Skip to main content

​
PythonVerifier

class PythonVerifier(BaseVerifier):
The PythonVerifier class verifies Python-based implementations by executing them in an isolated virtual environment. Features:
  • Creates a virtual environment with a specified Python version.
  • Installs required packages before executing the provided script.
  • Executes the script and compares the output against a ground truth, if supplied.
  • Automatically cleans up the virtual environment after execution.
The verification process ensures that the code runs in a controlled environment, minimizing external dependencies and conflicts.

​
init

def __init__(
    self,
    extractor: Optional[BaseExtractor] = None,
    timeout: Optional[float] = 30.0,
    required_packages: Optional[List[str]] = None,
    float_tolerance: Optional[float] = None,
    **kwargs
):
Initializes the PythonVerifier. Parameters:
  • extractor (Optional[BaseExtractor], optional): The extractor to use for extracting code from the solution. (default: :obj:None)
  • timeout (Optional[float], optional): The execution timeout in seconds. (default: :obj:30.0)
  • required_packages (Optional[List[str]], optional): A list of packages to install in the virtual environment. (default: :obj:None)
  • float_tolerance (Optional[float], optional): The tolerance for floating point comparisons. (default: :obj:None)

​
_cleanup_venv

def _cleanup_venv(self):
Clean up the virtual environment if it exists.

​
_is_uv_environment

def _is_uv_environment(self):
Detect whether the current Python runtime is managed by uv.

​
_setup_with_uv

def _setup_with_uv(self):
Create virtual environment and install packages using uv.

​
_is_expression

def _is_expression(self, code: str):
Determines whether a given string of code is a single expression. This utility uses Python’s AST module to parse the code and checks if it consists of a single expression node. Parameters:
  • code (str): The Python code to analyze.
Returns: bool: True if the code is a single expression, False otherwise.

​
_is_equal_with_tolerance

def _is_equal_with_tolerance(self, a: Any, b: Any):
Compares two Python objects for equality with optional float tolerance. This method recursively compares nested structures (lists, tuples, sets, and dictionaries) and applies floating point tolerance when comparing numerical values. If no float tolerance is set, a runtime error is raised. Parameters:
  • a (Any): First value to compare.
  • b (Any): Second value to compare.
Returns: bool: True if the values are considered equal within the specified float tolerance; False otherwise.
⌘I