Overhaul operator stubs

[AlexWaygood]

_operator.pyi currently has a large number of functions that are only annotated with Any. I believe we can do much better than that.

[github-actions]

Diff from mypy_primer, showing the effect of this PR on open source code:

pandera (https://github.com/pandera-dev/pandera)
+ pandera/strategies.py:565: error: Cannot call function of unknown type  [operator]
+ pandera/checks.py:731: error: Cannot call function of unknown type  [operator]

ibis (https://github.com/ibis-project/ibis)
+ ibis/backends/base/sql/registry/window.py:85: error: Cannot call function of unknown type

xarray (https://github.com/pydata/xarray)
+ xarray/core/accessor_str.py:2509: error: Argument 1 to "reduce" has incompatible type overloaded function; expected "Callable[[Iterable[Any], Iterable[Any]], Iterable[Any]]"  [arg-type]

rotki (https://github.com/rotki/rotki)
+ rotkehlchen/chain/ethereum/modules/balancer/balancer.py:151: error: No overload variant of "map" matches argument types "function", "List[AssetAmount]", "List[AssetAmount]"
+ rotkehlchen/chain/ethereum/modules/balancer/balancer.py:151: note: Possible overload variant:
+ rotkehlchen/chain/ethereum/modules/balancer/balancer.py:151: note:     def [_S, _T1, _T2] __init__(self, Callable[[_T1, _T2], _S], Iterable[_T1], Iterable[_T2]) -> map[_S]
+ rotkehlchen/chain/ethereum/modules/balancer/balancer.py:151: note:     <5 more non-matching overloads not shown>
+ rotkehlchen/chain/ethereum/modules/balancer/balancer.py:154: error: Cannot call function of unknown type
+ tools/data_faker/data_faker/statistics.py:154: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)

Tanjun (https://github.com/FasterSpeeding/Tanjun)
+ tanjun/conversion.py:603: error: Argument 1 to "index" has incompatible type "float"; expected "SupportsIndex"
+ tanjun/conversion.py:654: error: Argument 1 to "index" has incompatible type "float"; expected "SupportsIndex"

pandas (https://github.com/pandas-dev/pandas)
+ pandas/core/computation/ops.py:413: error: Cannot call function of unknown type  [operator]
+ pandas/core/computation/ops.py:455: error: Cannot call function of unknown type  [operator]
+ pandas/core/computation/ops.py:583: error: Cannot call function of unknown type  [operator]
+ pandas/core/algorithms.py:1585: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)  [assignment]
+ pandas/core/indexes/base.py:3854: error: Cannot call function of unknown type  [operator]
+ pandas/core/arrays/interval.py:1281: error: Cannot call function of unknown type  [operator]
+ pandas/core/indexes/datetimes.py:854: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)  [assignment]
+ pandas/core/indexes/datetimes.py:856: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)  [assignment]
+ pandas/core/indexes/datetimes.py:859: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)  [assignment]
+ pandas/core/indexes/datetimes.py:864: error: Incompatible types in assignment (expression has type overloaded function, variable has type overloaded function)  [assignment]
+ pandas/core/indexes/interval.py:630: error: Cannot call function of unknown type  [operator]
+ pandas/core/generic.py:1503: error: Argument 1 to "inv" has incompatible type "Union[ExtensionArray, ndarray[Any, Any]]"; expected "_SupportsInv[<nothing>]"  [arg-type]
+ pandas/core/generic.py:1505: error: Argument 1 to "neg" has incompatible type "Union[ExtensionArray, ndarray[Any, Any]]"; expected "_SupportsNeg[ndarray[Any, dtype[Any]]]"  [arg-type]
+ pandas/core/generic.py:1517: error: Argument 1 to "pos" has incompatible type "Union[ExtensionArray, ndarray[Any, Any]]"; expected "_SupportsPos[ndarray[Any, dtype[Any]]]"  [arg-type]
+ pandas/io/formats/style.py:3490: error: Incompatible types in assignment (expression has type "Tuple[overloaded function, overloaded function]", variable has type "Tuple[overloaded function, overloaded function]")  [assignment]
+ pandas/io/formats/style.py:3492: error: Incompatible types in assignment (expression has type "Tuple[overloaded function, overloaded function]", variable has type "Tuple[overloaded function, overloaded function]")  [assignment]
+ pandas/io/formats/style.py:3494: error: Incompatible types in assignment (expression has type "Tuple[overloaded function, overloaded function]", variable has type "Tuple[overloaded function, overloaded function]")  [assignment]
+ pandas/io/formats/style.py:3511: error: Unsupported left operand type for & ("object")  [operator]

[JelleZijlstra]

Nice, the Tanjun one found a real bug (or wrong annotation). Much of the other mypy-primer output looks like false positives though.

[AlexWaygood]

Nice, the Tanjun one found a real bug (or wrong annotation). Much of the other mypy-primer output looks like false positives though.

Yah, I kinda expected I'd get quite a few 🙂 but wanted to push it all in one go to see what would happen. Probably won't have a chance to look through the primer output until tomorrow.