forked from googleapis/python-bigquery-dataframes
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathexpression.py
More file actions
412 lines (328 loc) · 11.8 KB
/
Copy pathexpression.py
File metadata and controls
412 lines (328 loc) · 11.8 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
# Copyright 2023 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import annotations
import abc
import dataclasses
import itertools
import typing
from typing import Mapping, TypeVar, Union
import bigframes.core.identifiers as ids
import bigframes.dtypes as dtypes
import bigframes.operations
import bigframes.operations.aggregations as agg_ops
def const(
value: typing.Hashable, dtype: dtypes.ExpressionType = None
) -> ScalarConstantExpression:
return ScalarConstantExpression(value, dtype or dtypes.infer_literal_type(value))
def deref(name: str) -> DerefOp:
return DerefOp(ids.ColumnId(name))
def free_var(id: str) -> UnboundVariableExpression:
return UnboundVariableExpression(id)
@dataclasses.dataclass(frozen=True)
class Aggregation(abc.ABC):
"""Represents windowing or aggregation over a column."""
op: agg_ops.WindowOp = dataclasses.field()
@abc.abstractmethod
def output_type(
self, input_types: dict[ids.ColumnId, dtypes.ExpressionType]
) -> dtypes.ExpressionType:
...
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return ()
@abc.abstractmethod
def remap_column_refs(
self,
name_mapping: Mapping[ids.ColumnId, ids.ColumnId],
allow_partial_bindings: bool = False,
) -> Aggregation:
...
@dataclasses.dataclass(frozen=True)
class NullaryAggregation(Aggregation):
op: agg_ops.NullaryWindowOp = dataclasses.field()
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
return self.op.output_type()
def remap_column_refs(
self,
name_mapping: Mapping[ids.ColumnId, ids.ColumnId],
allow_partial_bindings: bool = False,
) -> NullaryAggregation:
return self
@dataclasses.dataclass(frozen=True)
class UnaryAggregation(Aggregation):
op: agg_ops.UnaryWindowOp = dataclasses.field()
arg: Union[DerefOp, ScalarConstantExpression] = dataclasses.field()
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
return self.op.output_type(self.arg.output_type(input_types))
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return self.arg.column_references
def remap_column_refs(
self,
name_mapping: Mapping[ids.ColumnId, ids.ColumnId],
allow_partial_bindings: bool = False,
) -> UnaryAggregation:
return UnaryAggregation(
self.op,
self.arg.remap_column_refs(
name_mapping, allow_partial_bindings=allow_partial_bindings
),
)
@dataclasses.dataclass(frozen=True)
class BinaryAggregation(Aggregation):
op: agg_ops.BinaryAggregateOp = dataclasses.field()
left: Union[DerefOp, ScalarConstantExpression] = dataclasses.field()
right: Union[DerefOp, ScalarConstantExpression] = dataclasses.field()
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
return self.op.output_type(
self.left.output_type(input_types), self.right.output_type(input_types)
)
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return (*self.left.column_references, *self.right.column_references)
def remap_column_refs(
self,
name_mapping: Mapping[ids.ColumnId, ids.ColumnId],
allow_partial_bindings: bool = False,
) -> BinaryAggregation:
return BinaryAggregation(
self.op,
self.left.remap_column_refs(
name_mapping, allow_partial_bindings=allow_partial_bindings
),
self.right.remap_column_refs(
name_mapping, allow_partial_bindings=allow_partial_bindings
),
)
TExpression = TypeVar("TExpression", bound="Expression")
@dataclasses.dataclass(frozen=True)
class Expression(abc.ABC):
"""An expression represents a computation taking N scalar inputs and producing a single output scalar."""
@property
def free_variables(self) -> typing.Tuple[str, ...]:
return ()
@property
@abc.abstractmethod
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
...
def remap_column_refs(
self: TExpression,
name_mapping: Mapping[ids.ColumnId, ids.ColumnId],
allow_partial_bindings: bool = False,
) -> TExpression:
return self.bind_refs(
{old_id: DerefOp(new_id) for old_id, new_id in name_mapping.items()}, # type: ignore
allow_partial_bindings=allow_partial_bindings,
)
@property
@abc.abstractmethod
def is_const(self) -> bool:
...
@abc.abstractmethod
def output_type(
self, input_types: dict[ids.ColumnId, dtypes.ExpressionType]
) -> dtypes.ExpressionType:
...
@abc.abstractmethod
def bind_refs(
self,
bindings: Mapping[ids.ColumnId, Expression],
allow_partial_bindings: bool = False,
) -> Expression:
"""Replace variables with expression given in `bindings`.
If allow_partial_bindings is False, validate that all free variables are bound to a new value.
"""
...
@abc.abstractmethod
def bind_variables(
self, bindings: Mapping[str, Expression], allow_partial_bindings: bool = False
) -> Expression:
"""Replace variables with expression given in `bindings`.
If allow_partial_bindings is False, validate that all free variables are bound to a new value.
"""
...
@property
def is_bijective(self) -> bool:
return False
@property
def is_identity(self) -> bool:
"""True for identity operation that does not transform input."""
return False
@dataclasses.dataclass(frozen=True)
class ScalarConstantExpression(Expression):
"""An expression representing a scalar constant."""
# TODO: Further constrain?
value: typing.Hashable
dtype: dtypes.ExpressionType = None
@property
def is_const(self) -> bool:
return True
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return ()
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
return self.dtype
def bind_variables(
self, bindings: Mapping[str, Expression], allow_partial_bindings: bool = False
) -> Expression:
return self
def bind_refs(
self,
bindings: Mapping[ids.ColumnId, Expression],
allow_partial_bindings: bool = False,
) -> ScalarConstantExpression:
return self
@property
def is_bijective(self) -> bool:
# () <-> value
return True
@dataclasses.dataclass(frozen=True)
class UnboundVariableExpression(Expression):
"""A variable expression representing an unbound variable."""
id: str
@property
def free_variables(self) -> typing.Tuple[str, ...]:
return (self.id,)
@property
def is_const(self) -> bool:
return False
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return ()
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
raise ValueError(f"Type of variable {self.id} has not been fixed.")
def bind_refs(
self,
bindings: Mapping[ids.ColumnId, Expression],
allow_partial_bindings: bool = False,
) -> UnboundVariableExpression:
return self
def bind_variables(
self, bindings: Mapping[str, Expression], allow_partial_bindings: bool = False
) -> Expression:
if self.id in bindings.keys():
return bindings[self.id]
elif not allow_partial_bindings:
raise ValueError(f"Variable {self.id} remains unbound")
return self
@property
def is_bijective(self) -> bool:
return True
@property
def is_identity(self) -> bool:
return True
@dataclasses.dataclass(frozen=True)
class DerefOp(Expression):
"""A variable expression representing an unbound variable."""
id: ids.ColumnId
@property
def column_references(self) -> typing.Tuple[ids.ColumnId, ...]:
return (self.id,)
@property
def is_const(self) -> bool:
return False
def output_type(
self, input_types: dict[ids.ColumnId, bigframes.dtypes.Dtype]
) -> dtypes.ExpressionType:
if self.id in input_types:
return input_types[self.id]
else:
raise ValueError(f"Type of variable {self.id} has not been fixed.")
def bind_variables(
self, bindings: Mapping[str, Expression], allow_partial_bindings: bool = False
) -> Expression:
return self
def bind_refs(
self,
bindings: Mapping[ids.ColumnId, Expression],
allow_partial_bindings: bool = False,
) -> Expression:
if self.id in bindings.keys():
return bindings[self.id]
elif not allow_partial_bindings:
raise ValueError(f"Variable {self.id} remains unbound")
return self
@property
def is_bijective(self) -> bool:
return True
@property
def is_identity(self) -> bool:
return True
@dataclasses.dataclass(frozen=True)
class OpExpression(Expression):
"""An expression representing a scalar operation applied to 1 or more argument sub-expressions."""
op: bigframes.operations.RowOp
inputs: typing.Tuple[Expression, ...]
@property
def column_references(
self,
) -> typing.Tuple[bigframes.core.identifiers.ColumnId, ...]:
return tuple(
itertools.chain.from_iterable(
map(lambda x: x.column_references, self.inputs)
)
)
@property
def free_variables(self) -> typing.Tuple[str, ...]:
return tuple(
itertools.chain.from_iterable(map(lambda x: x.free_variables, self.inputs))
)
@property
def is_const(self) -> bool:
return all(child.is_const for child in self.inputs)
def output_type(
self, input_types: dict[ids.ColumnId, dtypes.ExpressionType]
) -> dtypes.ExpressionType:
operand_types = tuple(
map(lambda x: x.output_type(input_types=input_types), self.inputs)
)
return self.op.output_type(*operand_types)
def bind_variables(
self, bindings: Mapping[str, Expression], allow_partial_bindings: bool = False
) -> OpExpression:
return OpExpression(
self.op,
tuple(
input.bind_variables(
bindings, allow_partial_bindings=allow_partial_bindings
)
for input in self.inputs
),
)
def bind_refs(
self,
bindings: Mapping[ids.ColumnId, Expression],
allow_partial_bindings: bool = False,
) -> OpExpression:
return OpExpression(
self.op,
tuple(
input.bind_refs(bindings, allow_partial_bindings=allow_partial_bindings)
for input in self.inputs
),
)
@property
def is_bijective(self) -> bool:
# TODO: Mark individual functions as bijective?
return False