FileDocCategorySizeDatePackage
Policy.javaAPI DocJava SE 6 API26490Tue Jun 10 00:25:46 BST 2008java.security

Policy

public abstract class Policy extends Object
A Policy object is responsible for determining whether code executing in the Java runtime environment has permission to perform a security-sensitive operation.

There is only one Policy object installed in the runtime at any given time. A Policy object can be installed by calling the setPolicy method. The installed Policy object can be obtained by calling the getPolicy method.

If no Policy object has been installed in the runtime, a call to getPolicy installs an instance of the default Policy implementation (a default subclass implementation of this abstract class). The default Policy implementation can be changed by setting the value of the "policy.provider" security property (in the Java security properties file) to the fully qualified name of the desired Policy subclass implementation. The Java security properties file is located in the file named <JAVA_HOME>/lib/security/java.security. <JAVA_HOME> refers to the value of the java.home system property, and specifies the directory where the JRE is installed.

Application code can directly subclass Policy to provide a custom implementation. In addition, an instance of a Policy object can be constructed by invoking one of the getInstance factory methods with a standard type. The default policy type is "JavaPolicy". See Appendix A in the Java Cryptography Architecture API Specification & Reference for a list of standard Policy types.

Once a Policy instance has been installed (either by default, or by calling setPolicy), the Java runtime invokes its implies when it needs to determine whether executing code (encapsulated in a ProtectionDomain) can perform SecurityManager-protected operations. How a Policy object retrieves its policy data is up to the Policy implementation itself. The policy data may be stored, for example, in a flat ASCII file, in a serialized binary file of the Policy class, or in a database.

The refresh method causes the policy object to refresh/reload its data. This operation is implementation-dependent. For example, if the policy object stores its data in configuration files, calling refresh will cause it to re-read the configuration policy files. If a refresh operation is not supported, this method does nothing. Note that refreshed policy may not have an effect on classes in a particular ProtectionDomain. This is dependent on the Policy provider's implementation of the implies method and its PermissionCollection caching strategy.

author
Roland Schemers
author
Gary Ellison
version
1.103, 11/17/06
see
java.security.Provider
see
java.security.ProtectionDomain
see
java.security.Permission

Fields Summary
public static final PermissionCollection
UNSUPPORTED_EMPTY_COLLECTION
A read-only empty PermissionCollection instance.
private static Policy
policy
the system-wide policy.
private static final Debug
debug
private WeakHashMap
pdMapping
Constructors Summary
Methods Summary
private voidaddStaticPerms(java.security.PermissionCollection perms, java.security.PermissionCollection statics)
add static permissions to provided permission collection

	if (statics != null) {
	    synchronized (statics) {
		Enumeration e = statics.elements();
		while (e.hasMoreElements()) {
		    perms.add((Permission)e.nextElement());
		}
	    }
	}
    
private static voidcheckPermission(java.lang.String type)

	SecurityManager sm = System.getSecurityManager();
	if (sm != null) {
	    sm.checkPermission(new SecurityPermission("createPolicy." + type));
	}
    
public static java.security.PolicygetInstance(java.lang.String type, java.security.Policy$Parameters params)
Returns a Policy object of the specified type.

This method traverses the list of registered security providers, starting with the most preferred Provider. A new Policy object encapsulating the PolicySpi implementation from the first Provider that supports the specified type is returned.

Note that the list of registered providers may be retrieved via the {@link Security#getProviders() Security.getProviders()} method.

param
type the specified Policy type. See Appendix A in the Java Cryptography Architecture API Specification & Reference for a list of standard Policy types.
param
params parameters for the Policy, which may be null.
return
the new Policy object.
exception
SecurityException if the caller does not have permission to get a Policy instance for the specified type.
exception
NullPointerException if the specified type is null.
exception
IllegalArgumentException if the specified parameters are not understood by the PolicySpi implementation from the selected Provider.
exception
NoSuchAlgorithmException if no Provider supports a PolicySpi implementation for the specified type.
see
Provider
since
1.6


	checkPermission(type);
	try {
	    GetInstance.Instance instance = GetInstance.getInstance("Policy",
							PolicySpi.class,
							type,
							params);
	    return new PolicyDelegate((PolicySpi)instance.impl,
							instance.provider,
							type,
							params);
	} catch (NoSuchAlgorithmException nsae) {
	    return handleException(nsae);
	}
    
public static java.security.PolicygetInstance(java.lang.String type, java.security.Policy$Parameters params, java.lang.String provider)
Returns a Policy object of the specified type.

A new Policy object encapsulating the PolicySpi implementation from the specified provider is returned. The specified provider must be registered in the provider list.

Note that the list of registered providers may be retrieved via the {@link Security#getProviders() Security.getProviders()} method.

param
type the specified Policy type. See Appendix A in the Java Cryptography Architecture API Specification & Reference for a list of standard Policy types.
param
params parameters for the Policy, which may be null.
param
provider the provider.
return
the new Policy object.
exception
SecurityException if the caller does not have permission to get a Policy instance for the specified type.
exception
NullPointerException if the specified type is null.
exception
IllegalArgumentException if the specified provider is null or empty, or if the specified parameters are not understood by the PolicySpi implementation from the specified provider.
exception
NoSuchProviderException if the specified provider is not registered in the security provider list.
exception
NoSuchAlgorithmException if the specified provider does not support a PolicySpi implementation for the specified type.
see
Provider
since
1.6


	if (provider == null || provider.length() == 0) {
	    throw new IllegalArgumentException("missing provider");
	}

	checkPermission(type);
	try {
	    GetInstance.Instance instance = GetInstance.getInstance("Policy",
							PolicySpi.class,
							type,
							params,
							provider);
	    return new PolicyDelegate((PolicySpi)instance.impl,
							instance.provider,
							type,
							params);
	} catch (NoSuchAlgorithmException nsae) {
	    return handleException (nsae);
	}
    
public static java.security.PolicygetInstance(java.lang.String type, java.security.Policy$Parameters params, java.security.Provider provider)
Returns a Policy object of the specified type.

A new Policy object encapsulating the PolicySpi implementation from the specified Provider object is returned. Note that the specified Provider object does not have to be registered in the provider list.

param
type the specified Policy type. See Appendix A in the Java Cryptography Architecture API Specification & Reference for a list of standard Policy types.
param
params parameters for the Policy, which may be null.
param
provider the Provider.
return
the new Policy object.
exception
SecurityException if the caller does not have permission to get a Policy instance for the specified type.
exception
NullPointerException if the specified type is null.
exception
IllegalArgumentException if the specified Provider is null, or if the specified parameters are not understood by the PolicySpi implementation from the specified Provider.
exception
NoSuchAlgorithmException if the specified Provider does not support a PolicySpi implementation for the specified type.
see
Provider
since
1.6


	if (provider == null) {
	    throw new IllegalArgumentException("missing provider");
	}

	checkPermission(type);
	try {
	    GetInstance.Instance instance = GetInstance.getInstance("Policy",
							PolicySpi.class,
							type,
							params,
							provider);
	    return new PolicyDelegate((PolicySpi)instance.impl,
							instance.provider,
							type,
							params);
	} catch (NoSuchAlgorithmException nsae) {
	    return handleException (nsae);
	}
    
public java.security.Policy$ParametersgetParameters()
Return Policy parameters.

This Policy instance will only have parameters if it was obtained via a call to Policy.getInstance. Otherwise this method returns null.

return
Policy parameters, or null.
since
1.6

	return null;
    
public java.security.PermissionCollectiongetPermissions(java.security.CodeSource codesource)
Return a PermissionCollection object containing the set of permissions granted to the specified CodeSource.

Applications are discouraged from calling this method since this operation may not be supported by all policy implementations. Applications should solely rely on the implies method to perform policy checks. If an application absolutely must call a getPermissions method, it should call getPermissions(ProtectionDomain).

The default implementation of this method returns Policy.UNSUPPORTED_EMPTY_COLLECTION. This method can be overridden if the policy implementation can return a set of permissions granted to a CodeSource.

param
codesource the CodeSource to which the returned PermissionCollection has been granted.
return
a set of permissions granted to the specified CodeSource. If this operation is supported, the returned set of permissions must be a new mutable instance and it must support heterogeneous Permission types. If this operation is not supported, Policy.UNSUPPORTED_EMPTY_COLLECTION is returned.

	return Policy.UNSUPPORTED_EMPTY_COLLECTION;
    
public java.security.PermissionCollectiongetPermissions(java.security.ProtectionDomain domain)
Return a PermissionCollection object containing the set of permissions granted to the specified ProtectionDomain.

Applications are discouraged from calling this method since this operation may not be supported by all policy implementations. Applications should rely on the implies method to perform policy checks.

The default implementation of this method first retrieves the permissions returned via getPermissions(CodeSource) (the CodeSource is taken from the specified ProtectionDomain), as well as the permissions located inside the specified ProtectionDomain. All of these permissions are then combined and returned in a new PermissionCollection object. If getPermissions(CodeSource) returns Policy.UNSUPPORTED_EMPTY_COLLECTION, then this method returns the permissions contained inside the specified ProtectionDomain in a new PermissionCollection object.

This method can be overridden if the policy implementation supports returning a set of permissions granted to a ProtectionDomain.

param
domain the ProtectionDomain to which the returned PermissionCollection has been granted.
return
a set of permissions granted to the specified ProtectionDomain. If this operation is supported, the returned set of permissions must be a new mutable instance and it must support heterogeneous Permission types. If this operation is not supported, Policy.UNSUPPORTED_EMPTY_COLLECTION is returned.
since
1.4

	PermissionCollection pc = null;

	if (domain == null)
	    return new Permissions();

	if (pdMapping == null) {
	    initPolicy(this);
	}

	synchronized (pdMapping) {
	    pc = (PermissionCollection)pdMapping.get(domain);
	}

	if (pc != null) {
	    Permissions perms = new Permissions();
	    synchronized (pc) {
		for (Enumeration e = pc.elements() ; e.hasMoreElements() ;) {
		    perms.add((Permission)e.nextElement());
		}
	    }
	    return perms;
	}

	pc = getPermissions(domain.getCodeSource());
	if (pc == null || pc == UNSUPPORTED_EMPTY_COLLECTION) {
	    pc = new Permissions();
	}

	addStaticPerms(pc, domain.getPermissions());
	return pc;
    
public static java.security.PolicygetPolicy()
Returns the installed Policy object. This value should not be cached, as it may be changed by a call to setPolicy. This method first calls SecurityManager.checkPermission with a SecurityPermission("getPolicy") permission to ensure it's ok to get the Policy object..

return
the installed Policy.
throws
SecurityException if a security manager exists and its checkPermission method doesn't allow getting the Policy object.
see
SecurityManager#checkPermission(Permission)
see
#setPolicy(java.security.Policy)

        SecurityManager sm = System.getSecurityManager();
        if (sm != null)
	    sm.checkPermission(SecurityConstants.GET_POLICY_PERMISSION);
	return getPolicyNoCheck();
    
static synchronized java.security.PolicygetPolicyNoCheck()
Returns the installed Policy object, skipping the security check. Used by SecureClassLoader and getPolicy.

return
the installed Policy.

	if (policy == null) {
	    String policy_class = null;
	    policy_class = (String)AccessController.doPrivileged(
                new PrivilegedAction() {
		    public Object run() {
			return Security.getProperty("policy.provider");
		    }
		});
	    if (policy_class == null) {
		policy_class = "sun.security.provider.PolicyFile";
	    }

	    /**
	     * Install a bootstrap (sandbox) policy to avoid recursion
	     * while the configured policy implementation initializes itself.
	     * After the configured implementation loads, install it over
	     * the bootstrap policy.
	     */
	    policy = new sun.security.provider.PolicyFile(true);

	    try {
		policy = (Policy)
		    Class.forName(policy_class).newInstance();
	    } catch (Exception e) {

	        // policy_class seems to be an extension 
			
		final String pc = policy_class;
		Policy p = (Policy)
		    AccessController.doPrivileged(new PrivilegedAction() {
			public Object run() {
			    try {
				ClassLoader cl =
					ClassLoader.getSystemClassLoader();
				// we want the extension loader 
				ClassLoader extcl = null;
				while (cl != null) {
				    extcl = cl;
				    cl = cl.getParent();
				} 
				return (extcl != null? Class.forName
					(pc, true, extcl).newInstance():
					null);
			    } catch (Exception e) {
				if (debug != null) {
				    debug.println("policy provider " +
						pc +
						" not available");
				    e.printStackTrace();
				}
				return null;
			    }
			}
		    });
		/*
		 * if it loaded install it as the policy provider. Otherwise
	         * continue to use the bootstrap implementation
		 */
		if (p != null) {
		    policy = p;
		} else {
		    if (debug != null) {
			debug.println("using sun.security.provider.PolicyFile");
		    }
		}
	    }
	}
	return policy;
    
public java.security.ProvidergetProvider()
Return the Provider of this Policy.

This Policy instance will only have a Provider if it was obtained via a call to Policy.getInstance. Otherwise this method returns null.

return
the Provider of this Policy, or null.
since
1.6

	return null;
    
public java.lang.StringgetType()
Return the type of this Policy.

This Policy instance will only have a type if it was obtained via a call to Policy.getInstance. Otherwise this method returns null.

return
the type of this Policy, or null.
since
1.6

	return null;
    
private static java.security.PolicyhandleException(java.security.NoSuchAlgorithmException nsae)

        Throwable cause = nsae.getCause();
        if (cause instanceof IllegalArgumentException) {
            throw (IllegalArgumentException)cause;
        }
        throw nsae;
    
public booleanimplies(java.security.ProtectionDomain domain, java.security.Permission permission)
Evaluates the global policy for the permissions granted to the ProtectionDomain and tests whether the permission is granted.

param
domain the ProtectionDomain to test
param
permission the Permission object to be tested for implication.
return
true if "permission" is a proper subset of a permission granted to this ProtectionDomain.
see
java.security.ProtectionDomain
since
1.4

	PermissionCollection pc;

	if (pdMapping == null) {
	    initPolicy(this);
	}

	synchronized (pdMapping) {
	    pc = (PermissionCollection)pdMapping.get(domain);
	}

	if (pc != null) {
	    return pc.implies(permission);
	} 
	
	pc = getPermissions(domain);
	if (pc == null) {
	    return false;
	}

	synchronized (pdMapping) {
	    // cache it 
	    pdMapping.put(domain, pc);
	}
	
	return pc.implies(permission);
    
private static voidinitPolicy(java.security.Policy p)
Initialize superclass state such that a legacy provider can handle queries for itself.

since
1.4

	/*
	 * A policy provider not on the bootclasspath could trigger
	 * security checks fulfilling a call to either Policy.implies
	 * or Policy.getPermissions. If this does occur the provider
	 * must be able to answer for it's own ProtectionDomain
	 * without triggering additional security checks, otherwise
	 * the policy implementation will end up in an infinite
	 * recursion.
	 * 
	 * To mitigate this, the provider can collect it's own
	 * ProtectionDomain and associate a PermissionCollection while
	 * it is being installed. The currently installed policy
	 * provider (if there is one) will handle calls to
	 * Policy.implies or Policy.getPermissions during this
	 * process.
	 * 
	 * This Policy superclass caches away the ProtectionDomain and
	 * statically binds permissions so that legacy Policy 
	 * implementations will continue to function.
	 */

	ProtectionDomain policyDomain = (ProtectionDomain)
		AccessController.doPrivileged(new PrivilegedAction() {
		    public Object run() {
			return p.getClass().getProtectionDomain();
		    }
		});

	/*
	 * Collect the permissions granted to this protection domain
	 * so that the provider can be security checked while processing
	 * calls to Policy.implies or Policy.getPermissions.
	 */
	PermissionCollection policyPerms = null;
	synchronized (p) {
	   if (p.pdMapping == null) {
	       p.pdMapping = new WeakHashMap();
	   }
	}

	if (policyDomain.getCodeSource() != null) {
	    if (Policy.isSet()) {
		policyPerms = policy.getPermissions(policyDomain);
	    }

	    if (policyPerms == null) { // assume it has all
		policyPerms = new Permissions();
		policyPerms.add(SecurityConstants.ALL_PERMISSION);
	    }

	    synchronized (p.pdMapping) {
		// cache of pd to permissions
		p.pdMapping.put(policyDomain, policyPerms);
	    }
	}
	return;
    
static booleanisSet()
package private for AccessControlContext


         
      
    
	return policy != null;
    
public voidrefresh()
Refreshes/reloads the policy configuration. The behavior of this method depends on the implementation. For example, calling refresh on a file-based policy will cause the file to be re-read.

The default implementation of this method does nothing. This method should be overridden if a refresh operation is supported by the policy implementation.

 
public static voidsetPolicy(java.security.Policy p)
Sets the system-wide Policy object. This method first calls SecurityManager.checkPermission with a SecurityPermission("setPolicy") permission to ensure it's ok to set the Policy.

param
p the new system Policy object.
throws
SecurityException if a security manager exists and its checkPermission method doesn't allow setting the Policy.
see
SecurityManager#checkPermission(Permission)
see
#getPolicy()

	SecurityManager sm = System.getSecurityManager();
	if (sm != null) sm.checkPermission(
				 new SecurityPermission("setPolicy"));
	if (p != null) {
	    initPolicy(p);
	}
	synchronized (Policy.class) {
	    Policy.policy = p;
	}