Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Appearance settings
This repository was archived by the owner on May 7, 2026. It is now read-only.

Commit ce5de57

Browse filesBrowse files
TrevorBergeronTrevorBergeron
authored
feat: Add str, dt accessors to pd.col Expression objects (#2488)
1 parent 47a0feb commit ce5de57
Copy full SHA for ce5de57

8 files changed

+174-109Lines changed: 174 additions & 109 deletions

File tree

Expand file treeCollapse file tree
Open diff view settings
Filter options
Expand file treeCollapse file tree
Open diff view settings
Collapse file

‎bigframes/core/col.py‎

Copy file name to clipboardExpand all lines: bigframes/core/col.py
+51-28Lines changed: 51 additions & 28 deletions
Original file line numberDiff line numberDiff line change
@@ -14,7 +14,7 @@
1414
from __future__ import annotations
1515

1616
import dataclasses
17-
from typing import Any, Hashable
17+
from typing import Any, Hashable, Literal, TYPE_CHECKING
1818

1919
import bigframes_vendored.pandas.core.col as pd_col
2020

@@ -23,6 +23,10 @@
2323
import bigframes.operations as bf_ops
2424
import bigframes.operations.aggregations as agg_ops
2525

26+
if TYPE_CHECKING:
27+
import bigframes.operations.datetimes as datetimes
28+
import bigframes.operations.strings as strings
29+
2630

2731
# Not to be confused with the Expression class in `bigframes.core.expressions`
2832
# Name collision unintended
@@ -32,7 +36,7 @@ class Expression:
3236

3337
_value: bf_expression.Expression
3438

35-
def _apply_unary(self, op: bf_ops.UnaryOp) -> Expression:
39+
def _apply_unary_op(self, op: bf_ops.UnaryOp) -> Expression:
3640
return Expression(op.as_expr(self._value))
3741

3842
def _apply_unary_agg(self, op: agg_ops.UnaryAggregateOp) -> Expression:
@@ -44,7 +48,14 @@ def _apply_unary_agg(self, op: agg_ops.UnaryAggregateOp) -> Expression:
4448
agg_expressions.WindowExpression(agg_expr, window_spec.unbound())
4549
)
4650

47-
def _apply_binary(self, other: Any, op: bf_ops.BinaryOp, reverse: bool = False):
51+
# alignment is purely for series compatibility, and is ignored here
52+
def _apply_binary_op(
53+
self,
54+
other: Any,
55+
op: bf_ops.BinaryOp,
56+
alignment: Literal["outer", "left"] = "outer",
57+
reverse: bool = False,
58+
):
4859
if isinstance(other, Expression):
4960
other_value = other._value
5061
else:
@@ -55,79 +66,79 @@ def _apply_binary(self, other: Any, op: bf_ops.BinaryOp, reverse: bool = False):
5566
return Expression(op.as_expr(self._value, other_value))
5667

5768
def __add__(self, other: Any) -> Expression:
58-
return self._apply_binary(other, bf_ops.add_op)
69+
return self._apply_binary_op(other, bf_ops.add_op)
5970

6071
def __radd__(self, other: Any) -> Expression:
61-
return self._apply_binary(other, bf_ops.add_op, reverse=True)
72+
return self._apply_binary_op(other, bf_ops.add_op, reverse=True)
6273

6374
def __sub__(self, other: Any) -> Expression:
64-
return self._apply_binary(other, bf_ops.sub_op)
75+
return self._apply_binary_op(other, bf_ops.sub_op)
6576

6677
def __rsub__(self, other: Any) -> Expression:
67-
return self._apply_binary(other, bf_ops.sub_op, reverse=True)
78+
return self._apply_binary_op(other, bf_ops.sub_op, reverse=True)
6879

6980
def __mul__(self, other: Any) -> Expression:
70-
return self._apply_binary(other, bf_ops.mul_op)
81+
return self._apply_binary_op(other, bf_ops.mul_op)
7182

7283
def __rmul__(self, other: Any) -> Expression:
73-
return self._apply_binary(other, bf_ops.mul_op, reverse=True)
84+
return self._apply_binary_op(other, bf_ops.mul_op, reverse=True)
7485

7586
def __truediv__(self, other: Any) -> Expression:
76-
return self._apply_binary(other, bf_ops.div_op)
87+
return self._apply_binary_op(other, bf_ops.div_op)
7788

7889
def __rtruediv__(self, other: Any) -> Expression:
79-
return self._apply_binary(other, bf_ops.div_op, reverse=True)
90+
return self._apply_binary_op(other, bf_ops.div_op, reverse=True)
8091

8192
def __floordiv__(self, other: Any) -> Expression:
82-
return self._apply_binary(other, bf_ops.floordiv_op)
93+
return self._apply_binary_op(other, bf_ops.floordiv_op)
8394

8495
def __rfloordiv__(self, other: Any) -> Expression:
85-
return self._apply_binary(other, bf_ops.floordiv_op, reverse=True)
96+
return self._apply_binary_op(other, bf_ops.floordiv_op, reverse=True)
8697

8798
def __ge__(self, other: Any) -> Expression:
88-
return self._apply_binary(other, bf_ops.ge_op)
99+
return self._apply_binary_op(other, bf_ops.ge_op)
89100

90101
def __gt__(self, other: Any) -> Expression:
91-
return self._apply_binary(other, bf_ops.gt_op)
102+
return self._apply_binary_op(other, bf_ops.gt_op)
92103

93104
def __le__(self, other: Any) -> Expression:
94-
return self._apply_binary(other, bf_ops.le_op)
105+
return self._apply_binary_op(other, bf_ops.le_op)
95106

96107
def __lt__(self, other: Any) -> Expression:
97-
return self._apply_binary(other, bf_ops.lt_op)
108+
return self._apply_binary_op(other, bf_ops.lt_op)
98109

99110
def __eq__(self, other: object) -> Expression: # type: ignore
100-
return self._apply_binary(other, bf_ops.eq_op)
111+
return self._apply_binary_op(other, bf_ops.eq_op)
101112

102113
def __ne__(self, other: object) -> Expression: # type: ignore
103-
return self._apply_binary(other, bf_ops.ne_op)
114+
return self._apply_binary_op(other, bf_ops.ne_op)
104115

105116
def __mod__(self, other: Any) -> Expression:
106-
return self._apply_binary(other, bf_ops.mod_op)
117+
return self._apply_binary_op(other, bf_ops.mod_op)
107118

108119
def __rmod__(self, other: Any) -> Expression:
109-
return self._apply_binary(other, bf_ops.mod_op, reverse=True)
120+
return self._apply_binary_op(other, bf_ops.mod_op, reverse=True)
110121

111122
def __and__(self, other: Any) -> Expression:
112-
return self._apply_binary(other, bf_ops.and_op)
123+
return self._apply_binary_op(other, bf_ops.and_op)
113124

114125
def __rand__(self, other: Any) -> Expression:
115-
return self._apply_binary(other, bf_ops.and_op, reverse=True)
126+
return self._apply_binary_op(other, bf_ops.and_op, reverse=True)
116127

117128
def __or__(self, other: Any) -> Expression:
118-
return self._apply_binary(other, bf_ops.or_op)
129+
return self._apply_binary_op(other, bf_ops.or_op)
119130

120131
def __ror__(self, other: Any) -> Expression:
121-
return self._apply_binary(other, bf_ops.or_op, reverse=True)
132+
return self._apply_binary_op(other, bf_ops.or_op, reverse=True)
122133

123134
def __xor__(self, other: Any) -> Expression:
124-
return self._apply_binary(other, bf_ops.xor_op)
135+
return self._apply_binary_op(other, bf_ops.xor_op)
125136

126137
def __rxor__(self, other: Any) -> Expression:
127-
return self._apply_binary(other, bf_ops.xor_op, reverse=True)
138+
return self._apply_binary_op(other, bf_ops.xor_op, reverse=True)
128139

129140
def __invert__(self) -> Expression:
130-
return self._apply_unary(bf_ops.invert_op)
141+
return self._apply_unary_op(bf_ops.invert_op)
131142

132143
def sum(self) -> Expression:
133144
return self._apply_unary_agg(agg_ops.sum_op)
@@ -147,6 +158,18 @@ def min(self) -> Expression:
147158
def max(self) -> Expression:
148159
return self._apply_unary_agg(agg_ops.max_op)
149160

161+
@property
162+
def dt(self) -> datetimes.DatetimeSimpleMethods:
163+
import bigframes.operations.datetimes as datetimes
164+
165+
return datetimes.DatetimeSimpleMethods(self)
166+
167+
@property
168+
def str(self) -> strings.StringMethods:
169+
import bigframes.operations.strings as strings
170+
171+
return strings.StringMethods(self)
172+
150173

151174
def col(col_name: Hashable) -> Expression:
152175
return Expression(bf_expression.free_var(col_name))
Collapse file

‎bigframes/core/compile/ibis_compiler/scalar_op_registry.py‎

Copy file name to clipboardExpand all lines: bigframes/core/compile/ibis_compiler/scalar_op_registry.py
+3-1Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -982,7 +982,9 @@ def isin_op_impl(x: ibis_types.Value, op: ops.IsInOp):
982982

983983
@scalar_op_compiler.register_unary_op(ops.ToDatetimeOp, pass_op=True)
984984
def to_datetime_op_impl(x: ibis_types.Value, op: ops.ToDatetimeOp):
985-
if x.type() in (ibis_dtypes.str, ibis_dtypes.Timestamp("UTC")): # type: ignore
985+
if x.type() == ibis_dtypes.Timestamp(None): # type: ignore
986+
return x # already a timestamp, no-op
987+
elif x.type() in (ibis_dtypes.str, ibis_dtypes.Timestamp("UTC")): # type: ignore
986988
return x.try_cast(ibis_dtypes.Timestamp(None)) # type: ignore
987989
else:
988990
# Numerical inputs.
Collapse file

‎bigframes/operations/datetime_ops.py‎

Copy file name to clipboardExpand all lines: bigframes/operations/datetime_ops.py
+3Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -74,6 +74,7 @@ def output_type(self, *input_types: dtypes.ExpressionType) -> dtypes.ExpressionT
7474
dtypes.STRING_DTYPE,
7575
dtypes.DATE_DTYPE,
7676
dtypes.TIMESTAMP_DTYPE,
77+
dtypes.DATETIME_DTYPE,
7778
):
7879
raise TypeError("expected string or numeric input")
7980
return pd.ArrowDtype(pa.timestamp("us", tz=None))
@@ -87,6 +88,8 @@ class ToTimestampOp(base_ops.UnaryOp):
8788

8889
def output_type(self, *input_types: dtypes.ExpressionType) -> dtypes.ExpressionType:
8990
# Must be numeric or string
91+
if input_types[0] == dtypes.TIMESTAMP_DTYPE:
92+
raise TypeError("Already tz-aware.")
9093
if input_types[0] not in (
9194
dtypes.FLOAT_DTYPE,
9295
dtypes.INT_DTYPE,

0 commit comments

Comments
0 (0)
Morty Proxy This is a proxified and sanitized view of the page, visit original site.