Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[SPARK-47500][PYTHON][CONNECT] Factor column name handling out of plan.py #45636

Closed
wants to merge 1 commit into from
Closed

[SPARK-47500][PYTHON][CONNECT] Factor column name handling out of plan.py #45636

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
70 changes: 37 additions & 33 deletions python/pyspark/sql/connect/dataframe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,8 +182,10 @@ def isEmpty(self) -> bool:
def select(self, *cols: "ColumnOrName") -> "DataFrame":
if len(cols) == 1 and isinstance(cols[0], list):
cols = cols[0]

return DataFrame(plan.Project(self._plan, *cols), session=self._session)
return DataFrame(
plan.Project(self._plan, [F._to_col(c) for c in cols]),
session=self._session,
)

select.__doc__ = PySparkDataFrame.select.__doc__

Expand All @@ -197,7 +199,7 @@ def selectExpr(self, *expr: Union[str, List[str]]) -> "DataFrame":
else:
sql_expr.extend([F.expr(e) for e in element])

return DataFrame(plan.Project(self._plan, *sql_expr), session=self._session)
return DataFrame(plan.Project(self._plan, sql_expr), session=self._session)

selectExpr.__doc__ = PySparkDataFrame.selectExpr.__doc__

Expand Down Expand Up @@ -309,18 +311,20 @@ def repartition( # type: ignore[misc]
)
if len(cols) == 0:
return DataFrame(
plan.Repartition(self._plan, num_partitions=numPartitions, shuffle=True),
plan.Repartition(self._plan, numPartitions, shuffle=True),
self._session,
)
else:
return DataFrame(
plan.RepartitionByExpression(self._plan, numPartitions, list(cols)),
plan.RepartitionByExpression(
self._plan, numPartitions, [F._to_col(c) for c in cols]
),
self.sparkSession,
)
elif isinstance(numPartitions, (str, Column)):
cols = (numPartitions,) + cols
return DataFrame(
plan.RepartitionByExpression(self._plan, None, list(cols)),
plan.RepartitionByExpression(self._plan, None, [F._to_col(c) for c in cols]),
self.sparkSession,
)
else:
Expand All @@ -345,14 +349,14 @@ def repartitionByRange(self, *cols: "ColumnOrName") -> "DataFrame":
def repartitionByRange( # type: ignore[misc]
self, numPartitions: Union[int, "ColumnOrName"], *cols: "ColumnOrName"
) -> "DataFrame":
def _convert_col(col: "ColumnOrName") -> "ColumnOrName":
def _convert_col(col: "ColumnOrName") -> Column:
if isinstance(col, Column):
if isinstance(col._expr, SortOrder):
return col
else:
return Column(SortOrder(col._expr))
return col.asc()
else:
return Column(SortOrder(ColumnReference(col)))
return F.col(col).asc()

if isinstance(numPartitions, int):
if not numPartitions > 0:
Expand All @@ -369,18 +373,17 @@ def _convert_col(col: "ColumnOrName") -> "ColumnOrName":
message_parameters={"item": "cols"},
)
else:
sort = []
sort.extend([_convert_col(c) for c in cols])
return DataFrame(
plan.RepartitionByExpression(self._plan, numPartitions, sort),
plan.RepartitionByExpression(
self._plan, numPartitions, [_convert_col(c) for c in cols]
),
self.sparkSession,
)
elif isinstance(numPartitions, (str, Column)):
cols = (numPartitions,) + cols
sort = []
sort.extend([_convert_col(c) for c in cols])
return DataFrame(
plan.RepartitionByExpression(self._plan, None, sort),
plan.RepartitionByExpression(
self._plan, None, [_convert_col(c) for c in [numPartitions] + list(cols)]
),
self.sparkSession,
)
else:
Expand Down Expand Up @@ -648,12 +651,18 @@ def _joinAsOf(
if tolerance is not None:
assert isinstance(tolerance, Column), "tolerance should be Column"

def _convert_col(df: "DataFrame", col: "ColumnOrName") -> Column:
if isinstance(col, Column):
return col
else:
return Column(ColumnReference(col, df._plan._plan_id))

return DataFrame(
plan.AsOfJoin(
left=self._plan,
right=other._plan,
left_as_of=leftAsOfColumn,
right_as_of=rightAsOfColumn,
left_as_of=_convert_col(self, leftAsOfColumn),
right_as_of=_convert_col(other, rightAsOfColumn),
on=on,
how=how,
tolerance=tolerance,
Expand Down Expand Up @@ -940,24 +949,21 @@ def unpivot(
) -> "DataFrame":
assert ids is not None, "ids must not be None"

def to_jcols(
def _convert_cols(
cols: Optional[Union["ColumnOrName", List["ColumnOrName"], Tuple["ColumnOrName", ...]]]
) -> List["ColumnOrName"]:
) -> List[Column]:
if cols is None:
lst = []
elif isinstance(cols, tuple):
lst = list(cols)
elif isinstance(cols, list):
lst = cols
return []
elif isinstance(cols, (tuple, list)):
return [F._to_col(c) for c in cols]
else:
lst = [cols]
return lst
return [F._to_col(cols)]

return DataFrame(
plan.Unpivot(
self._plan,
to_jcols(ids),
to_jcols(values) if values is not None else None,
_convert_cols(ids),
_convert_cols(values) if values is not None else None,
variableColumnName,
valueColumnName,
),
Expand Down Expand Up @@ -1645,9 +1651,7 @@ def freqItems(
def sampleBy(
self, col: "ColumnOrName", fractions: Dict[Any, float], seed: Optional[int] = None
) -> "DataFrame":
if isinstance(col, str):
col = Column(ColumnReference(col))
elif not isinstance(col, Column):
if not isinstance(col, (str, Column)):
raise PySparkTypeError(
error_class="NOT_COLUMN_OR_STR",
message_parameters={"arg_name": "col", "arg_type": type(col).__name__},
Expand All @@ -1671,7 +1675,7 @@ def sampleBy(
fractions[k] = float(v)
seed = seed if seed is not None else random.randint(0, sys.maxsize)
return DataFrame(
plan.StatSampleBy(child=self._plan, col=col, fractions=fractions, seed=seed),
plan.StatSampleBy(child=self._plan, col=F._to_col(col), fractions=fractions, seed=seed),
session=self._session,
)

Expand Down
Loading