle if the type of the exception thrown by {@code target} is not
* assignable to the first parameter type of {@code cleanup}. Note that
* various exception types of {@code VirtualMachineError},
* {@code LinkageError}, and {@code RuntimeException} can in principle be
* thrown by almost any kind of Java code, and a finally clause that
* catches (say) only {@code IOException} would mask any of the others
* behind a {@code ClassCastException}.
*
* @param target the handle whose execution is to be wrapped in a {@code try} block.
* @param cleanup the handle that is invoked in the finally block.
*
* @return a method handle embodying the {@code try-finally} block composed of the two arguments.
* @throws NullPointerException if any argument is null
* @throws IllegalArgumentException if {@code cleanup} does not accept
* the required leading arguments, or if the method handle types do
* not match in their return types and their
* corresponding trailing parameters
*
* @see MethodHandles#catchException(MethodHandle, Class, MethodHandle)
* @since 9
*/
public static MethodHandle tryFinally(MethodHandle target, MethodHandle cleanup) {
Class<?>[] targetParamTypes = target.type().ptypes();
Class<?> rtype = target.type().returnType();
tryFinallyChecks(target, cleanup);
// Match parameter lists: if the cleanup has a shorter parameter list than the target, add ignored arguments.
// The cleanup parameter list (minus the leading Throwable and result parameters) must be a sublist of the
// target parameter list.
cleanup = dropArgumentsToMatch(cleanup, (rtype == void.class ? 1 : 2), targetParamTypes, 0, false);
// Ensure that the intrinsic type checks the instance thrown by the
// target against the first parameter of cleanup
cleanup = cleanup.asType(cleanup.type().changeParameterType(0, Throwable.class));
// Use asFixedArity() to avoid unnecessary boxing of last argument for VarargsCollector case.
return MethodHandleImpl.makeTryFinally(target.asFixedArity(), cleanup.asFixedArity(), rtype, targetParamTypes);
}
private static void tryFinallyChecks(MethodHandle target, MethodHandle cleanup) {
Class<?> rtype = target.type().returnType();
if (rtype != cleanup.type().returnType()) {
throw misMatchedTypes("target and return types", cleanup.type().returnType(), rtype);
}
MethodType cleanupType = cleanup.type();
if (!Throwable.class.isAssignableFrom(cleanupType.parameterType(0))) {
throw misMatchedTypes("cleanup first argument and Throwable", cleanup.type(), Throwable.class);
}
if (rtype != void.class && cleanupType.parameterType(1) != rtype) {
throw misMatchedTypes("cleanup second argument and target return type", cleanup.type(), rtype);
}
// The cleanup parameter list (minus the leading Throwable and result parameters) must be a sublist of the
// target parameter list.
int cleanupArgIndex = rtype == void.class ? 1 : 2;
if (!cleanupType.effectivelyIdenticalParameters(cleanupArgIndex, target.type().parameterList())) {
throw misMatchedTypes("cleanup parameters after (Throwable,result) and target parameter list prefix",
cleanup.type(), target.type());
}
}
/**
* Creates a table switch method handle, which can be used to switch over a set of target
* method handles, based on a given target index, called selector.
* <p>
* For a selector value of {@code n}, where {@code n} falls in the range {@code [0, N)},
* and where {@code N} is the number of target method handles, the table switch method
* handle will invoke the n-th target method handle from the list of target method handles.
* <p>
* For a selector value that does not fall in the range {@code [0, N)}, the table switch
* method handle will invoke the given fallback method handle.
* <p>
* All method handles passed to this method must have the same type, with the additional
* requirement that the leading parameter be of type {@code int}. The leading parameter
* represents the selector.
* <p>
* Any trailing parameters present in the type will appear on the returned table switch
* method handle as well. Any arguments assigned to these parameters will be forwarded,
* together with the selector value, to the selected method handle when invoking it.
*
* @apiNote Example:
* The cases each drop the {@code selector} value they are given, and take an additional
* {@code String} argument, which is concatenated (using {@link String#concat(String)})
* to a specific constant label string for each case:
* {@snippet lang="java" :
* MethodHandles.Lookup lookup = MethodHandles.lookup();
* MethodHandle caseMh = lookup.findVirtual(String.class, "concat",
* MethodType.methodType(String.class, String.class));
* caseMh = MethodHandles.dropArguments(caseMh, 0, int.class);
*
* MethodHandle caseDefault = MethodHandles.insertArguments(caseMh, 1, "default: ");
* MethodHandle case0 = MethodHandles.insertArguments(caseMh, 1, "case 0: ");
* MethodHandle case1 = MethodHandles.insertArguments(caseMh, 1, "case 1: ");
*
* MethodHandle mhSwitch = MethodHandles.tableSwitch(
* caseDefault,
* case0,
* case1
* );
*
* assertEquals("default: data", (String) mhSwitch.invokeExact(-1, "data"));
* assertEquals("case 0: data", (String) mhSwitch.invokeExact(0, "data"));
* assertEquals("case 1: data", (String) mhSwitch.invokeExact(1, "data"));
* assertEquals("default: data", (String) mhSwitch.invokeExact(2, "data"));
* }
*
* @param fallback the fallback method handle that is called when the selector is not
* within the range {@code [0, N)}.
* @param targets array of target method handles.
* @return the table switch method handle.
* @throws NullPointerException if {@code fallback}, the {@code targets} array, or any
* any of the elements of the {@code targets} array are
* {@code null}.
* @throws IllegalArgumentException if the {@code targets} array is empty, if the leading
* parameter of the fallback handle or any of the target
* handles is not {@code int}, or if the types of
* the fallback handle and all of target handles are
* not the same.
*
* @since 17
*/
public static MethodHandle tableSwitch(MethodHandle fallback, MethodHandle... targets) {
Objects.requireNonNull(fallback);
Objects.requireNonNull(targets);
targets = targets.clone();
MethodType type = tableSwitchChecks(fallback, targets);
return MethodHandleImpl.makeTableSwitch(type, fallback, targets);
}
private static MethodType tableSwitchChecks(MethodHandle defaultCase, MethodHandle[] caseActions) {
if (caseActions.length == 0)
throw new IllegalArgumentException("Not enough cases: " + Arrays.toString(caseActions));
MethodType expectedType = defaultCase.type();
if (!(expectedType.parameterCount() >= 1) || expectedType.parameterType(0) != int.class)
throw new IllegalArgumentException(
"Case actions must have int as leading parameter: " + Arrays.toString(caseActions));
for (MethodHandle mh : caseActions) {
Objects.requireNonNull(mh);
if (mh.type() != expectedType)
throw new IllegalArgumentException(
"Case actions must have the same type: " + Arrays.toString(caseActions));
}
return expectedType;
}
/**
* Adapts a target var handle by pre-processing incoming and outgoing values using a pair of filter functions.
* <p>
* When calling e.g. {@link VarHandle#set(Object...)} on the resulting var handle, the incoming value (of type {@code T}, where
* {@code T} is the <em>last</em> parameter type of the first filter function) is processed using the first filter and then passed
* to the target var handle.
* Conversely, when calling e.g. {@link VarHandle#get(Object...)} on the resulting var handle, the return value obtained from
* the target var handle (of type {@code T}, where {@code T} is the <em>last</em> parameter type of the second filter function)
* is processed using the second filter and returned to the caller. More advanced access mode types, such as
* {@link VarHandle.AccessMode#COMPARE_AND_EXCHANGE} might apply both filters at the same time.
* <p>
* For the boxing and unboxing filters to be well-formed, their types must be of the form {@code (A... , S) -> T} and
* {@code (A... , T) -> S}, respectively, where {@code T} is the type of the target var handle. If this is the case,
* the resulting var handle will have type {@code S} and will feature the additional coordinates {@code A...} (which
* will be appended to the coordinates of the target var handle).
* <p>
* If the boxing and unboxing filters throw any checked exceptions when invoked, the resulting var handle will
* throw an {@link IllegalStateException}.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
*
* @param target the target var handle
* @param filterToTarget a filter to convert some type {@code S} into the type of {@code target}
* @param filterFromTarget a filter to convert the type of {@code target} to some type {@code S}
* @return an adapter var handle which accepts a new type, performing the provided boxing/unboxing conversions.
* @throws IllegalArgumentException if {@code filterFromTarget} and {@code filterToTarget} are not well-formed, that is, they have types
* other than {@code (A... , S) -> T} and {@code (A... , T) -> S}, respectively, where {@code T} is the type of the target var handle,
* or if it's determined that either {@code filterFromTarget} or {@code filterToTarget} throws any checked exceptions.
* @throws NullPointerException if any of the arguments is {@code null}.
* @since 22
*/
public static VarHandle filterValue(VarHandle target, MethodHandle filterToTarget, MethodHandle filterFromTarget) {
return VarHandles.filterValue(target, filterToTarget, filterFromTarget);
}
/**
* Adapts a target var handle by pre-processing incoming coordinate values using unary filter functions.
* <p>
* When calling e.g. {@link VarHandle#get(Object...)} on the resulting var handle, the incoming coordinate values
* starting at position {@code pos} (of type {@code C1, C2 ... Cn}, where {@code C1, C2 ... Cn} are the return types
* of the unary filter functions) are transformed into new values (of type {@code S1, S2 ... Sn}, where {@code S1, S2 ... Sn} are the
* parameter types of the unary filter functions), and then passed (along with any coordinate that was left unaltered
* by the adaptation) to the target var handle.
* <p>
* For the coordinate filters to be well-formed, their types must be of the form {@code S1 -> T1, S2 -> T1 ... Sn -> Tn},
* where {@code T1, T2 ... Tn} are the coordinate types starting at position {@code pos} of the target var handle.
* <p>
* If any of the filters throws a checked exception when invoked, the resulting var handle will
* throw an {@link IllegalStateException}.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
*
* @param target the target var handle
* @param pos the position of the first coordinate to be transformed
* @param filters the unary functions which are used to transform coordinates starting at position {@code pos}
* @return an adapter var handle which accepts new coordinate types, applying the provided transformation
* to the new coordinate values.
* @throws IllegalArgumentException if the handles in {@code filters} are not well-formed, that is, they have types
* other than {@code S1 -> T1, S2 -> T2, ... Sn -> Tn} where {@code T1, T2 ... Tn} are the coordinate types starting
* at position {@code pos} of the target var handle, if {@code pos} is not between 0 and the target var handle coordinate arity, inclusive,
* or if more filters are provided than the actual number of coordinate types available starting at {@code pos},
* or if it's determined that any of the filters throws any checked exceptions.
* @throws NullPointerException if any of the arguments is {@code null} or {@code filters} contains {@code null}.
* @since 22
*/
public static VarHandle filterCoordinates(VarHandle target, int pos, MethodHandle... filters) {
return VarHandles.filterCoordinates(target, pos, filters);
}
/**
* Provides a target var handle with one or more <em>bound coordinates</em>
* in advance of the var handle's invocation. As a consequence, the resulting var handle will feature less
* coordinate types than the target var handle.
* <p>
* When calling e.g. {@link VarHandle#get(Object...)} on the resulting var handle, incoming coordinate values
* are joined with bound coordinate values, and then passed to the target var handle.
* <p>
* For the bound coordinates to be well-formed, their types must be {@code T1, T2 ... Tn },
* where {@code T1, T2 ... Tn} are the coordinate types starting at position {@code pos} of the target var handle.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
*
* @param target the var handle to invoke after the bound coordinates are inserted
* @param pos the position of the first coordinate to be inserted
* @param values the series of bound coordinates to insert
* @return an adapter var handle which inserts additional coordinates,
* before calling the target var handle
* @throws IllegalArgumentException if {@code pos} is not between 0 and the target var handle coordinate arity, inclusive,
* or if more values are provided than the actual number of coordinate types available starting at {@code pos}.
* @throws ClassCastException if the bound coordinates in {@code values} are not well-formed, that is, they have types
* other than {@code T1, T2 ... Tn }, where {@code T1, T2 ... Tn} are the coordinate types starting at position {@code pos}
* of the target var handle.
* @throws NullPointerException if any of the arguments is {@code null} or {@code values} contains {@code null}.
* @since 22
*/
public static VarHandle insertCoordinates(VarHandle target, int pos, Object... values) {
return VarHandles.insertCoordinates(target, pos, values);
}
/**
* Provides a var handle which adapts the coordinate values of the target var handle, by re-arranging them
* so that the new coordinates match the provided ones.
* <p>
* The given array controls the reordering.
* Call {@code #I} the number of incoming coordinates (the value
* {@code newCoordinates.size()}), and call {@code #O} the number
* of outgoing coordinates (the number of coordinates associated with the target var handle).
* Then the length of the reordering array must be {@code #O},
* and each element must be a non-negative number less than {@code #I}.
* For every {@code N} less than {@code #O}, the {@code N}-th
* outgoing coordinate will be taken from the {@code I}-th incoming
* coordinate, where {@code I} is {@code reorder[N]}.
* <p>
* No coordinate value conversions are applied.
* The type of each incoming coordinate, as determined by {@code newCoordinates},
* must be identical to the type of the corresponding outgoing coordinate
* in the target var handle.
* <p>
* The reordering array need not specify an actual permutation.
* An incoming coordinate will be duplicated if its index appears
* more than once in the array, and an incoming coordinate will be dropped
* if its index does not appear in the array.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
* @param target the var handle to invoke after the coordinates have been reordered
* @param newCoordinates the new coordinate types
* @param reorder an index array which controls the reordering
* @return an adapter var handle which re-arranges the incoming coordinate values,
* before calling the target var handle
* @throws IllegalArgumentException if the index array length is not equal to
* the number of coordinates of the target var handle, or if any index array element is not a valid index for
* a coordinate of {@code newCoordinates}, or if two corresponding coordinate types in
* the target var handle and in {@code newCoordinates} are not identical.
* @throws NullPointerException if any of the arguments is {@code null} or {@code newCoordinates} contains {@code null}.
* @since 22
*/
public static VarHandle permuteCoordinates(VarHandle target, List<Class<?>> newCoordinates, int... reorder) {
return VarHandles.permuteCoordinates(target, newCoordinates, reorder);
}
/**
* Adapts a target var handle by pre-processing
* a sub-sequence of its coordinate values with a filter (a method handle).
* The pre-processed coordinates are replaced by the result (if any) of the
* filter function and the target var handle is then called on the modified (usually shortened)
* coordinate list.
* <p>
* If {@code R} is the return type of the filter, then:
* <ul>
* <li>if {@code R} <em>is not</em> {@code void}, the target var handle must have a coordinate of type {@code R} in
* position {@code pos}. The parameter types of the filter will replace the coordinate type at position {@code pos}
* of the target var handle. When the returned var handle is invoked, it will be as if the filter is invoked first,
* and its result is passed in place of the coordinate at position {@code pos} in a downstream invocation of the
* target var handle.</li>
* <li> if {@code R} <em>is</em> {@code void}, the parameter types (if any) of the filter will be inserted in the
* coordinate type list of the target var handle at position {@code pos}. In this case, when the returned var handle
* is invoked, the filter essentially acts as a side effect, consuming some of the coordinate values, before a
* downstream invocation of the target var handle.</li>
* </ul>
* <p>
* If any of the filters throws a checked exception when invoked, the resulting var handle will
* throw an {@link IllegalStateException}.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
*
* @param target the var handle to invoke after the coordinates have been filtered
* @param pos the position in the coordinate list of the target var handle where the filter is to be inserted
* @param filter the filter method handle
* @return an adapter var handle which filters the incoming coordinate values,
* before calling the target var handle
* @throws IllegalArgumentException if the return type of {@code filter}
* is not void, and it is not the same as the {@code pos} coordinate of the target var handle,
* if {@code pos} is not between 0 and the target var handle coordinate arity, inclusive,
* if the resulting var handle's type would have <a href="MethodHandle.html#maxarity">too many coordinates</a>,
* or if it's determined that {@code filter} throws any checked exceptions.
* @throws NullPointerException if any of the arguments is {@code null}.
* @since 22
*/
public static VarHandle collectCoordinates(VarHandle target, int pos, MethodHandle filter) {
return VarHandles.collectCoordinates(target, pos, filter);
}
/**
* Returns a var handle which will discard some dummy coordinates before delegating to the
* target var handle. As a consequence, the resulting var handle will feature more
* coordinate types than the target var handle.
* <p>
* The {@code pos} argument may range between zero and <i>N</i>, where <i>N</i> is the arity of the
* target var handle's coordinate types. If {@code pos} is zero, the dummy coordinates will precede
* the target's real arguments; if {@code pos} is <i>N</i> they will come after.
* <p>
* The resulting var handle will feature the same access modes (see {@link VarHandle.AccessMode}) and
* atomic access guarantees as those featured by the target var handle.
*
* @param target the var handle to invoke after the dummy coordinates are dropped
* @param pos position of the first coordinate to drop (zero for the leftmost)
* @param valueTypes the type(s) of the coordinate(s) to drop
* @return an adapter var handle which drops some dummy coordinates,
* before calling the target var handle
* @throws IllegalArgumentException if {@code pos} is not between 0 and the target var handle coordinate arity, inclusive.
* @throws NullPointerException if any of the arguments is {@code null} or {@code valueTypes} contains {@code null}.
* @since 22
*/
public static VarHandle dropCoordinates(VarHandle target, int pos, Class<?>... valueTypes) {
return VarHandles.dropCoordinates(target, pos, valueTypes);
}
}