Jbpm task management implementation

A process definition has a taskmgmtdefinition; A taskmgmtdefinition corresponds to multiple actors and multiple tasks at the same time;
An actor has multiple tasks. You can directly obtain the corresponding task from taskmgmtdefinition using the task name;

One process instance has one taskmgmtinstance; one taskmgmtinstance corresponds to multiple actorinstances and multiple taskinstance;
An actorinstance has multiple taskinstance, which can be directly obtained from taskmgmtinstance;

One tasknode corresponds to multiple tasks.
For such a process definition:
<Task-node name = 'A'>
<Task Name = 'laundry '/>
<Task Name = 'dishes'/>
<Task Name = 'change nappy'/>
<Transition to = 'B'/>
</Task-node>
The process enters the subsequent node only after all three tasks on the node are completed.
 
For such a process definition:
<Task-node name = 'A' signal = 'first'>
<Task Name = 'laundry '/>
<Task Name = 'dishes'/>
<Task Name = 'change nappy'/>
<Transition to = 'B'/>
</Task-node>
After the first task is completed, the token points to the following node.
 
For such a process definition:
<Task-node name = 'A' signal = 'never '>
<Task Name = 'laundry '/>
<Task Name = 'dishes'/>
<Task Name = 'change nappy'/>
<Transition to = 'B'/>
</Task-node>
After the three tasks are completed, the token still does not point to the following node; you need to manually call processinstance. Signal ()
To drive the process to the following nodes.

For such a process definition:
<Task-node name = 'A' signal = 'unsynchronized'>
<Task Name = 'laundry '/>
<Task Name = 'dishes'/>
<Task Name = 'change nappy'/>
<Transition to = 'B'/>
</Task-node>
The token does not stop at the current node, but directly goes to the following node.

Jbpm uses join, fork, merge, and demo-te to implement the same functions. However, they have different implementation methods.
Node. One is implemented by task.