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
22 changes: 14 additions & 8 deletions src/dve/core_engine/backends/implementations/duckdb/rules.py
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,7 @@
ConstantExpression,
DuckDBPyConnection,
DuckDBPyRelation,
FunctionExpression,
StarExpression,
)
from duckdb.typing import DuckDBPyType
Expand All @@ -22,7 +23,6 @@
)
from dve.core_engine.backends.exceptions import ConstraintError
from dve.core_engine.backends.implementations.duckdb.duckdb_helpers import (
DDBStruct,
ddb_filter_contract_errors,
duckdb_get_entity_count,
duckdb_read_parquet,
Expand Down Expand Up @@ -361,8 +361,7 @@ def join_header(self, entities: DuckDBEntities, *, config: HeaderJoin) -> Messag
target_rel: DuckDBPyRelation = entities[config.target_name]
target_rel = target_rel.set_alias(config.target_name)

target_rows = target_rel.pl().to_struct("header").to_list()
n_target_rows = len(target_rows)
n_target_rows = target_rel.shape[0]
if n_target_rows != 1:
raise ConstraintError(
f"Unable to join header {config.target_name!r} to {config.entity_name!r} "
Expand All @@ -373,11 +372,18 @@ def join_header(self, entities: DuckDBEntities, *, config: HeaderJoin) -> Messag
),
)

target_schema = DDBStruct(dict(zip(target_rel.columns, target_rel.dtypes)))() # type: ignore # pylint:disable=C0301

joined_rel = source_rel.select(
StarExpression(exclude=[]),
ConstantExpression(target_rows[0]).cast(target_schema).alias(config.header_column_name),
joined_rel = (
source_rel
.cross(target_rel)
.select(
StarExpression(
exclude=[f"{config.target_name}.{c}" for c in target_rel.columns]
),
FunctionExpression(
"struct_pack",
*[ColumnExpression(f"{config.target_name}.{c}") for c in target_rel.columns],
).alias(config.header_column_name)
)
)

entities[config.new_entity_name or config.entity_name] = joined_rel
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -70,6 +70,13 @@ def value_literal_1_header(duckdb_connection) -> Iterator[DuckDBPyRelation]:
yield duckdb_connection.sql("SELECT 1 AS Value")


@pytest.fixture(scope="function")
def value_mixed_header(duckdb_connection) -> Iterator[DuckDBPyRelation]:
yield duckdb_connection.sql(
"SELECT make_date(0000,1,1) as test_date, 'test' as test_str, 123 as test_int, make_date(10000,1,1) as test_date2"
)


def test_column_addition(planets_rel: DuckDBPyRelation):
"""Test that columns can be added to entities."""
entities = EntityManager({"planets": planets_rel})
Expand Down Expand Up @@ -566,6 +573,31 @@ def test_header_join_planets(
assert planet_row == actual_row, "More than header changed as result of header join"


def test_header_join_with_out_of_range_date(
planets_rel: DuckDBPyRelation,
value_mixed_header: DuckDBPyRelation
):
"""Test that a header with an invalid date does not error"""
header_join = HeaderJoin(
entity_name="planets",
target_name="header",
new_entity_name="planets_header",
header_column_name="_Header"
)
entities = EntityManager({
"planets": planets_rel,
"header": value_mixed_header,
})
DUCKDB_STEP_BACKEND.evaluate(entities, config=header_join)

assert entities.entities["planets_header"].select("_header").distinct().shape[0] == 1
assert entities.entities["planets_header"].shape[0] == 9
assert (
entities.entities["planets_header"].select("typeof(_header)").fetchone()[0]
== 'STRUCT(test_date DATE, test_str VARCHAR, test_int INTEGER, test_date2 DATE)'
)


def test_header_multi_rows_raises(
planets_rel: DuckDBPyRelation, value_literal_1_header: DuckDBPyRelation
):
Expand Down
Loading