Chapter 80. Execution control in the decision engine

Each rule has an integer salience attribute that determines the order of execution. Rules with a higher salience value are given higher priority when ordered in the activation queue. The default salience value for rules is zero, but the salience can be negative or positive.

For example, the following sample DRL rules are listed in the decision engine stack in the order shown:

rule "RuleA"
salience 95
when
    $fact : MyFact( field1 == true )
then
    System.out.println("Rule2 : " + $fact);
    update($fact);
end

rule "RuleB"
salience 100
when
   $fact : MyFact( field1 == false )
then
   System.out.println("Rule1 : " + $fact);
   $fact.setField1(true);
   update($fact);
end

The RuleB rule is listed second, but it has a higher salience value than the RuleA rule and is therefore executed first.

An agenda group is a set of rules bound together by the same agenda-group rule attribute. Agenda groups partition rules on the decision engine agenda. At any one time, only one group has a focus that gives that group of rules priority for execution before rules in other agenda groups. You determine the focus with a setFocus() call for the agenda group. You can also define rules with an auto-focus attribute so that the next time the rule is activated, the focus is automatically given to the entire agenda group to which the rule is assigned.

Each time the setFocus() call is made in a Java application, the decision engine adds the specified agenda group to the top of the rule stack. The default agenda group "MAIN" contains all rules that do not belong to a specified agenda group and is executed first in the stack unless another group has the focus.

For example, the following sample DRL rules belong to specified agenda groups and are listed in the decision engine stack in the order shown:

rule "Increase balance for credits"
  agenda-group "calculation"
when
  ap : AccountPeriod()
  acc : Account( $accountNo : accountNo )
  CashFlow( type == CREDIT,
            accountNo == $accountNo,
            date >= ap.start && <= ap.end,
            $amount : amount )
then
  acc.balance  += $amount;
end

rule "Print balance for AccountPeriod"
  agenda-group "report"
when
  ap : AccountPeriod()
  acc : Account()
then
  System.out.println( acc.accountNo +
                      " : " + acc.balance );
end

For this example, the rules in the "report" agenda group must always be executed first and the rules in the "calculation" agenda group must always be executed second. Any remaining rules in other agenda groups can then be executed. Therefore, the "report" and "calculation" groups must receive the focus to be executed in that order, before other rules can be executed:

Agenda agenda = ksession.getAgenda();
agenda.getAgendaGroup( "report" ).setFocus();
agenda.getAgendaGroup( "calculation" ).setFocus();
ksession.fireAllRules();

You can also use the clear() method to cancel all the activations generated by the rules belonging to a given agenda group before each has had a chance to be executed:

ksession.getAgenda().getAgendaGroup( "Group A" ).clear();

An activation group is a set of rules bound together by the same activation-group rule attribute. In this group, only one rule can be executed. After conditions are met for a rule in that group to be executed, all other pending rule executions from that activation group are removed from the agenda.

For example, the following sample DRL rules belong to the specified activation group and are listed in the decision engine stack in the order shown:

rule "Print balance for AccountPeriod1"
  activation-group "report"
when
  ap : AccountPeriod1()
  acc : Account()
then
  System.out.println( acc.accountNo +
                      " : " + acc.balance );
end

rule "Print balance for AccountPeriod2"
  activation-group "report"
when
  ap : AccountPeriod2()
  acc : Account()
then
  System.out.println( acc.accountNo +
                      " : " + acc.balance );
end

For this example, if the first rule in the "report" activation group is executed, the second rule in the group and all other executable rules on the agenda are removed from the agenda.

The decision engine supports the following rule execution modes that determine how and when the decision engine executes rules:

Although you should avoid using both fireAllRules() and fireUntilHalt() calls, especially from different threads, the decision engine can handle such situations safely using thread-safety logic and an internal state machine. If a fireAllRules() call is in progress and you call fireUntilHalt(), the decision engine continues to run in passive mode until the fireAllRules() operation is complete and then starts in active mode in response to the fireUntilHalt() call. However, if the decision engine is running in active mode following a fireUntilHalt() call and you call fireAllRules(), the fireAllRules() call is ignored and the decision engine continues to run in active mode until you call halt().

For added thread safety in active mode, the decision engine supports a submit() method that you can use to group and perform operations on a KIE session in a thread-safe, atomic action:

KieSession session = ...;

new Thread( new Runnable() {
  @Override
  public void run() {
      session.fireUntilHalt();
  }
} ).start();

final FactHandle fh = session.insert( fact_a );

... Thread.sleep( 1000L ); ...

session.submit( new KieSession.AtomicAction() {
  @Override
  public void execute( KieSession kieSession ) {
    fact_a.setField("value");
    kieSession.update( fh, fact_a );
    kieSession.insert( fact_1 );
    kieSession.insert( fact_2 );
    kieSession.insert( fact_3 );
  }
} );

... Thread.sleep( 1000L ); ...

session.insert( fact_z );

session.halt();
session.dispose();

Thread safety and atomic operations are also helpful from a client-side perspective. For example, you might need to insert more than one fact at a given time, but require the decision engine to consider the insertions as an atomic operation and to wait until all the insertions are complete before evaluating the rules again.

The decision engine supports the following fact propagation modes that determine how the decision engine progresses inserted facts through the engine network in preparation for rule execution:

By default, the Phreak rule algorithm in the decision engine uses lazy fact propagation for improved rule evaluation overall. However, in few cases, this lazy propagation behavior can alter the expected result of certain rule executions that may require immediate or eager propagation.

For example, the following rule uses a specified query with a ? prefix to invoke the query in pull-only or passive fashion:

query Q (Integer i)
    String( this == i.toString() )
end

rule "Rule"
  when
    $i : Integer()
    ?Q( $i; )
  then
    System.out.println( $i );
end

For this example, the rule should be executed only when a String that satisfies the query is inserted before the Integer, such as in the following example commands:

KieSession ksession = ...
ksession.insert("1");
ksession.insert(1);
ksession.fireAllRules();

However, due to the default lazy propagation behavior in Phreak, the decision engine does not detect the insertion sequence of the two facts in this case, so this rule is executed regardless of String and Integer insertion order. For this example, immediate propagation is required for the expected rule evaluation.

To alter the decision engine propagation mode to achieve the expected rule evaluation in this case, you can add the @Propagation(<type>) tag to your rule and set <type> to LAZY, IMMEDIATE, or EAGER.

In the same example rule, the immediate propagation annotation enables the rule to be evaluated only when a String that satisfies the query is inserted before the Integer, as expected:

query Q (Integer i)
    String( this == i.toString() )
end

rule "Rule" @Propagation(IMMEDIATE)
  when
    $i : Integer()
    ?Q( $i; )
  then
    System.out.println( $i );
end

The decision engine supports an AgendaFilter object in the filter interface that you can use to allow or deny the evaluation of specified rules during agenda evaluation. You can specify an agenda filter as part of a fireAllRules() call.

The following example code permits only rules ending with the string "Test" to be evaluated and executed. All other rules are filtered out of the decision engine agenda.

ksession.fireAllRules( new RuleNameEndsWithAgendaFilter( "Test" ) );

Rule units are groups of data sources, global variables, and DRL rules that function together for a specific purpose. You can use rule units to partition a rule set into smaller units, bind different data sources to those units, and then execute the individual unit. Rule units are an enhanced alternative to rule-grouping DRL attributes such as rule agenda groups or activation groups for execution control.

Rule units are helpful when you want to coordinate rule execution so that the complete execution of one rule unit triggers the start of another rule unit and so on. For example, assume that you have a set of rules for data enrichment, another set of rules that processes that data, and another set of rules that extract the output from the processed data. If you add these rule sets into three distinct rule units, you can coordinate those rule units so that complete execution of the first unit triggers the start of the second unit and the complete execution of the second unit triggers the start of third unit.

To define a rule unit, implement the RuleUnit interface as shown in the following example:

package org.mypackage.myunit;

public static class AdultUnit implements RuleUnit {
    private int adultAge;
    private DataSource<Person> persons;

    public AdultUnit( ) { }

    public AdultUnit( DataSource<Person> persons, int age ) {
        this.persons = persons;
        this.age = age;
    }

    // A data source of `Persons` in this rule unit:
    public DataSource<Person> getPersons() {
        return persons;
    }

    // A global variable in this rule unit:
    public int getAdultAge() {
        return adultAge;
    }

    // Life-cycle methods:
    @Override
    public void onStart() {
        System.out.println("AdultUnit started.");
    }

    @Override
    public void onEnd() {
        System.out.println("AdultUnit ended.");
    }
}

In this example, persons is a source of facts of type Person. A rule unit data source is a source of the data processed by a given rule unit and represents the entry point that the decision engine uses to evaluate the rule unit. The adultAge global variable is accessible from all the rules belonging to this rule unit. The last two methods are part of the rule unit life cycle and are invoked by the decision engine.

The decision engine supports the following optional life-cycle methods for rule units:

You can add one or more rules to a rule unit. By default, all the rules in a DRL file are automatically associated with a rule unit that follows the naming convention of the DRL file name. If the DRL file is in the same package and has the same name as a class that implements the RuleUnit interface, then all of the rules in that DRL file implicitly belong to that rule unit. For example, all the rules in the AdultUnit.drl file in the org.mypackage.myunit package are automatically part of the rule unit org.mypackage.myunit.AdultUnit.

To override this naming convention and explicitly declare the rule unit that the rules in a DRL file belong to, use the unit keyword in the DRL file. The unit declaration must immediately follow the package declaration and contain the name of the class in that package that the rules in the DRL file are part of.

package org.mypackage.myunit
unit AdultUnit

rule Adult
  when
    $p : Person(age >= adultAge) from persons
  then
    System.out.println($p.getName() + " is adult and greater than " + adultAge);
end

You can also rewrite the same pattern in a more convenient way using OOPath notation, as shown in the following example:

package org.mypackage.myunit
unit AdultUnit

rule Adult
  when
    $p : /persons[age >= adultAge]
  then
    System.out.println($p.getName() + " is adult and greater than " + adultAge);
end

In this example, any matching facts in the rule conditions are retrieved from the persons data source defined in the DataSource definition in the rule unit class. The rule condition and action use the adultAge variable in the same way that a global variable is defined at the DRL file level.

To execute one or more rule units defined in a KIE base, create a new RuleUnitExecutor class bound to the KIE base, create the rule unit from the relevant data source, and run the rule unit executer:

// Create a `RuleUnitExecutor` class and bind it to the KIE base:
KieBase kbase = kieContainer.getKieBase();
RuleUnitExecutor executor = RuleUnitExecutor.create().bind( kbase );

// Create the `AdultUnit` rule unit using the `persons` data source and run the executor:
RuleUnit adultUnit = new AdultUnit(persons, 18);
executor.run( adultUnit );

Rules are executed by the RuleUnitExecutor class. The RuleUnitExecutor class creates KIE sessions and adds the required DataSource objects to those sessions, and then executes the rules based on the RuleUnit that is passed as a parameter to the run() method.

The example execution code produces the following output when the relevant Person facts are inserted in the persons data source:

org.mypackage.myunit.AdultUnit started.
Jane is adult and greater than 18
John is adult and greater than 18
org.mypackage.myunit.AdultUnit ended.

Instead of explicitly creating the rule unit instance, you can register the rule unit variables in the executor and pass to the executor the rule unit class that you want to run, and then the executor creates an instance of the rule unit. You can then set the DataSource definition and other variables as needed before running the rule unit.

executor.bindVariable( "persons", persons );
        .bindVariable( "adultAge", 18 );
executor.run( AdultUnit.class );

The name that you pass to the RuleUnitExecutor.bindVariable() method is used at run time to bind the variable to the field of the rule unit class with the same name. In the previous example, the RuleUnitExecutor inserts into the new rule unit the data source bound to the "persons" name and inserts the value 18 bound to the String "adultAge" into the fields with the corresponding names inside the AdultUnit class.

To override this default variable-binding behavior, use the @UnitVar annotation to explicitly define a logical binding name for each field of the rule unit class. For example, the field bindings in the following class are redefined with alternative names:

package org.mypackage.myunit;

public static class AdultUnit implements RuleUnit {
    @UnitVar("minAge")
    private int adultAge = 18;

    @UnitVar("data")
    private DataSource<Person> persons;
}

You can then bind the variables to the executor using those alternative names and run the rule unit:

executor.bindVariable( "data", persons );
        .bindVariable( "minAge", 18 );
executor.run( AdultUnit.class );

You can execute a rule unit in passive mode by using the run() method (equivalent to invoking fireAllRules() on a KIE session) or in active mode using the runUntilHalt() method (equivalent to invoking fireUntilHalt() on a KIE session). By default, the decision engine runs in passive mode and evaluates rule units only when a user or an application explicitly calls run() (or fireAllRules() for standard rules). If a user or application calls runUntilHalt() for rule units (or fireUntilHalt() for standard rules), the decision engine starts in active mode and evaluates rule units continually until the user or application explicitly calls halt().

If you use the runUntilHalt() method, invoke the method on a separate execution thread to avoid blocking the main thread:

new Thread( () -> executor.runUntilHalt( adultUnit ) ).start();

DataSource<Person> persons = DataSource.create( new Person( "John", 42 ),
                                                new Person( "Jane", 44 ),
                                                new Person( "Sally", 4 ) );

// Insert a fact:
Person john = new Person( "John", 42 );
FactHandle johnFh = persons.insert( john );

// Modify the fact and optionally specify modified properties (for property reactivity):
john.setAge( 43 );
persons.update( johnFh, john, "age" );

// Delete the fact:
persons.delete( johnFh );

package org.mypackage.myunit
unit AdultUnit

rule Adult
  when
    Person(age >= 18, $name : name) from persons
  then
    System.out.println($name + " is adult");
end

package org.mypackage.myunit
unit NotAdultUnit

rule NotAdult
  when
    $p : Person(age < 18, $name : name) from persons
  then
    System.out.println($name + " is NOT adult");
    modify($p) { setAge(18); }
    drools.run( AdultUnit.class );
end

RuleUnitExecutor executor = RuleUnitExecutor.create().bind( kbase );
DataSource<Person> persons = executor.newDataSource( "persons",
                                                     new Person( "John", 42 ),
                                                     new Person( "Jane", 44 ),
                                                     new Person( "Sally", 4 ) );

Sally is NOT adult
John is adult
Jane is adult
Sally is adult

public class BoxOffice {
    private boolean open;

    public BoxOffice( boolean open ) {
        this.open = open;
    }

    public boolean isOpen() {
        return open;
    }

    public void setOpen( boolean open ) {
        this.open = open;
    }
}

public class BoxOfficeUnit implements RuleUnit {
    private DataSource<BoxOffice> boxOffices;

    public DataSource<BoxOffice> getBoxOffices() {
        return boxOffices;
    }
}

public class TicketIssuerUnit implements RuleUnit {
    private DataSource<Person> persons;
    private DataSource<AdultTicket> tickets;

    private List<String> results;

    public TicketIssuerUnit() { }

    public TicketIssuerUnit( DataSource<Person> persons, DataSource<AdultTicket> tickets ) {
        this.persons = persons;
        this.tickets = tickets;
    }

    public DataSource<Person> getPersons() {
        return persons;
    }

    public DataSource<AdultTicket> getTickets() {
        return tickets;
    }

    public List<String> getResults() {
        return results;
    }
}

package org.mypackage.myunit;
unit TicketIssuerUnit;

rule IssueAdultTicket when
    $p: /persons[ age >= 18 ]
then
    tickets.insert(new AdultTicket($p));
end
rule RegisterAdultTicket when
    $t: /tickets
then
    results.add( $t.getPerson().getName() );
end

package org.mypackage.myunit;
unit BoxOfficeUnit;

rule BoxOfficeIsOpen
  when
    $box: /boxOffices[ open ]
  then
    drools.guard( TicketIssuerUnit.class );
end

DataSource<Person> persons = executor.newDataSource( "persons" );
DataSource<BoxOffice> boxOffices = executor.newDataSource( "boxOffices" );
DataSource<AdultTicket> tickets = executor.newDataSource( "tickets" );

List<String> list = new ArrayList<>();
executor.bindVariable( "results", list );

// Two box offices are open:
BoxOffice office1 = new BoxOffice(true);
FactHandle officeFH1 = boxOffices.insert( office1 );
BoxOffice office2 = new BoxOffice(true);
FactHandle officeFH2 = boxOffices.insert( office2 );

persons.insert(new Person("John", 40));

// Execute `BoxOfficeIsOpen` rule, run `TicketIssuerUnit` rule unit, and execute `RegisterAdultTicket` rule:
executor.run(BoxOfficeUnit.class);

assertEquals( 1, list.size() );
assertEquals( "John", list.get(0) );
list.clear();

persons.insert(new Person("Matteo", 30));

// Execute `RegisterAdultTicket` rule:
executor.run(BoxOfficeUnit.class);

assertEquals( 1, list.size() );
assertEquals( "Matteo", list.get(0) );
list.clear();

// One box office is closed, the other is open:
office1.setOpen(false);
boxOffices.update(officeFH1, office1);
persons.insert(new Person("Mark", 35));
executor.run(BoxOfficeUnit.class);

assertEquals( 1, list.size() );
assertEquals( "Mark", list.get(0) );
list.clear();

// All box offices are closed:
office2.setOpen(false);
boxOffices.update(officeFH2, office2); // Guarding rule is no longer true.
persons.insert(new Person("Edson", 35));
executor.run(BoxOfficeUnit.class); // No execution

assertEquals( 0, list.size() );

default Identity getUnitIdentity() {
    return new Identity( getClass() );
}

public class Unit0 implements RuleUnit {
    private DataSource<Object> input;

    public DataSource<Object> getInput() {
        return input;
    }
}

package org.mypackage.myunit
unit Unit0

rule GuardAgeCheck
  when
    $i: /input#Integer
    $s: /input#String
  then
    drools.guard( new AgeCheckUnit($i) );
    drools.guard( new AgeCheckUnit($s.length()) );
end

public class AgeCheckUnit implements RuleUnit {
    private final int minAge;
    private DataSource<Person> persons;
    private List<String> results;

    public AgeCheckUnit( int minAge ) {
        this.minAge = minAge;
    }

    public DataSource<Person> getPersons() {
        return persons;
    }

    public int getMinAge() {
        return minAge;
    }

    public List<String> getResults() {
        return results;
    }
}

package org.mypackage.myunit
unit AgeCheckUnit

rule CheckAge
  when
    $p : /persons{ age > minAge }
  then
    results.add($p.getName() + ">" + minAge);
end

RuleUnitExecutor executor = RuleUnitExecutor.create().bind( kbase );

DataSource<Object> input = executor.newDataSource( "input" );
DataSource<Person> persons = executor.newDataSource( "persons",
                                                     new Person( "John", 42 ),
                                                     new Person( "Sally", 4 ) );

List<String> results = new ArrayList<>();
executor.bindVariable( "results", results );

ds.insert("test");
ds.insert(3);
ds.insert(4);
executor.run(Unit0.class);

[Sally>3, John>3]

public class AgeCheckUnit implements RuleUnit {

    ...

    @Override
    public Identity getUnitIdentity() {
        return new Identity(getClass(), minAge);
    }
}

[John>4, Sally>3, John>3]