PDDLReader

class unified_planning.io.PDDLReader(environment: Environment | None = None)[source]

Bases: object

Parse a PDDL domain file and, optionally, a PDDL problem file and generate the equivalent Problem.

Note: in the error report messages, a tabulation counts as one column; and due to PDDL case-insensitivity, everything in the PDDL files will be turned to lower case, so the names of fluents, actions etc. and the error report will all be in lower-case.

parse_problem(domain_filename: str, problem_filename: str | None = None) → Problem[source]

Takes in input a filename containing the PDDL domain and optionally a filename containing the PDDL problem and returns the parsed Problem.

Note: that if the problem_filename is None, an incomplete Problem will be returned.

Note: due to PDDL case-insensitivity, everything in the PDDL files will be turned to lower case, so the names of fluents, actions etc. and the error report will all be in lower-case.

Parameters:

  • domain_filename – The path to the file containing the PDDL domain.

  • problem_filename – Optionally the path to the file containing the PDDL problem.

Returns:

The Problem parsed from the given pddl domain + problem.

parse_problem_string(domain_str: str, problem_str: str | None = None) → Problem[source]

Takes in input a str representing the PDDL domain and optionally a str representing the PDDL problem and returns the parsed Problem.

Note that if the problem_str is None, an incomplete Problem will be returned.

Note: due to PDDL case-insensitivity, everything in the PDDL files will be turned to lower case, so the names of fluents, actions etc. and the error report will all be in lower-case.

Parameters:

  • domain_filename – The string representing the PDDL domain.

  • problem_filename – Optionally the string representing the PDDL problem.

Returns:

The Problem parsed from the given pddl domain + problem.

parse_plan(problem: Problem, plan_filename: str, get_item_named: Callable[[str], Type | Action | Fluent | Object | Parameter | Variable | Agent | Method | Task] | None = None) → Plan[source]

Takes a problem, a filename and optionally a map of renaming and returns the plan parsed from the file.

The format of the file must be: (action-name param1 param2 ... paramN) in each line for SequentialPlans start-time: (action-name param1 param2 ... paramN) [duration] in each line for TimeTriggeredPlans, where [duration] is optional and not specified for InstantaneousActions.

Parameters:

  • problem – The up.model.problem.Problem instance for which the plan is generated.

  • plan_filename – The path of the file in which the plan is written.

  • get_item_named – A function that takes a name and returns the original up.model element instance linked to that renaming; if None the problem is used to retrieve the actions and objects in the plan from their name.

Returns:

The up.plans.Plan corresponding to the parsed plan from the file

parse_plan_string(problem: Problem, plan_str: str, get_item_named: Callable[[str], Type | Action | Fluent | Object | Parameter | Variable | Agent | Method | Task] | None = None) → Plan[source]

Takes a problem, a string and optionally a map of renaming and returns the plan parsed from the string.

The format of the file must be: (action-name param1 param2 ... paramN) in each line for SequentialPlans start-time: (action-name param1 param2 ... paramN) [duration] in each line for TimeTriggeredPlans, where [duration] is optional and not specified for InstantaneousActions.

Parameters:

  • problem – The up.model.problem.Problem instance for which the plan is generated.

  • plan_str – The plan in string.

  • get_item_named – A function that takes a name and returns the original up.model element instance linked to that renaming; if None the problem is used to retrieve the actions and objects in the plan from their name.:return: The up.plans.Plan corresponding to the parsed plan from the string