shark工作流源代码学习(二)
org.enhydra.shark.Shark 在启动Shark引擎前,要对它进行配置,配置要做的操作就是Shark.configure("shark.conf"),下面按先后顺序说明;
1) 整个系统只用配置一次,如果已经配置过,将不会再配置而直接返回,
如下:
if (isConfigured)
{
SharkEngineManager.getInstance().getCallbackUtilities().info( "Trying to configure shark instance that is already configured !!!");
return;
}
2)如果没有为配置文件给绝对路径,系统会自动进行获取:
if (!configFile.isAbsolute()) {
configFile=configFile.getAbsoluteFile();
}
3)然后就是获取properties,进行配置,并将isConfigured设置为true:
try {
fis=new FileInputStream(configFile);
Properties props=new Properties();
props.load(fis);
fis.close();
adjustSharkProperties(props);
} catch (Exception ex) {
throw new Error ("Something went wrong during reading of configuration from the file");
}
isConfigured=true;
4)然后对SharkEngineManager进行配置,它利用回调函数,对该管理器的各子管理器和工厂进行初始化和配置,以后可以直接调用:
try {
callbackUtilities=(CallbackUtil)cl.loadClass(cbuClassName).newInstance(); callbackUtilities.setProperties(properties); objectFactory =(SharkObjectFactory)cl.loadClass(objectFactoryClassName).newInstance();
toolAgentManager = (ToolAgentManager) cl.loadClass(tamClassName).newInstance();
xmlInterface=new XMLInterfaceForJDK13();
} catch (Throwable ex) {
throw new Error("Problems instantiating core managers");
}
5)最后,对系统的缓存和LimitManager进行配置: shark.initCaches(); shark.reevaluateAssignments(); shark.initLimitManager();