规则引擎：Drools与JRuleEngine

 

规则引擎：Drools与JRuleEngine

译文：《规则引擎：Drools与JRuleEngine》
译者：chszs
作者：Vivek Tomar
原文：《Rule engine : Drools / JRuleEngine》
原文见http://www.naxos-software.de/blog/index.php?/archives/78-Rule-engine-Drools-JRuleEngine.html

一、规则引擎
规则引擎有助于基于存储在规则中的知识和推理来执行判断。这些规则基本上只有条件和动作，别无它物。

规则引擎的优点：
1、分隔应用程序的条件和控制流
(1) 规则都存储在单独的文件。
(2) 规则可以被技术人士和商业人士修改的。
(3) 规则改变后应用程序不必重新部署。
(4) 使用集中的规则使得应用程序更易于管理和维护。
 
2、规则替换了代码中的if else语句
(1) 规则脚本更容易提取。
(2) 即使是非技术人员也能轻易地遵循规则。
(3) 集中可以解决问题，而不是实现。
(4) 与实现代码相比，规则聚团更容易编写。 

3、为什么作决定能很容易地概念化

总结：规则有助于消除代码中大量的if else语句，使代码更易于维护。

二、Drools介绍
Drools是一个开源实现。它是一个Java库，以Apache许可证发布，其二进制代码或源码下载均有效。

推理机的核心是一个规则引擎。它执行模式匹配，在执行动作中做出决策。RETE算法用于模式匹配。 

知识被表述为规则。规则有两个主要部分：条件和动作。

例如：
如果用户（年龄>17），那么System.out.println("User is greater then 17");

在人工智能系统，主要有两种响应的方法。
1、正向链（Forward Chaining）
这是基于事实根据的。在工作区域检查中规则。当规则条件为真的规则不再有时，模式匹配结束。

2、反向链（Backward Chaining）
只检查规则的动作可以匹配目标的规则。如果满足条件，然后进行评估。

3、兼容JDK1.4，且需要下面的库。
(1) drools-all-jdk1.4.2.1.jar
(2) xercesImpl-2.6.2.jar
(3) antlr-2.7.5.jar
(4) janino-2.3.2.jar

4、代码示例（ApplyRule.java）

 

/*
  /* $Header$
  */

  package com.vivek.drools.example;

  import java.io.IOException;
  import java.io.InputStream;
  import org.apache.commons.logging.Log;
  import org.apache.commons.logging.LogFactory;
  import org.drools.FactException;
  import org.drools.IntegrationException;
  import org.drools.RuleBase;
  import org.drools.WorkingMemory;
  import org.drools.io.RuleBaseLoader;
  import org.xml.sax.SAXException;

  /**
  * Demonstration of a sample rule using Java.
  *
  * @version <tt>$Revision: 1.0 $</tt>
  * @author <a href="mailto:{[email protected]}" mce_href="mailto:{[email protected]}">{Vivek Tomar}</a>.
  */

  public class ApplyRule {

  // Constants -----------------------------------------------------

  static Log log = LogFactory.getLog(ApplyRule.class.getName());
  private static final String RULE_FILE = "/rules/rules.drl";
  //private static final String RULE_IS_TURNOVER_INSURABLE = "turnoverInsurable";
  //private static final String RULE_IS_TURNOVER_NOT_INSURABLE = "turnoverNotInsurable";

  // Attributes ----------------------------------------------------

  private double turnover = 1000000;

  // Constructors --------------------------------------------------

  public ApplyRule () {
  try {
  InputStream in = this.getClass().getResourceAsStream(RULE_FILE);
  RuleBase ruleBase = RuleBaseLoader.loadFromInputStream(in);
  WorkingMemory workingMemory = ruleBase.newWorkingMemory();
  workingMemory.assertObject(this);
  // Fire specific rule
  //workingMemory.fireAllRules(new RuleNameEqualsAgendaFilter(RULE_IS_TURNOVER_INSURABLE));

  //Fire all rules
  workingMemory.fireAllRules();
  } catch (IntegrationException e) {
  e.printStackTrace();
  } catch (FactException e) {
  e.printStackTrace();
  } catch (SAXException e) {
  e.printStackTrace();
  } catch (IOException e) {
  e.printStackTrace();
  }
  }

  // Public --------------------------------------------------------

  public static void main (String args[]) {
  new ApplyRule();
  }

  public double getTurnover () {
  return turnover;
  }

  public void setTurnover (double turnover) {
  this.turnover = turnover;
  }

  public void printCompanyInsurable () {
  log.info("******************************");
  log.info("This company is Insurable.");
  log.info("******************************");
  }

  public void printCompanyNotInsurable () {
  log.info("==============================");
  log.info("This company is not Insurable.");
  log.info("==============================");
  }
  }

/* /* $Header$ */ package com.vivek.drools.example; import java.io.IOException; import java.io.InputStream; import org.apache.commons.logging.Log; import org.apache.commons.logging.LogFactory; import org.drools.FactException; import org.drools.IntegrationException; import org.drools.RuleBase; import org.drools.WorkingMemory; import org.drools.io.RuleBaseLoader; import org.xml.sax.SAXException; /** * Demonstration of a sample rule using Java. * * @version <tt>$Revision: 1.0 $</tt> * @author <a href="mailto:{[email protected]}" mce_href="mailto:{[email protected]}">{Vivek Tomar}</a>. */ public class ApplyRule { // Constants ----------------------------------------------------- static Log log = LogFactory.getLog(ApplyRule.class.getName()); private static final String RULE_FILE = "/rules/rules.drl"; //private static final String RULE_IS_TURNOVER_INSURABLE = "turnoverInsurable"; //private static final String RULE_IS_TURNOVER_NOT_INSURABLE = "turnoverNotInsurable"; // Attributes ---------------------------------------------------- private double turnover = 1000000; // Constructors -------------------------------------------------- public ApplyRule () { try { InputStream in = this.getClass().getResourceAsStream(RULE_FILE); RuleBase ruleBase = RuleBaseLoader.loadFromInputStream(in); WorkingMemory workingMemory = ruleBase.newWorkingMemory(); workingMemory.assertObject(this); // Fire specific rule //workingMemory.fireAllRules(new RuleNameEqualsAgendaFilter(RULE_IS_TURNOVER_INSURABLE)); //Fire all rules workingMemory.fireAllRules(); } catch (IntegrationException e) { e.printStackTrace(); } catch (FactException e) { e.printStackTrace(); } catch (SAXException e) { e.printStackTrace(); } catch (IOException e) { e.printStackTrace(); } } // Public -------------------------------------------------------- public static void main (String args[]) { new ApplyRule(); } public double getTurnover () { return turnover; } public void setTurnover (double turnover) { this.turnover = turnover; } public void printCompanyInsurable () { log.info("******************************"); log.info("This company is Insurable."); log.info("******************************"); } public void printCompanyNotInsurable () { log.info("=============================="); log.info("This company is not Insurable."); log.info("=============================="); } }

规则定义（rules.drl）

 

 

  <?xml version="1.0"?>
<rule-set name="cheese rules"
  xmlns="http://drools.org/rules"
  xmlns:java="http://drools.org/semantics/java"
  xmlns:xs="http://www.w3.org/2001/XMLSchema-instance"
  xs:schemaLocation="http://drools.org/rules rules.xsd
  http://drools.org/semantics/java java.xsd">
<rule name="turnoverInsurable">
<parameter identifier="applyRule">
<class>com.vivek.drools.example.ApplyRule</class>
</parameter>
<java:condition>applyRule.getTurnover() > 2000000 </java:condition>
<java:consequence>
  applyRule.printCompanyInsurable();
</java:consequence>
</rule>
<rule name="turnoverNotInsurable">
<parameter identifier="applyRule">
<class>com.vivek.drools.example.ApplyRule</class>
</parameter>
<java:condition>applyRule.getTurnover() < 2000000 </java:condition>
<java:consequence>
  applyRule.printCompanyNotInsurable();
</java:consequence>
</rule>
</rule-set>

<?xml version="1.0"?> <rule-set name="cheese rules" xmlns="http://drools.org/rules" xmlns:java="http://drools.org/semantics/java" xmlns:xs="http://www.w3.org/2001/XMLSchema-instance" xs:schemaLocation="http://drools.org/rules rules.xsd http://drools.org/semantics/java java.xsd"> <rule name="turnoverInsurable"> <parameter identifier="applyRule"> <class>com.vivek.drools.example.ApplyRule</class> </parameter> <java:condition>applyRule.getTurnover() > 2000000 </java:condition> <java:consequence> applyRule.printCompanyInsurable(); </java:consequence> </rule> <rule name="turnoverNotInsurable"> <parameter identifier="applyRule"> <class>com.vivek.drools.example.ApplyRule</class> </parameter> <java:condition>applyRule.getTurnover() < 2000000 </java:condition> <java:consequence> applyRule.printCompanyNotInsurable(); </java:consequence> </rule> </rule-set>

三、JRuleEngine介绍
JRuleEngine是一个基于Java的规则引擎，遵循JSR94规范，版本1.1。

1、JRuleEngine共同特征： 
(1) 在JRuleEngine输入对象被称为事实，而输出对象被称为结论。 
(2) 一个类的方法可直接从规则中调用。 

2、JRuleEngine基于前向链算法。

3、规则是类org.jruleengine.rule.RuleImpl的对象，它们以下列方式载入：
(1) 可以从XML文件中读入
(2) 通过RuleImpl对象创建，也可以取自数据库。

4、会话是客户端和规则引擎之间的运行时的胶水。会话与单一的规则执行集相关。会话规则的类型：
(1) 一个有状态的规则会话可持续很长时间，可一次又一次地查询。
(2) 一个无状态的规则给出了实现，但只能持续一定时期。

5、JRuleEngine需要两个库：
(1) jsr94.jar
(2) jruleengine.jar

6、添加这些文件到您应用程序的类路径。

7、通过实例化一个有状态的规则会话（StatefulRuleSession）或无状态的规则会话（StatelessRuleSession）来使用这个库。

8、代码例子(ApplyRule.java)

 

 

/*
* $Header$
*/
package com.vivek.jruleengine.example;

import java.io.InputStream;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;

import javax.rules.RuleRuntime;
import javax.rules.RuleServiceProvider;
import javax.rules.RuleServiceProviderManager;
import javax.rules.StatelessRuleSession;
import javax.rules.admin.RuleAdministrator;
import javax.rules.admin.RuleExecutionSet;

import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;

/**
* Demonstration of a sample rule using Java.
*
* @version <tt>$Revision: 1.0 $</tt>
* @author <a href="mailto:{[email protected]}" mce_href="mailto:{[email protected]}">{Vivek Tomar}</a>.
*/
public class ApplyRule {

// Constants -----------------------------------------------------

static Log log = LogFactory.getLog(ApplyRule.class.getName());
private static final String RULE_FILE = "/rules/rules.xml";

// Attributes ----------------------------------------------------

private double turnover = 1000000;

// Constructors --------------------------------------------------

public ApplyRule () {

try {
// Load the rule service provider of the reference
// implementation.
// Loading this class will automatically register this
// provider with the provider manager.
Class.forName("org.jruleengine.RuleServiceProviderImpl");

// Get the rule service provider from the provider manager.
RuleServiceProvider serviceProvider = RuleServiceProviderManager.getRuleServiceProvider("org.jruleengine");

// get the RuleAdministrator
RuleAdministrator ruleAdministrator = serviceProvider.getRuleAdministrator();
log.info("Administration API");
log.info("======================");
log.info("Acquired RuleAdministrator: ");

// get an input stream to a test XML ruleset
// This rule execution set is part of the TCK.
InputStream inStream = this.getClass().getResourceAsStream(RULE_FILE);


// parse the ruleset from the XML document
RuleExecutionSet res1 =
ruleAdministrator.getLocalRuleExecutionSetProvider(null).createRuleExecutionSet(inStream, null);
inStream.close();
log.info("Loaded RuleExecutionSet: ");

// register the RuleExecutionSet
String uri = res1.getName();
ruleAdministrator.registerRuleExecutionSet(uri, res1, null);
log.info("Bound RuleExecutionSet to URI: " + uri);

RuleRuntime ruleRuntime = serviceProvider.getRuleRuntime();
log.info("Acquired RuleRuntime: ");

// create a StatelessRuleSession
StatelessRuleSession statelessRuleSession = (StatelessRuleSession)
ruleRuntime.createRuleSession(uri, new HashMap(),
RuleRuntime.STATELESS_SESSION_TYPE);

log.info("Got Stateless Rule Session: " + statelessRuleSession);

// call executeRules with some input objects

// Create a input list.
List input = new ArrayList();
input.add(this);

// Print the input.
log.info("Calling rule session with the following data");
log.info("Customer turnover input: " + this.getTurnover());

// Execute the rules without a filter.
List results = statelessRuleSession.executeRules(input);

// Release the session.
statelessRuleSession.release();
log.info("Released Stateless Rule Session.");

} catch (NoClassDefFoundError e) {
if (e.getMessage().indexOf("Exception") != -1) {
log.error("Error: The Rule Engine Implementation could not be found.");
} else {
log.error("Error: " + e.getMessage());
}
} catch (Exception e) {
e.printStackTrace();
}
}

// Public --------------------------------------------------------

public static void main (String args[]) {
new ApplyRule();
}

public double getTurnover () {
return turnover;
}

public void setTurnover (double turnover) {
this.turnover = turnover;
}

public void printCompanyInsurable () {
log.info("******************************");
log.info("This company is Insurable.");
log.info("******************************");
}

public void printCompanyNotInsurable () {
log.info("==============================");
log.info("This company is not Insurable.");
log.info("==============================");
}
}

/* * $Header$ */ package com.vivek.jruleengine.example; import java.io.InputStream; import java.util.ArrayList; import java.util.HashMap; import java.util.List; import javax.rules.RuleRuntime; import javax.rules.RuleServiceProvider; import javax.rules.RuleServiceProviderManager; import javax.rules.StatelessRuleSession; import javax.rules.admin.RuleAdministrator; import javax.rules.admin.RuleExecutionSet; import org.apache.commons.logging.Log; import org.apache.commons.logging.LogFactory; /** * Demonstration of a sample rule using Java. * * @version <tt>$Revision: 1.0 $</tt> * @author <a href="mailto:{[email protected]}" mce_href="mailto:{[email protected]}">{Vivek Tomar}</a>. */ public class ApplyRule { // Constants ----------------------------------------------------- static Log log = LogFactory.getLog(ApplyRule.class.getName()); private static final String RULE_FILE = "/rules/rules.xml"; // Attributes ---------------------------------------------------- private double turnover = 1000000; // Constructors -------------------------------------------------- public ApplyRule () { try { // Load the rule service provider of the reference // implementation. // Loading this class will automatically register this // provider with the provider manager. Class.forName("org.jruleengine.RuleServiceProviderImpl"); // Get the rule service provider from the provider manager. RuleServiceProvider serviceProvider = RuleServiceProviderManager.getRuleServiceProvider("org.jruleengine"); // get the RuleAdministrator RuleAdministrator ruleAdministrator = serviceProvider.getRuleAdministrator(); log.info("Administration API"); log.info("======================"); log.info("Acquired RuleAdministrator: "); // get an input stream to a test XML ruleset // This rule execution set is part of the TCK. InputStream inStream = this.getClass().getResourceAsStream(RULE_FILE); // parse the ruleset from the XML document RuleExecutionSet res1 = ruleAdministrator.getLocalRuleExecutionSetProvider(null).createRuleExecutionSet(inStream, null); inStream.close(); log.info("Loaded RuleExecutionSet: "); // register the RuleExecutionSet String uri = res1.getName(); ruleAdministrator.registerRuleExecutionSet(uri, res1, null); log.info("Bound RuleExecutionSet to URI: " + uri); RuleRuntime ruleRuntime = serviceProvider.getRuleRuntime(); log.info("Acquired RuleRuntime: "); // create a StatelessRuleSession StatelessRuleSession statelessRuleSession = (StatelessRuleSession) ruleRuntime.createRuleSession(uri, new HashMap(), RuleRuntime.STATELESS_SESSION_TYPE); log.info("Got Stateless Rule Session: " + statelessRuleSession); // call executeRules with some input objects // Create a input list. List input = new ArrayList(); input.add(this); // Print the input. log.info("Calling rule session with the following data"); log.info("Customer turnover input: " + this.getTurnover()); // Execute the rules without a filter. List results = statelessRuleSession.executeRules(input); // Release the session. statelessRuleSession.release(); log.info("Released Stateless Rule Session."); } catch (NoClassDefFoundError e) { if (e.getMessage().indexOf("Exception") != -1) { log.error("Error: The Rule Engine Implementation could not be found."); } else { log.error("Error: " + e.getMessage()); } } catch (Exception e) { e.printStackTrace(); } } // Public -------------------------------------------------------- public static void main (String args[]) { new ApplyRule(); } public double getTurnover () { return turnover; } public void setTurnover (double turnover) { this.turnover = turnover; } public void printCompanyInsurable () { log.info("******************************"); log.info("This company is Insurable."); log.info("******************************"); } public void printCompanyNotInsurable () { log.info("=============================="); log.info("This company is not Insurable."); log.info("=============================="); } }

 

规则定义(rules.xml)

 

  <?xml version="1.0" encoding="UTF-8"?>
<rule-execution-set>
<name>RuleExecutionSet1</name>
<description>Rule Execution Set</description>
<synonymn name="applyRule" class="com.vivek.jruleengine.example.ApplyRule" />
<rule name="turnoverInsurable" description="Check if turnover insurable" >
<if leftTerm="applyRule.getTurnover" op=">" rightTerm="2000000" />
<then method="applyRule.printCompanyInsurable()" />
</rule>
<rule name="turnoverNotInsurable" description="Check if turnover not insurable" >
<if leftTerm="applyRule.getTurnover" op="<" rightTerm="2000000" />
<then method="applyRule.printCompanyNotInsurable" />
</rule>
</rule-execution-set>

<?xml version="1.0" encoding="UTF-8"?> <rule-execution-set> <name>RuleExecutionSet1</name> <description>Rule Execution Set</description> <synonymn name="applyRule" class="com.vivek.jruleengine.example.ApplyRule" /> <rule name="turnoverInsurable" description="Check if turnover insurable" > <if leftTerm="applyRule.getTurnover" op=">" rightTerm="2000000" /> <then method="applyRule.printCompanyInsurable()" /> </rule> <rule name="turnoverNotInsurable" description="Check if turnover not insurable" > <if leftTerm="applyRule.getTurnover" op="<" rightTerm="2000000" /> <then method="applyRule.printCompanyNotInsurable" /> </rule> </rule-execution-set>

结论：
Drools除了提供正常的规则引擎的能力，还有以下额外的优点：
(1) 无论是技术人士还是商业人士，Drools都是用户友好的，它提供了一个巨大的支持工具集。
(2) Drools的Reteoo算法可加速和可扩展。
(3) Drools提供的Eclipse插件带有自动完成智能感知和调试视图、规则流GUI等。
(4) 基于Web的工具（Guvnor）：是一个业务规则管理系统（BRMS），它提供了高级规则授权、版本控制和管理。

至于其他的规则引擎，我个人建议在项目中使用Drools，因为它有一个很大的支持社区。在IDE支持和基于Web的规则管理工具（Guvnor）也有很大优势。