from functools import partial, wraps
from pathlib import Path
from types import TracebackType
from typing import (
TYPE_CHECKING,
Any,
Callable,
Dict,
Iterable,
List,
NewType,
Sequence,
Tuple,
Type,
Union,
)
import numpy as np
from typing_extensions import TypedDict, get_args
if TYPE_CHECKING:
import dask.array
import zarr
from magicgui.widgets import FunctionGui
from qtpy.QtWidgets import QWidget
try:
from numpy.typing import DTypeLike # requires numpy 1.20
except ImportError:
# Anything that can be coerced into numpy.dtype.
# Reference: https://docs.scipy.org/doc/numpy/reference/arrays.dtypes.html
from typing import TypeVar
from typing_extensions import Protocol
_DType_co = TypeVar("_DType_co", covariant=True, bound=np.dtype)
# A protocol for anything with the dtype attribute
class _SupportsDType(Protocol[_DType_co]):
@property
def dtype(self) -> _DType_co:
...
DTypeLike = Union[ # type: ignore
np.dtype, # default data type (float64)
None,
type, # array-scalar types and generic types
_SupportsDType[np.dtype], # anything with a dtype attribute
str, # character codes, type strings, e.g. 'float64'
]
# This is a WOEFULLY inadequate stub for a duck-array type.
# Mostly, just a placeholder for the concept of needing an ArrayLike type.
# Ultimately, this should come from https://github.com/napari/image-types
# and should probably be replaced by a typing.Protocol
# note, numpy.typing.ArrayLike (in v1.20) is not quite what we want either,
# since it includes all valid arguments for np.array() ( int, float, str...)
ArrayLike = Union[np.ndarray, 'dask.array.Array', 'zarr.Array']
# layer data may be: (data,) (data, meta), or (data, meta, layer_type)
# using "Any" for the data type until ArrayLike is more mature.
FullLayerData = Tuple[Any, Dict, str]
LayerData = Union[Tuple[Any], Tuple[Any, Dict], FullLayerData]
PathLike = Union[str, Path]
PathOrPaths = Union[str, Sequence[str]]
ReaderFunction = Callable[[PathOrPaths], List[LayerData]]
WriterFunction = Callable[[str, List[FullLayerData]], List[str]]
ExcInfo = Union[
Tuple[Type[BaseException], BaseException, TracebackType],
Tuple[None, None, None],
]
# Types for GUI HookSpecs
WidgetCallable = Callable[..., Union['FunctionGui', 'QWidget']]
AugmentedWidget = Union[WidgetCallable, Tuple[WidgetCallable, dict]]
# Sample Data for napari_provide_sample_data hookspec is either a string/path
# or a function that returns an iterable of LayerData tuples
SampleData = Union[PathLike, Callable[..., Iterable[LayerData]]]
# or... they can provide a dict as follows:
[docs]class SampleDict(TypedDict):
display_name: str
data: SampleData
# these types are mostly "intentionality" placeholders. While it's still hard
# to use actual types to define what is acceptable data for a given layer,
# these types let us point to a concrete namespace to indicate "this data is
# intended to be (and is capable of) being turned into X layer type".
# while their names should not change (without deprecation), their typing
# implementations may... or may be rolled over to napari/image-types
if tuple(np.__version__.split('.')) < ('1', '20'):
# this hack is because NewType doesn't allow `Any` as a base type
# and numpy <=1.20 didn't provide type stubs for np.ndarray
# https://github.com/python/mypy/issues/6701#issuecomment-609638202
class ArrayBase(np.ndarray):
def __getattr__(self, name: str) -> Any:
return object.__getattribute__(self, name)
else:
ArrayBase = np.ndarray # type: ignore
ImageData = NewType("ImageData", ArrayBase)
LabelsData = NewType("LabelsData", ArrayBase)
PointsData = NewType("PointsData", ArrayBase)
ShapesData = NewType("ShapesData", List[ArrayBase])
SurfaceData = NewType("SurfaceData", Tuple[ArrayBase, ArrayBase, ArrayBase])
TracksData = NewType("TracksData", ArrayBase)
VectorsData = NewType("VectorsData", ArrayBase)
_LayerData = Union[
ImageData,
LabelsData,
PointsData,
ShapesData,
SurfaceData,
TracksData,
VectorsData,
]
LayerDataTuple = NewType("LayerDataTuple", tuple)
[docs]def image_reader_to_layerdata_reader(
func: Callable[[PathOrPaths], ArrayLike]
) -> ReaderFunction:
"""Convert a PathLike -> ArrayLike function to a PathLike -> LayerData.
Parameters
----------
func : Callable[[PathLike], ArrayLike]
A function that accepts a string or list of strings, and returns an
ArrayLike.
Returns
-------
reader_function : Callable[[PathLike], List[LayerData]]
A function that accepts a string or list of strings, and returns data
as a list of LayerData: List[Tuple[ArrayLike]]
"""
@wraps(func)
def reader_function(*args, **kwargs) -> List[LayerData]:
result = func(*args, **kwargs)
return [(result,)]
return reader_function
def _register_types_with_magicgui():
"""Register ``napari.types`` objects with magicgui."""
import sys
from concurrent.futures import Future
from magicgui import register_type
from .utils import _magicgui as _mgui
for _type in (LayerDataTuple, List[LayerDataTuple]):
register_type(
_type,
return_callback=_mgui.add_layer_data_tuples_to_viewer,
)
if sys.version_info >= (3, 9):
future_type = Future[_type] # type: ignore
register_type(future_type, return_callback=_mgui.add_future_data)
for data_type in get_args(_LayerData):
register_type(
data_type,
choices=_mgui.get_layers_data,
return_callback=_mgui.add_layer_data_to_viewer,
)
if sys.version_info >= (3, 9):
register_type(
Future[data_type], # type: ignore
choices=_mgui.get_layers_data,
return_callback=partial(
_mgui.add_future_data, _from_tuple=False
),
)
_register_types_with_magicgui()