|
| 1 | +from __future__ import annotations |
| 2 | + |
| 3 | +from collections.abc import Callable, Hashable, MutableMapping |
| 4 | +from typing import TYPE_CHECKING, Any, Union |
| 5 | + |
| 6 | +import numpy as np |
| 7 | + |
| 8 | +from xarray.core import indexing |
| 9 | +from xarray.core.variable import Variable |
| 10 | +from xarray.namedarray.parallelcompat import get_chunked_array_type |
| 11 | +from xarray.namedarray.pycompat import is_chunked_array |
| 12 | + |
| 13 | +if TYPE_CHECKING: |
| 14 | + T_VarTuple = tuple[tuple[Hashable, ...], Any, dict, dict] |
| 15 | + T_Name = Union[Hashable, None] |
| 16 | + |
| 17 | + |
| 18 | +class SerializationWarning(RuntimeWarning): |
| 19 | + """Warnings about encoding/decoding issues in serialization.""" |
| 20 | + |
| 21 | + |
| 22 | +class VariableCoder: |
| 23 | + """Base class for encoding and decoding transformations on variables. |
| 24 | +
|
| 25 | + We use coders for transforming variables between xarray's data model and |
| 26 | + a format suitable for serialization. For example, coders apply CF |
| 27 | + conventions for how data should be represented in netCDF files. |
| 28 | +
|
| 29 | + Subclasses should implement encode() and decode(), which should satisfy |
| 30 | + the identity ``coder.decode(coder.encode(variable)) == variable``. If any |
| 31 | + options are necessary, they should be implemented as arguments to the |
| 32 | + __init__ method. |
| 33 | +
|
| 34 | + The optional name argument to encode() and decode() exists solely for the |
| 35 | + sake of better error messages, and should correspond to the name of |
| 36 | + variables in the underlying store. |
| 37 | + """ |
| 38 | + |
| 39 | + def encode(self, variable: Variable, name: T_Name = None) -> Variable: |
| 40 | + """Convert an encoded variable to a decoded variable""" |
| 41 | + raise NotImplementedError() |
| 42 | + |
| 43 | + def decode(self, variable: Variable, name: T_Name = None) -> Variable: |
| 44 | + """Convert a decoded variable to an encoded variable""" |
| 45 | + raise NotImplementedError() |
| 46 | + |
| 47 | + |
| 48 | +class _ElementwiseFunctionArray(indexing.ExplicitlyIndexedNDArrayMixin): |
| 49 | + """Lazily computed array holding values of elemwise-function. |
| 50 | +
|
| 51 | + Do not construct this object directly: call lazy_elemwise_func instead. |
| 52 | +
|
| 53 | + Values are computed upon indexing or coercion to a NumPy array. |
| 54 | + """ |
| 55 | + |
| 56 | + def __init__(self, array, func: Callable, dtype: np.typing.DTypeLike): |
| 57 | + assert not is_chunked_array(array) |
| 58 | + self.array = indexing.as_indexable(array) |
| 59 | + self.func = func |
| 60 | + self._dtype = dtype |
| 61 | + |
| 62 | + @property |
| 63 | + def dtype(self) -> np.dtype: |
| 64 | + return np.dtype(self._dtype) |
| 65 | + |
| 66 | + def _oindex_get(self, key): |
| 67 | + return type(self)(self.array.oindex[key], self.func, self.dtype) |
| 68 | + |
| 69 | + def _vindex_get(self, key): |
| 70 | + return type(self)(self.array.vindex[key], self.func, self.dtype) |
| 71 | + |
| 72 | + def __getitem__(self, key): |
| 73 | + return type(self)(self.array[key], self.func, self.dtype) |
| 74 | + |
| 75 | + def get_duck_array(self): |
| 76 | + return self.func(self.array.get_duck_array()) |
| 77 | + |
| 78 | + def __repr__(self) -> str: |
| 79 | + return f"{type(self).__name__}({self.array!r}, func={self.func!r}, dtype={self.dtype!r})" |
| 80 | + |
| 81 | + |
| 82 | +def lazy_elemwise_func(array, func: Callable, dtype: np.typing.DTypeLike): |
| 83 | + """Lazily apply an element-wise function to an array. |
| 84 | + Parameters |
| 85 | + ---------- |
| 86 | + array : any valid value of Variable._data |
| 87 | + func : callable |
| 88 | + Function to apply to indexed slices of an array. For use with dask, |
| 89 | + this should be a pickle-able object. |
| 90 | + dtype : coercible to np.dtype |
| 91 | + Dtype for the result of this function. |
| 92 | +
|
| 93 | + Returns |
| 94 | + ------- |
| 95 | + Either a dask.array.Array or _ElementwiseFunctionArray. |
| 96 | + """ |
| 97 | + if is_chunked_array(array): |
| 98 | + chunkmanager = get_chunked_array_type(array) |
| 99 | + |
| 100 | + return chunkmanager.map_blocks(func, array, dtype=dtype) # type: ignore[arg-type] |
| 101 | + else: |
| 102 | + return _ElementwiseFunctionArray(array, func, dtype) |
| 103 | + |
| 104 | + |
| 105 | +def safe_setitem(dest, key: Hashable, value, name: T_Name = None): |
| 106 | + if key in dest: |
| 107 | + var_str = f" on variable {name!r}" if name else "" |
| 108 | + raise ValueError( |
| 109 | + f"failed to prevent overwriting existing key {key} in attrs{var_str}. " |
| 110 | + "This is probably an encoding field used by xarray to describe " |
| 111 | + "how a variable is serialized. To proceed, remove this key from " |
| 112 | + "the variable's attributes manually." |
| 113 | + ) |
| 114 | + dest[key] = value |
| 115 | + |
| 116 | + |
| 117 | +def pop_to( |
| 118 | + source: MutableMapping, dest: MutableMapping, key: Hashable, name: T_Name = None |
| 119 | +) -> Any: |
| 120 | + """ |
| 121 | + A convenience function which pops a key k from source to dest. |
| 122 | + None values are not passed on. If k already exists in dest an |
| 123 | + error is raised. |
| 124 | + """ |
| 125 | + value = source.pop(key, None) |
| 126 | + if value is not None: |
| 127 | + safe_setitem(dest, key, value, name=name) |
| 128 | + return value |
| 129 | + |
| 130 | + |
| 131 | +def unpack_for_encoding(var: Variable) -> T_VarTuple: |
| 132 | + return var.dims, var.data, var.attrs.copy(), var.encoding.copy() |
| 133 | + |
| 134 | + |
| 135 | +def unpack_for_decoding(var: Variable) -> T_VarTuple: |
| 136 | + return var.dims, var._data, var.attrs.copy(), var.encoding.copy() |
![pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?v=4&size=40){kind=avatar}
0 commit comments