py-trees-meet-groot 0.1

Simple library to load XML begavior Trees done using Groot GUI into the py_trees library

Import Groot XML files into py_trees

Groot is a Graphical Editor, written in C++ and Qt, to create BehaviorTrees. It is only compatible with BehaviorTree.CPP. This module is an attempt to use Groot to generate trees using py_trees instead of BehaviorTree.CPP.

Groot supported features

BehaviorTree.CPP and py_trees use different control sequences, decorators and tree leaves. This section explains the mapping between one and the other in order to use Groot in py_trees.

Control nodes to Composite nodes:

Groot	py_trees
Sequence	Sequence with memory=True
ReactiveSequence	Sequence with memory=False
Fallback	Selector with memory=True
ReactiveFallback	Selector with memory=False
Parallel with success_threshold=1	Parallel with policy=SuccessOnOne
Parallel with success_threshold <> 1	Parallel with policy=SuccessOnAll
IfThenElse	not available
WhileDoElse	not available
SwitchX	not available

Decorators to Decorators

Groot	py_trees
Timeout	Timeout
ForceFailure	SuccessIsFailure
ForceSuccess	FailureIsSuccess
Inverter	Inverter
BlackboardCheckX	EternalGuard?
KeepRunningUntilFailure	Condition?
Repeat	not available
RetryUntilSuccessful	not available
Delay	not available
User defined Decorator	Decorator in user_decorators

Additional py_trees decorators, both core defined as OneShot, StatusToBlackboard, FailureIsRunning, RunningIsSuccess... or completely created by the use must be added in the user_decorators diccionare and pass to the groot_xml_to_py_trees function to be used.

Actions and Conditions to Behaviors

Groot	py_trees
AlwaysFailure	Failure
AlwaysSuccess	Success
SetBlackBoard	SetBlackboardVariable output_key=value
user defined Action	Behavior in user_behaviors
user defined Condition	Behavior in user_behaviors

All manually defined Actions or Conditions in Groot has to be defined as py_trees behaviors and added into the user_behaviors list in function groot_xml_to_py_trees to be used.