Add conditional

[erichlf]

This PR implements the conditional feature using a simple if keyword.

Fixes #5

Motivation and Context

To be able to include or exclude elements from a behavior tree in a programmatic way a conditional was necessary.

Changes

• Add if keyword

Type of changes

• New feature

Checklist

• Update dependencies
• Update version
• Update README

Testing

1. colcon build --packages-up-to py_trees_parser
2. colcon test --event-handlers console_cohesion+ --packages-select py_trees_parser

[erichlf]

@pakagronglb I think we may go with this implementation instead. If we decide to take this one I want to make sure you know that I really appreciate your contribution. Your PR forced me to think about this a bit more and spurred me to write the code.

[erichlf]

@SyZbidi this will be especially useful for simulation and excluding hardware specific behaviors.

[pakagronglb]

@pakagronglb I think we may go with this implementation instead. If we decide to take this one I want to make sure you know that I really appreciate your contribution. Your PR forced me to think about this a bit more and spurred me to write the code.

Thank you so much for the opportunity! I really appreciate it. It was a fun and challenging one.

[pakagronglb]

I hope my feedback or output will be useful somehow. Thank you again for the opportunity.

Looking at the specific implementation code, I have a few technical observations:

1. Attribute Handling: The code removes the if attribute but doesn't appear to have attribute type validation. Consider adding type checking for condition values.

2. Function Signature Analysis: The code has good handling of different function signatures (behaviour, subtrees, tasks). However, for consistency, the conditional processing could be abstracted into its own helper method to keep _create_node cleaner.

3. Error Propagation: When raising BTParseError, it might be helpful to include more context about the failed condition in debug logs to aid troubleshooting.

4. Advanced Pattern: You could consider implementing a lightweight condition parser instead of direct evaluation for better security and flexibility:

def _parse_condition(self, condition_str, args):
    # Safer condition parsing with explicit operator support
    # This would allow for more controlled evaluation

5. Performance Enhancement: Since removing dictionary keys is an O(1) operation, you're handling things efficiently, but consider grouping the attribute processing:

# Extract and process special attributes in one block
special_attrs = {}
for attr in ["name", "if", "other_special_attr"]:
    if attr in node_attribs:
        special_attrs[attr] = node_attribs.pop(attr)

This implementation looks solid overall - the approach of handling conditions as attributes is clean and leverages the existing parsing architecture well. Nice work!!!

[erichlf]

I hope my feedback or output will be useful somehow. Thank you again for the opportunity.

Looking at the specific implementation code, I have a few technical observations:

1. Attribute Handling: The code removes the if attribute but doesn't appear to have attribute type validation. Consider adding type checking for condition values.
2. Function Signature Analysis: The code has good handling of different function signatures (behaviour, subtrees, tasks). However, for consistency, the conditional processing could be abstracted into its own helper method to keep _create_node cleaner.
3. Error Propagation: When raising BTParseError, it might be helpful to include more context about the failed condition in debug logs to aid troubleshooting.
4. Advanced Pattern: You could consider implementing a lightweight condition parser instead of direct evaluation for better security and flexibility:

def _parse_condition(self, condition_str, args):
    # Safer condition parsing with explicit operator support
    # This would allow for more controlled evaluation

5. Performance Enhancement: Since removing dictionary keys is an O(1) operation, you're handling things efficiently, but consider grouping the attribute processing:

# Extract and process special attributes in one block
special_attrs = {}
for attr in ["name", "if", "other_special_attr"]:
    if attr in node_attribs:
        special_attrs[attr] = node_attribs.pop(attr)

This implementation looks solid overall - the approach of handling conditions as attributes is clean and leverages the existing parsing architecture well. Nice work!!!

1. I am not sure this is possible since the conditional is a statement.
2. Not sure how this applies.
3. The BTParseError is propogated and should be logged by the user. It typically contains plenty of info to determine what happened.
4. I may make a ticket for this as handling the python the way I do is not exactly secure, so this would be for an general python parsing overhaul to happen in the future.
5. Not sure how this actually applies to the current code.