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

feat: Add SQL support for NULLS FIRST/LAST ordering #16711

Merged
merged 2 commits into from
Jun 5, 2024
Merged

feat: Add SQL support for NULLS FIRST/LAST ordering #16711

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a561f7c
feat: Add SQL support for `NULLS FIRST/LAST` ordering directives
alexander-beedie Jun 2, 2024
1ba22f1
rebase (again)
alexander-beedie Jun 4, 2024
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
13 changes: 7 additions & 6 deletions crates/polars-sql/src/context.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -849,21 +849,22 @@ impl SQLContext {
) -> PolarsResult<LazyFrame> {
let mut by = Vec::with_capacity(ob.len());
let mut descending = Vec::with_capacity(ob.len());
let mut nulls_last = Vec::with_capacity(ob.len());

let schema = Some(lf.schema_with_arenas(&mut self.lp_arena, &mut self.expr_arena)?);
for ob in ob {
// note: if not specified 'NULLS FIRST' is default for DESC, 'NULLS LAST' otherwise
// https://www.postgresql.org/docs/current/queries-order.html
by.push(parse_sql_expr(&ob.expr, self, schema.as_deref())?);
descending.push(!ob.asc.unwrap_or(true));
polars_ensure!(
ob.nulls_first.is_none(),
SQLInterface: "NULLS FIRST/LAST is not (yet) supported",
);
let desc_order = !ob.asc.unwrap_or(true);
nulls_last.push(!ob.nulls_first.unwrap_or(desc_order));
descending.push(desc_order);
}

Ok(lf.sort_by_exprs(
&by,
SortMultipleOptions::default()
.with_order_descending_multi(descending)
.with_nulls_last_multi(nulls_last)
.with_maintain_order(true),
))
}
Expand Down
167 changes: 3 additions & 164 deletions py-polars/tests/unit/sql/test_miscellaneous.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
import pytest

import polars as pl
from polars.exceptions import SQLInterfaceError
from polars.testing import assert_frame_equal


Expand Down Expand Up @@ -81,7 +82,7 @@ def test_distinct() -> None:

# test unregistration
ctx.unregister("df")
with pytest.raises(pl.SQLInterfaceError, match="relation 'df' was not found"):
with pytest.raises(SQLInterfaceError, match="relation 'df' was not found"):
ctx.execute("SELECT * FROM df")


Expand All @@ -94,7 +95,7 @@ def test_frame_sql_globals_error() -> None:
FROM df2 JOIN df1 ON df1.a = df2.a
ORDER BY b DESC
"""
with pytest.raises(pl.SQLInterfaceError, match=".*not found.*"):
with pytest.raises(SQLInterfaceError, match="relation.*not found.*"):
df1.sql(query=query)

res = pl.sql(query=query, eager=True)
Expand Down Expand Up @@ -134,168 +135,6 @@ def test_limit_offset() -> None:
assert len(out) == min(limit, n_values - offset)


def test_order_by(foods_ipc_path: Path) -> None:
foods = pl.scan_ipc(foods_ipc_path)
nums = pl.LazyFrame({"x": [1, 2, 3], "y": [4, 3, 2]})

order_by_distinct_res = pl.SQLContext(foods1=foods).execute(
"""
SELECT DISTINCT category
FROM foods1
ORDER BY category DESC
""",
eager=True,
)
assert order_by_distinct_res.to_dict(as_series=False) == {
"category": ["vegetables", "seafood", "meat", "fruit"]
}

order_by_group_by_res = pl.SQLContext(foods1=foods).execute(
"""
SELECT category
FROM foods1
GROUP BY category
ORDER BY category DESC
""",
eager=True,
)
assert order_by_group_by_res.to_dict(as_series=False) == {
"category": ["vegetables", "seafood", "meat", "fruit"]
}

order_by_constructed_group_by_res = pl.SQLContext(foods1=foods).execute(
"""
SELECT category, SUM(calories) as summed_calories
FROM foods1
GROUP BY category
ORDER BY summed_calories DESC
""",
eager=True,
)
assert order_by_constructed_group_by_res.to_dict(as_series=False) == {
"category": ["seafood", "meat", "fruit", "vegetables"],
"summed_calories": [1250, 540, 410, 192],
}

order_by_unselected_res = pl.SQLContext(foods1=foods).execute(
"""
SELECT SUM(calories) as summed_calories
FROM foods1
GROUP BY category
ORDER BY summed_calories DESC
""",
eager=True,
)
assert order_by_unselected_res.to_dict(as_series=False) == {
"summed_calories": [1250, 540, 410, 192],
}

order_by_unselected_nums_res = pl.SQLContext(df=nums).execute(
"""
SELECT
df.x,
df.y as y_alias
FROM df
ORDER BY y_alias
""",
eager=True,
)
assert order_by_unselected_nums_res.to_dict(as_series=False) == {
"x": [3, 2, 1],
"y_alias": [2, 3, 4],
}

order_by_wildcard_res = pl.SQLContext(df=nums).execute(
"""
SELECT
*,
df.y as y_alias
FROM df
ORDER BY y
""",
eager=True,
)
assert order_by_wildcard_res.to_dict(as_series=False) == {
"x": [3, 2, 1],
"y": [2, 3, 4],
"y_alias": [2, 3, 4],
}

order_by_qualified_wildcard_res = pl.SQLContext(df=nums).execute(
"""
SELECT
df.*
FROM df
ORDER BY y
""",
eager=True,
)
assert order_by_qualified_wildcard_res.to_dict(as_series=False) == {
"x": [3, 2, 1],
"y": [2, 3, 4],
}

order_by_exclude_res = pl.SQLContext(df=nums).execute(
"""
SELECT
* EXCLUDE y
FROM df
ORDER BY y
""",
eager=True,
)
assert order_by_exclude_res.to_dict(as_series=False) == {
"x": [3, 2, 1],
}

order_by_qualified_exclude_res = pl.SQLContext(df=nums).execute(
"""
SELECT
df.* EXCLUDE y
FROM df
ORDER BY y
""",
eager=True,
)
assert order_by_qualified_exclude_res.to_dict(as_series=False) == {
"x": [3, 2, 1],
}

order_by_expression_res = pl.SQLContext(df=nums).execute(
"""
SELECT
x % y as modded
FROM df
ORDER BY x % y
""",
eager=True,
)
assert order_by_expression_res.to_dict(as_series=False) == {
"modded": [1, 1, 2],
}


def test_order_by_misc() -> None:
res = pl.DataFrame(
{
"x": ["apple", "orange"],
"y": ["sheep", "alligator"],
"z": ["hello", "world"],
}
).sql(
"""
SELECT z, y, x
FROM self ORDER BY y DESC
"""
)
assert res.columns == ["z", "y", "x"]
assert res.to_dict(as_series=False) == {
"z": ["hello", "world"],
"y": ["sheep", "alligator"],
"x": ["apple", "orange"],
}


def test_register_context() -> None:
# use as context manager unregisters tables created within each scope
# on exit from that scope; arbitrary levels of nesting are supported.
Expand Down
2 changes: 1 addition & 1 deletion py-polars/tests/unit/sql/test_numeric.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ def test_div() -> None:
{
"a": [20.5, None, 10.0, 5.0, 2.5],
"b": [6, 12, 24, None, 5],
}
},
)
res = df.sql("SELECT DIV(a, b) AS a_div_b, DIV(b, a) AS b_div_a FROM self")
assert res.to_dict(as_series=False) == {
Expand Down
Loading