Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
89 changes: 89 additions & 0 deletions bigframes/core/compile/sqlglot/expressions/unary_compiler.py
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,8 @@

import typing

import pandas as pd
import pyarrow as pa
import sqlglot
import sqlglot.expressions as sge

Expand Down Expand Up @@ -105,6 +107,12 @@ def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
)


@UNARY_OP_REGISTRATION.register(ops.AsTypeOp)
def _(op: ops.AsTypeOp, expr: TypedExpr) -> sge.Expression:
# TODO: Support more types for casting, such as JSON, etc.
return sge.Cast(this=expr.expr, to=op.to_type)


@UNARY_OP_REGISTRATION.register(ops.ArrayToStringOp)
def _(op: ops.ArrayToStringOp, expr: TypedExpr) -> sge.Expression:
return sge.ArrayToString(this=expr.expr, expression=f"'{op.delimiter}'")
Expand Down Expand Up @@ -234,6 +242,12 @@ def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
) - sge.convert(1)


@UNARY_OP_REGISTRATION.register(ops.FloorDtOp)
def _(op: ops.FloorDtOp, expr: TypedExpr) -> sge.Expression:
# TODO: Remove this method when it is covered by ops.FloorOp
return sge.TimestampTrunc(this=expr.expr, unit=sge.Identifier(this=op.freq))


@UNARY_OP_REGISTRATION.register(ops.floor_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.Floor(this=expr.expr)
Expand All @@ -249,6 +263,26 @@ def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("ST_ASTEXT", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.geo_st_boundary_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("ST_BOUNDARY", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.geo_st_geogfromtext_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("SAFE.ST_GEOGFROMTEXT", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.geo_st_isclosed_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("ST_ISCLOSED", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.GeoStLengthOp)
def _(op: ops.GeoStLengthOp, expr: TypedExpr) -> sge.Expression:
return sge.func("ST_LENGTH", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.geo_x_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("SAFE.ST_X", expr.expr)
Expand All @@ -274,6 +308,11 @@ def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.BitwiseNot(this=expr.expr)


@UNARY_OP_REGISTRATION.register(ops.IsInOp)
def _(op: ops.IsInOp, expr: TypedExpr) -> sge.Expression:
return sge.In(this=expr.expr, expressions=[sge.convert(v) for v in op.values])


@UNARY_OP_REGISTRATION.register(ops.isalnum_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.RegexpLike(this=expr.expr, expression=sge.convert(r"^(\p{N}|\p{L})+$"))
Expand Down Expand Up @@ -517,6 +556,26 @@ def _(op: ops.StrSliceOp, expr: TypedExpr) -> sge.Expression:
)


@UNARY_OP_REGISTRATION.register(ops.StrftimeOp)
def _(op: ops.StrftimeOp, expr: TypedExpr) -> sge.Expression:
return sge.func("FORMAT_TIMESTAMP", sge.convert(op.date_format), expr.expr)


@UNARY_OP_REGISTRATION.register(ops.StructFieldOp)
def _(op: ops.StructFieldOp, expr: TypedExpr) -> sge.Expression:
if isinstance(op.name_or_index, str):
name = op.name_or_index
else:
pa_type = typing.cast(pd.ArrowDtype, expr.dtype)
pa_struct_type = typing.cast(pa.StructType, pa_type.pyarrow_dtype)
name = pa_struct_type.field(op.name_or_index).name

return sge.Column(
this=sge.to_identifier(name, quoted=True),
catalog=expr.expr,
)


@UNARY_OP_REGISTRATION.register(ops.tan_op)
def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.func("TAN", expr.expr)
Expand All @@ -537,6 +596,36 @@ def _(op: ops.base_ops.UnaryOp, expr: TypedExpr) -> sge.Expression:
return sge.Floor(this=expr.expr)


@UNARY_OP_REGISTRATION.register(ops.ToDatetimeOp)
def _(op: ops.ToDatetimeOp, expr: TypedExpr) -> sge.Expression:
return sge.Cast(this=sge.func("TIMESTAMP_SECONDS", expr.expr), to="DATETIME")


@UNARY_OP_REGISTRATION.register(ops.ToTimestampOp)
def _(op: ops.ToTimestampOp, expr: TypedExpr) -> sge.Expression:
return sge.func("TIMESTAMP_SECONDS", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.ToTimedeltaOp)
def _(op: ops.ToTimedeltaOp, expr: TypedExpr) -> sge.Expression:
return sge.Interval(this=expr.expr, unit=sge.Identifier(this="SECOND"))


@UNARY_OP_REGISTRATION.register(ops.UnixMicros)
def _(op: ops.UnixMicros, expr: TypedExpr) -> sge.Expression:
return sge.func("UNIX_MICROS", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.UnixMillis)
def _(op: ops.UnixMillis, expr: TypedExpr) -> sge.Expression:
return sge.func("UNIX_MILLIS", expr.expr)


@UNARY_OP_REGISTRATION.register(ops.UnixSeconds)
def _(op: ops.UnixSeconds, expr: TypedExpr) -> sge.Expression:
return sge.func("UNIX_SECONDS", expr.expr)


# JSON Ops
@UNARY_OP_REGISTRATION.register(ops.JSONExtract)
def _(op: ops.JSONExtract, expr: TypedExpr) -> sge.Expression:
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`timestamp_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
TIMESTAMP_TRUNC(`bfcol_0`, DAY) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `timestamp_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`geography_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
ST_BOUNDARY(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `geography_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`string_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
SAFE.ST_GEOGFROMTEXT(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `string_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`geography_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
ST_ISCLOSED(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `geography_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`geography_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
ST_LENGTH(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `geography_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`int64_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
`bfcol_0` IN (1, 2, 3) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `int64_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`timestamp_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
FORMAT_TIMESTAMP('%Y-%m-%d', `bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `timestamp_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`people` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`nested_structs_types`
), `bfcte_1` AS (
SELECT
*,
`bfcol_0`.`name` AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `people`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`int64_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
CAST(TIMESTAMP_SECONDS(`bfcol_0`) AS DATETIME) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `int64_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`int64_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
INTERVAL `bfcol_0` SECOND AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `int64_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`int64_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
TIMESTAMP_SECONDS(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `int64_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`timestamp_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
UNIX_MICROS(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `timestamp_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`timestamp_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
UNIX_MILLIS(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `timestamp_col`
FROM `bfcte_1`
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
WITH `bfcte_0` AS (
SELECT
`timestamp_col` AS `bfcol_0`
FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
), `bfcte_1` AS (
SELECT
*,
UNIX_SECONDS(`bfcol_0`) AS `bfcol_1`
FROM `bfcte_0`
)
SELECT
`bfcol_1` AS `timestamp_col`
FROM `bfcte_1`
Loading