Update Typing Comments to Python3 Syntax

[WillAyd]

This should be done after #25789

Not necessarily analyzing types yet, but I think it makes sense from a readability perspective to start switching over to Python3 syntax for types. A quick grep shows 70 of these in the code base:

pandas/core/dtypes/base.py:        # type: () -> Optional[List[str]]
pandas/core/dtypes/base.py:        # type: () -> bool
pandas/core/dtypes/base.py:        # type: () -> bool
pandas/core/dtypes/base.py:        # type: () -> Type
pandas/core/dtypes/base.py:        # type: () -> str
pandas/core/groupby/groupby.py:            # type: (np.ndarray) -> (np.ndarray, Type)
pandas/core/groupby/groupby.py:            # type: (np.ndarray, Type) -> np.ndarray
pandas/core/groupby/groupby.py:            # type: (np.ndarray) -> (np.ndarray, Optional[Type])
pandas/core/groupby/groupby.py:            # type: (np.ndarray, Optional[Type]) -> np.ndarray
pandas/core/internals/blocks.py:              periods,                  # type: int
pandas/core/internals/blocks.py:              axis=0,                   # type: libinternals.BlockPlacement
pandas/core/internals/blocks.py:              fill_value=None):         # type: Any
pandas/core/internals/blocks.py:        # type: (...) -> List[ExtensionBlock]
pandas/core/internals/managers.py:    # type: (List[Block]) -> Optional[Union[np.dtype, ExtensionDtype]]
pandas/core/common.py:    # type: (Any) -> bool
pandas/core/arrays/datetimes.py:    _dtype = None  # type: Union[np.dtype, DatetimeTZDtype]
pandas/core/arrays/datetimes.py:        # type: () -> Union[np.dtype, DatetimeTZDtype]
pandas/core/arrays/integer.py:        # type: () -> np.ndarray
pandas/core/arrays/integer.py:        # type: () -> np.ndarray
pandas/core/arrays/period.py:        # type: (Sequence[Optional[Period]], PeriodDtype, bool) -> PeriodArray
pandas/core/arrays/period.py:        # type: (Union[Period, NaTType]) -> int
pandas/core/arrays/period.py:        # type: (str) -> Period
pandas/core/arrays/period.py:            other,      # type: Union[ExtensionArray, np.ndarray[int]]
pandas/core/arrays/period.py:            op          # type: Callable[Any, Any]
pandas/core/arrays/period.py:        # type: (...) -> PeriodArray
pandas/core/arrays/period.py:    # type: (Sequence[Optional[Period]], Optional[Tick], bool) -> PeriodArray
pandas/core/arrays/sparse.py:        # type: (Union[str, np.dtype, 'ExtensionDtype', type], Any) -> None
pandas/core/arrays/sparse.py:    # type: (SparseArray) -> np.ndarray
pandas/core/arrays/sparse.py:    # type: (SparseArray, SparseArray, Callable, str) -> Any
pandas/core/arrays/sparse.py:        # type: (np.ndarray, SparseIndex, SparseDtype) -> 'SparseArray'
pandas/core/arrays/datetimelike.py:        # type: () -> Union[type, Tuple[type]]
pandas/core/arrays/datetimelike.py:        # type: (str) -> Union[Period, Timestamp, Timedelta, NaTType]
pandas/core/arrays/datetimelike.py:        # type: (Union[Period, Timestamp, Timedelta, NaTType]) -> int
pandas/core/arrays/datetimelike.py:        # type: (Union[Period, Timestamp, Timedelta, NaTType]) -> None
pandas/core/arrays/datetimelike.py:        # type: () -> np.ndarray
pandas/core/arrays/datetimelike.py:        # type: () -> int
pandas/core/arrays/datetimelike.py:            key,    # type: Union[int, Sequence[int], Sequence[bool], slice]
pandas/core/arrays/datetimelike.py:            value,  # type: Union[NaTType, Any, Sequence[Any]]
pandas/core/arrays/datetimelike.py:        # type: (...) -> None
pandas/core/arrays/array_.py:def array(data,         # type: Sequence[object]
pandas/core/arrays/array_.py:          dtype=None,   # type: Optional[Union[str, np.dtype, ExtensionDtype]]
pandas/core/arrays/array_.py:          copy=True,    # type: bool
pandas/core/arrays/array_.py:    # type: (...) -> ExtensionArray
pandas/core/arrays/base.py:        # type: (Union[int, np.ndarray], Any) -> None
pandas/core/arrays/base.py:        # type: () -> int
pandas/core/arrays/base.py:        # type: () -> ExtensionDtype
pandas/core/arrays/base.py:        # type: () -> Tuple[int, ...]
pandas/core/arrays/base.py:        # type: () -> int
pandas/core/arrays/base.py:        # type: () -> int
pandas/core/arrays/base.py:        # type: () -> Union[ExtensionArray, np.ndarray]
pandas/core/arrays/base.py:        # type: () -> np.ndarray
pandas/core/arrays/base.py:        # type: (int, object) -> ExtensionArray
pandas/core/arrays/base.py:        # type: () -> Tuple[np.ndarray, Any]
pandas/core/arrays/base.py:        # type: (int) -> Tuple[np.ndarray, ExtensionArray]
pandas/core/arrays/base.py:        # type: (Sequence[int], bool, Optional[Any]) -> ExtensionArray
pandas/core/arrays/base.py:        # type: (bool) -> ExtensionArray
pandas/core/arrays/base.py:        # type: (bool) -> Callable[[Any], Optional[str]]
pandas/core/arrays/base.py:        # type: () -> np.ndarray
pandas/core/arrays/base.py:        # type: (Sequence[ExtensionArray]) -> ExtensionArray
pandas/core/arrays/base.py:        # type: () -> np.ndarray
pandas/core/frame.py:                 key,           # type: Union[str, List[str]]
pandas/core/frame.py:                 ndim,          # type: int
pandas/core/frame.py:                 subset=None    # type: Union[Series, DataFrame, None]
pandas/core/frame.py:        # type: (...) -> Union[Series, DataFrame]
pandas/core/base.py:        # type: () -> ExtensionArray
pandas/core/base.py:        # type: () -> np.ndarray
pandas/core/indexes/period.py:    _data = None  # type: PeriodArray
pandas/core/indexes/datetimelike.py:    _data = None  # type: DatetimeLikeArrayMixin
pandas/core/indexes/datetimelike.py:        # type: () -> np.ndarray
pandas/core/indexes/base.py:        # type: () -> Union[ExtensionArray, Index, np.ndarray]

I think it makes sense to break this up into 2 PRs:

• Everything in pandas/core/arrays
• Everything else

[gwrome]

I’ll work on these.