if length.value > len(inferred_sequence.elts): # In this case, this will raise a ValueError raise UseInferenceDefault try: elts = random.sample(inferred_sequence.elts, length.value) except ValueError as exc: raise UseInferenceDefault from exc new_node = List(lineno=node.lineno, col_offset=node.col_offset, parent=node.scope()) new_elts = [ _clone_node_with_lineno(elt, parent=new_node, lineno=new_node.lineno) for elt in elts ] new_node.postinit(new_elts) return iter((new_node,)) def _looks_like_random_sample(node): func = node.func if isinstance(func, Attribute): return func.attrname == "sample" if isinstance(func, Name): return func.name == "sample" return False AstroidManager().register_transform( Call, inference_tip(infer_random_sample), _looks_like_random_sample )
return False for decorator in node.decorators.nodes: if not isinstance(decorator, Call): continue if _looks_like_functools_member(decorator, "lru_cache"): return True return False def _looks_like_functools_member(node, member) -> bool: """Check if the given Call node is a functools.partial call""" if isinstance(node.func, Name): return node.func.name == member if isinstance(node.func, Attribute): return (node.func.attrname == member and isinstance(node.func.expr, Name) and node.func.expr.name == "functools") return False _looks_like_partial = partial(_looks_like_functools_member, member="partial") AstroidManager().register_transform(FunctionDef, _transform_lru_cache, _looks_like_lru_cache) AstroidManager().register_transform( Call, inference_tip(_functools_partial_inference), _looks_like_partial, )
# Licensed under the LGPL: https://www.gnu.org/licenses/old-licenses/lgpl-2.1.en.html # For details: https://github.com/PyCQA/astroid/blob/main/LICENSE """Astroid hooks for numpy.core.function_base module.""" import functools from astroid.brain.brain_numpy_utils import infer_numpy_member, looks_like_numpy_member from astroid.inference_tip import inference_tip from astroid.manager import AstroidManager from astroid.nodes.node_classes import Attribute METHODS_TO_BE_INFERRED = { "linspace": """def linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=None, axis=0): return numpy.ndarray([0, 0])""", "logspace": """def logspace(start, stop, num=50, endpoint=True, base=10.0, dtype=None, axis=0): return numpy.ndarray([0, 0])""", "geomspace": """def geomspace(start, stop, num=50, endpoint=True, dtype=None, axis=0): return numpy.ndarray([0, 0])""", } for func_name, func_src in METHODS_TO_BE_INFERRED.items(): inference_function = functools.partial(infer_numpy_member, func_src) AstroidManager().register_transform( Attribute, inference_tip(inference_function), functools.partial(looks_like_numpy_member, func_name), )
def swapaxes(self, axis1, axis2): return np.ndarray([0, 0]) def take(self, indices, axis=None, out=None, mode='raise'): return np.ndarray([0, 0]) def tobytes(self, order='C'): return b'' def tofile(self, fid, sep="", format="%s"): return None def tolist(self, ): return [] def tostring(self, order='C'): return b'' def trace(self, offset=0, axis1=0, axis2=1, dtype=None, out=None): return np.ndarray([0, 0]) def transpose(self, *axes): return np.ndarray([0, 0]) def var(self, axis=None, dtype=None, out=None, ddof=0, keepdims=False): return np.ndarray([0, 0]) def view(self, dtype=None, type=None): return np.ndarray([0, 0]) """ if numpy_supports_type_hints(): ndarray += """ @classmethod def __class_getitem__(cls, value): return cls """ node = extract_node(ndarray) return node.infer(context=context) def _looks_like_numpy_ndarray(node): return isinstance(node, Attribute) and node.attrname == "ndarray" AstroidManager().register_transform( Attribute, inference_tip(infer_numpy_ndarray), _looks_like_numpy_ndarray, )
# In this case, this will raise a ValueError raise UseInferenceDefault try: elts = random.sample(inferred_sequence.elts, length.value) except ValueError as exc: raise UseInferenceDefault from exc new_node = List(lineno=node.lineno, col_offset=node.col_offset, parent=node.scope()) new_elts = [ _clone_node_with_lineno(elt, parent=new_node, lineno=new_node.lineno) for elt in elts ] new_node.postinit(new_elts) return iter((new_node, )) def _looks_like_random_sample(node): func = node.func if isinstance(func, Attribute): return func.attrname == "sample" if isinstance(func, Name): return func.name == "sample" return False AstroidManager().register_transform(Call, inference_tip(infer_random_sample), _looks_like_random_sample)