CLN: parts of #29667 #29677
CLN: parts of #29667 #29677
Conversation
| @@ -197,7 +197,9 @@ class Op: | |||
| Hold an operator of arbitrary arity. | |||
| """ | |||
|
|
|||
| def __init__(self, op, operands, *args, **kwargs): | |||
| op: str | |||
There was a problem hiding this comment.
no, there is a op attribute that is a str
There was a problem hiding this comment.
is it necessary?
reveal_type(_bool_op_map)
reveal_type(_bool_op_map.get("not", "not"))
pandas\core\computation\ops.py:193: note: Revealed type is 'builtins.dict[builtins.str*, builtins.str*]'
pandas\core\computation\ops.py:194: note: Revealed type is 'builtins.str'
There was a problem hiding this comment.
is it necessary?
i guess im not clear on the policy. i thought it was a "more is better" kind of thing
There was a problem hiding this comment.
I'm just suspicious when I see unnecessary variable annotations since they could be casts in disguise.
In this case, it could have been (but wasn't) hiding that the return type of _bool_op_map.get(op,op) might be Any.
If the return type of some_function is Any and we do some variable: some_type = some_function() it is effectively a cast.
def some_function():
return "foo"
x: int = some_function()
reveal_type(x)test.py:5: note: Revealed type is 'builtins.int'
I think we should just allow mypy to use the return type of the initialization unless needed.
| from pandas.core.computation.common import _ensure_decoded | ||
| from pandas.core.computation.expr import BaseExprVisitor | ||
| from pandas.core.computation.ops import UndefinedVariableError, is_term | ||
|
|
||
| from pandas.io.formats.printing import pprint_thing, pprint_thing_encoded | ||
|
|
||
|
|
||
| class Scope(expr.Scope): | ||
| class Scope(_scope.Scope): | ||
| __slots__ = ("queryables",) |
There was a problem hiding this comment.
the medium-term goal that motivates this and the other pytables PR is that I think we can type queryables as Dict[str, Any], but it'll take a few steps to get there
|
thanks |
Breaks off easier parts