JBPM3中文用户手册——第三章：教程(1)

 [译：liuu  对应版本: 3.3.1/3.2.3 ]        

3       教程

本教程将为你展示使用jpdl的基本流程构建和管理运行时执行的API用法。

 

教程的形式是解说一组示例，每个示例关注特定的主题并包含丰富的注释。这些示例可以在下载包目录src/java.examples下找到。

 

最佳的学习方式，是基于已提供的例子，新建一个项目，边修改，边试验。

 

Eclipse用户开始步骤如下：

1、  下载jbpm分发包，并解压

2、  “文件”à“导入”à选择“现有项目到工作空间”，点击“下一步”，然后浏览选中jbpm的根目录，点“完成”。现在，在你的工作空间里，有了一个jbpm3项目。

3、  你可以在“src/java.examples/...”目录下，找到本教程的示例。打开这些例子后，你可以通过“运行”à“运行为”à“JUnit Test”运行它们。

 

jBPM包含一个图形化设计器工具，用于编辑示例中展示的那些XML。你可以在2.1可用下载概览一节中，找到这个工具的下载指导。不过，你并不必须要这个图形化设计器，来完成这个教程。

3.1    Hello World示例

一个流程定义是一张有向图，由节点和迁移组成。HelloWorld流程有3个节点。为了看清这些零零散散的东西，是如何放到一起的工作，我们将从一个简单的流程开始，而且也不使用流程设计工具。下图是HelloWorld流程的图示：

 

 

 

 

 

图表 3.1 Hello World流程图

 

public void testHelloWorldProcess() {   // This method shows a process definition and one execution   // of the process definition.  The process definition has   // 3 nodes: an unnamed start-state, a state 's' and an   // end-state named 'end'.   // The next line parses a piece of xml text into a   // ProcessDefinition.  A ProcessDefinition is the formal   // description of a process represented as a java object.   ProcessDefinition processDefinition = ProcessDefinition.parseXmlString(     "<process-definition>" +     "  <start-state>" +     "    <transition to='s' />" +     "  </start-state>" +     "  <state name='s'>" +     "    <transition to='end' />" +     "  </state>" +     "  <end-state name='end' />" +     "</process-definition>"   );     // The next line creates one execution of the process definition.   // After construction, the process execution has one main path   // of execution (=the root token) that is positioned in the   // start-state.   ProcessInstance processInstance =       new ProcessInstance(processDefinition);     // After construction, the process execution has one main path   // of execution (=the root token).   Token token = processInstance.getRootToken();     // Also after construction, the main path of execution is positioned   // in the start-state of the process definition.   assertSame(processDefinition.getStartState(), token.getNode());     // Let's start the process execution, leaving the start-state   // over its default transition.   token.signal();   // The signal method will block until the process execution   // enters a wait state.     // The process execution will have entered the first wait state   // in state 's'. So the main path of execution is now   // positioned in state 's'   assertSame(processDefinition.getNode("s"), token.getNode());     // Let's send another signal.  This will resume execution by   // leaving the state 's' over its default transition.   token.signal();   // Now the signal method returned because the process instance   // has arrived in the end-state.     assertSame(processDefinition.getNode("end"), token.getNode()); }