This repository was archived by the owner on May 7, 2026. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 68
Expand file tree
/
Copy pathas_sql.py
More file actions
306 lines (263 loc) · 10.3 KB
/
Copy pathas_sql.py
File metadata and controls
306 lines (263 loc) · 10.3 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
# Copyright 2026 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 dataclasses
import itertools
from typing import Optional, Sequence, Union
from bigframes.core import (
agg_expressions,
expression,
guid,
identifiers,
nodes,
ordering,
sql_nodes,
)
import bigframes.core.rewrite
def _limit(select: sql_nodes.SqlSelectNode, limit: int) -> sql_nodes.SqlSelectNode:
new_limit = limit if select.limit is None else min([select.limit, limit])
return dataclasses.replace(select, limit=new_limit)
def _try_sort(
select: sql_nodes.SqlSelectNode, sort_by: Sequence[ordering.OrderingExpression]
) -> Optional[sql_nodes.SqlSelectNode]:
new_order_exprs = []
for sort_expr in sort_by:
new_expr = _try_bind(
sort_expr.scalar_expression, select.get_id_mapping(), analytic_allowed=False
)
if new_expr is None:
return None
new_order_exprs.append(
dataclasses.replace(sort_expr, scalar_expression=new_expr)
)
return dataclasses.replace(select, sorting=tuple(new_order_exprs))
def _sort(
node: nodes.BigFrameNode, sort_by: Sequence[ordering.OrderingExpression]
) -> sql_nodes.SqlSelectNode:
if isinstance(node, sql_nodes.SqlSelectNode):
merged = _try_sort(node, sort_by)
if merged:
return merged
result = _try_sort(_create_noop_select(node), sort_by)
assert result is not None
return result
def _try_bind(
expr: expression.Expression,
bindings: dict[identifiers.ColumnId, expression.Expression],
analytic_allowed: bool = False, # means block binding to an analytic even if original is scalar
) -> Optional[expression.Expression]:
if not expr.is_scalar_expr or not analytic_allowed:
for ref in expr.column_references:
if ref in bindings and not bindings[ref].is_scalar_expr:
return None
return expr.bind_refs(bindings)
def _try_add_cdefs(
select: sql_nodes.SqlSelectNode, cdefs: Sequence[nodes.ColumnDef]
) -> Optional[sql_nodes.SqlSelectNode]:
# TODO: add up complexity measure while inlining refs
new_defs = []
for cdef in cdefs:
cdef_expr = cdef.expression
merged_expr = _try_bind(
cdef_expr, select.get_id_mapping(), analytic_allowed=True
)
if merged_expr is None:
return None
new_defs.append(nodes.ColumnDef(merged_expr, cdef.id))
return dataclasses.replace(select, selections=(*select.selections, *new_defs))
def _add_cdefs(
node: nodes.BigFrameNode, cdefs: Sequence[nodes.ColumnDef]
) -> sql_nodes.SqlSelectNode:
if isinstance(node, sql_nodes.SqlSelectNode):
merged = _try_add_cdefs(node, cdefs)
if merged:
return merged
# Otherwise, wrap the child in a SELECT and add the columns
result = _try_add_cdefs(_create_noop_select(node), cdefs)
assert result is not None
return result
def _try_add_filter(
select: sql_nodes.SqlSelectNode, predicates: Sequence[expression.Expression]
) -> Optional[sql_nodes.SqlSelectNode]:
# Filter implicitly happens first, so merging it into ths select will modify non-scalar col expressions
if not all(cdef.expression.is_scalar_expr for cdef in select.selections):
return None
if not all(
sort_expr.scalar_expression.is_scalar_expr for sort_expr in select.sorting
):
return None
# Constraint: filters can only be merged if they are scalar expression after binding
new_predicates = []
# bind variables, merge predicates
for predicate in predicates:
merged_pred = _try_bind(predicate, select.get_id_mapping())
if not merged_pred:
return None
new_predicates.append(merged_pred)
return dataclasses.replace(select, predicates=(*select.predicates, *new_predicates))
def _add_filter(
node: nodes.BigFrameNode, predicates: Sequence[expression.Expression]
) -> sql_nodes.SqlSelectNode:
if isinstance(node, sql_nodes.SqlSelectNode):
result = _try_add_filter(node, predicates)
if result:
return result
new_node = _try_add_filter(_create_noop_select(node), predicates)
assert new_node is not None
return new_node
def _create_noop_select(node: nodes.BigFrameNode) -> sql_nodes.SqlSelectNode:
return sql_nodes.SqlSelectNode(
node,
selections=tuple(
nodes.ColumnDef(expression.ResolvedDerefOp.from_field(field), field.id)
for field in node.fields
),
)
def _try_remap_select_cols(
select: sql_nodes.SqlSelectNode, cols: Sequence[nodes.AliasedRef]
):
new_defs = []
for aliased_ref in cols:
new_defs.append(
nodes.ColumnDef(select.get_id_mapping()[aliased_ref.ref.id], aliased_ref.id)
)
return dataclasses.replace(select, selections=tuple(new_defs))
def _remap_select_cols(node: nodes.BigFrameNode, cols: Sequence[nodes.AliasedRef]):
if isinstance(node, sql_nodes.SqlSelectNode):
result = _try_remap_select_cols(node, cols)
if result:
return result
new_node = _try_remap_select_cols(_create_noop_select(node), cols)
assert new_node is not None
return new_node
def _get_added_cdefs(node: Union[nodes.ProjectionNode, nodes.WindowOpNode]):
# TODO: InNode
if isinstance(node, nodes.ProjectionNode):
return tuple(nodes.ColumnDef(expr, id) for expr, id in node.assignments)
if isinstance(node, nodes.WindowOpNode):
new_cdefs = []
for cdef in node.agg_exprs:
assert isinstance(cdef.expression, agg_expressions.Aggregation)
window_expr = agg_expressions.WindowExpression(
cdef.expression, node.window_spec
)
# TODO: we probably should do this as another step
rewritten_window_expr = bigframes.core.rewrite.simplify_complex_windows(
window_expr
)
new_cdefs.append(nodes.ColumnDef(rewritten_window_expr, cdef.id))
return tuple(new_cdefs)
else:
raise ValueError(f"Unexpected node type: {type(node)}")
def _as_sql_node(node: nodes.BigFrameNode) -> nodes.BigFrameNode:
# case one, can be converted to select
if isinstance(node, nodes.ReadTableNode):
leaf = sql_nodes.SqlDataSource(source=node.source)
mappings = [
nodes.AliasedRef(expression.deref(scan_item.source_id), scan_item.id)
for scan_item in node.scan_list.items
]
return _remap_select_cols(leaf, mappings)
elif isinstance(node, (nodes.ProjectionNode, nodes.WindowOpNode)):
cdefs = _get_added_cdefs(node)
return _add_cdefs(node.child, cdefs)
elif isinstance(node, (nodes.SelectionNode)):
return _remap_select_cols(node.child, node.input_output_pairs)
elif isinstance(node, nodes.FilterNode):
return _add_filter(node.child, [node.predicate])
elif isinstance(node, nodes.ResultNode):
result = node.child
if node.order_by is not None:
result = _sort(result, node.order_by.all_ordering_columns)
result = _remap_select_cols(
result,
[
nodes.AliasedRef(ref, identifiers.ColumnId(name))
for ref, name in node.output_cols
],
)
if node.limit is not None:
result = _limit(result, node.limit) # type: ignore
return result
else:
return node
# In the future, we will have sql nodes for each of these node types.
_LOGICAL_NODE_TYPES_TO_WRAP = (
nodes.ReadLocalNode,
nodes.ExplodeNode,
nodes.InNode,
nodes.AggregateNode,
nodes.FromRangeNode,
nodes.ConcatNode,
sql_nodes.SqlSelectNode,
)
def _insert_cte_markers(root: nodes.BigFrameNode) -> nodes.BigFrameNode:
# important not to wrap nodes that are already wrapped
wrapped_nodes = set(
node.child for node in root.unique_nodes() if isinstance(node, nodes.CteNode)
)
# don't wrap child nodes of ConcatNode
union_child_nodes = set(
itertools.chain.from_iterable(
node.child_nodes
for node in root.unique_nodes()
if isinstance(node, nodes.ConcatNode)
)
)
def maybe_insert_cte_marker(node: nodes.BigFrameNode) -> nodes.BigFrameNode:
if node == root:
return node
if (
isinstance(node, _LOGICAL_NODE_TYPES_TO_WRAP)
and node not in wrapped_nodes
and node not in union_child_nodes
):
wrapped_nodes.add(node)
return nodes.CteNode(node)
return node
return root.top_down(maybe_insert_cte_marker)
def _extract_ctes_to_with_expr(
root: nodes.BigFrameNode, uid_gen: guid.SequentialUIDGenerator
) -> nodes.BigFrameNode:
topological_ctes = list(
filter(
lambda n: isinstance(n, nodes.CteNode),
root.iter_nodes_topo(),
)
)
cte_names = tuple(
next(uid_gen.get_uid_stream("bfcte_")) for _ in range(len(topological_ctes))
)
if len(topological_ctes) == 0:
return root
mapping = {
cte_node: sql_nodes.SqlCteRefNode(cte_name, tuple(cte_node.fields))
for cte_node, cte_name in zip(topological_ctes, cte_names)
}
# Replace all CTEs with CTE references and wrap the new root in a WITH clause
return sql_nodes.SqlWithCtesNode(
root.top_down(lambda x: mapping.get(x, x)),
cte_names,
tuple(
cte_node.child.top_down(lambda x: mapping.get(x, x)) for cte_node in topological_ctes # type: ignore
),
)
def as_sql_nodes(
root: nodes.BigFrameNode, uid_gen: guid.SequentialUIDGenerator
) -> nodes.BigFrameNode:
root = nodes.bottom_up(root, _as_sql_node)
# Insert CTE markers to indicate where we want to split the query.
root = _insert_cte_markers(root)
root = _extract_ctes_to_with_expr(root, uid_gen)
return root