Visitar URL original
fix: Add project filter to apply_data_source and delete_data_source by mailtoboggavarapu-coder · Pull Request #6298 · feast-dev/feast · GitHub
Skip to content
Closed
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
19 changes: 14 additions & 5 deletions sdk/python/feast/infra/registry/registry.py
Comment thread
devin-ai-integration[bot] marked this conversation as resolved.
Original file line number Diff line number Diff line change
Expand Up @@ -386,6 +386,12 @@ def list_data_sources(
def apply_data_source(
self, data_source: DataSource, project: str, commit: bool = True
):
"""Apply a data source to the registry with project-scoped deduplication.

Filters existing data sources by both name and project (fixes feast-dev/feast#6206),
preserving the original created_timestamp if the source already exists in the
target project.
"""
now = _utc_now()
if not data_source.created_timestamp:
data_source.created_timestamp = now
Expand All @@ -394,7 +400,10 @@ def apply_data_source(
registry = self._prepare_registry_for_changes(project)

for idx, existing_data_source_proto in enumerate(registry.data_sources):
if existing_data_source_proto.name == data_source.name:
if (
existing_data_source_proto.name == data_source.name
and existing_data_source_proto.project == project
):
existing_data_source = DataSource.from_proto(existing_data_source_proto)
# Check if the data source has actually changed
if existing_data_source == data_source:
Expand Down Expand Up @@ -423,7 +432,7 @@ def delete_data_source(self, name: str, project: str, commit: bool = True):
for idx, data_source_proto in enumerate(
self.cached_registry_proto.data_sources
):
if data_source_proto.name == name:
if data_source_proto.name == name and data_source_proto.project == project:
del self.cached_registry_proto.data_sources[idx]
if commit:
self.commit()
Expand Down Expand Up @@ -688,13 +697,13 @@ def apply_feature_view(

if not is_latest:
# Explicit version: check if it exists (pin/revert) or not (forward declaration).
# Note: The file registry is last-write-wins for true concurrent races —
# Note: The file registry is last-write-wins for true concurrent races —
# this is a pre-existing limitation for all file registry operations.
# For multi-client environments, use the SQL registry.
record = self._get_version_record(feature_view.name, project, pin_version)

if record is not None:
# Version exists → pin/revert to that snapshot
# Version exists → pin/revert to that snapshot
# Check that the user hasn't also modified the definition.
# Compare user's FV (with version="latest") against active FV.
self._prepare_registry_for_changes(project)
Expand Down Expand Up @@ -735,7 +744,7 @@ def apply_feature_view(
# Apply the restored FV using the standard path below
feature_view = restored_fv
else:
# Version doesn't exist → forward declaration: create it
# Version doesn't exist → forward declaration: create it
feature_view.current_version_number = pin_version
feature_view_proto = feature_view.to_proto()
feature_view_proto.spec.project = project
Expand Down
105 changes: 105 additions & 0 deletions sdk/python/tests/integration/registration/test_universal_registry.py
Original file line number Diff line number Diff line change
Expand Up @@ -899,6 +899,111 @@ def test_apply_data_source_with_timestamps(test_registry):
test_registry.teardown()


@pytest.mark.integration
@pytest.mark.parametrize(
"test_registry",
all_fixtures,
)
def test_apply_data_source_cross_project_isolation(test_registry):
"""Test that apply_data_source uses project-scoped filtering.

Regression test for https://github.com/feast-dev/feast/issues/6206:
applying a data source to one project must not overwrite the data source
with the same name in a different project.

See: feast-dev/feast#6298
"""
project_a = "project_a"
project_b = "project_b"

source_a = FileSource(
name="shared_source_name",
file_format=ParquetFormat(),
path="file://feast/project_a.parquet",
timestamp_field="ts_col",
)
source_b = FileSource(
name="shared_source_name",
file_format=ParquetFormat(),
path="file://feast/project_b.parquet",
timestamp_field="ts_col",
)

test_registry.apply_data_source(source_a, project_a, commit=True)
test_registry.apply_data_source(source_b, project_b, commit=True)

# Each project should have exactly its own source
sources_a = test_registry.list_data_sources(project_a)
sources_b = test_registry.list_data_sources(project_b)
assert len(sources_a) == 1
assert len(sources_b) == 1

# Paths must be project-specific — not overwritten cross-project
assert sources_a[0].path == "file://feast/project_a.parquet"
assert sources_b[0].path == "file://feast/project_b.parquet"

# Re-apply source_b with updated path: must not bleed into project_a
source_b_updated = FileSource(
name="shared_source_name",
file_format=ParquetFormat(),
path="file://feast/project_b_v2.parquet",
timestamp_field="ts_col",
)
test_registry.apply_data_source(source_b_updated, project_b, commit=True)

sources_a_after = test_registry.list_data_sources(project_a)
assert len(sources_a_after) == 1
assert sources_a_after[0].path == "file://feast/project_a.parquet", (
"apply_data_source for project_b must not overwrite project_a's source"
)

test_registry.teardown()


@pytest.mark.integration
@pytest.mark.parametrize(
"test_registry",
all_fixtures,
)
def test_delete_data_source_project_scoped(test_registry):
"""Test that delete_data_source only removes the source from the given project.

Regression test for https://github.com/feast-dev/feast/issues/6206:
deleting a data source from one project must not delete the data source
with the same name from another project.
"""
project_a = "project_a"
project_b = "project_b"

source_a = FileSource(
name="shared_source_name",
file_format=ParquetFormat(),
path="file://feast/project_a.parquet",
timestamp_field="ts_col",
)
source_b = FileSource(
name="shared_source_name",
file_format=ParquetFormat(),
path="file://feast/project_b.parquet",
timestamp_field="ts_col",
)

test_registry.apply_data_source(source_a, project_a, commit=True)
test_registry.apply_data_source(source_b, project_b, commit=True)

# Delete the source from project_a only
test_registry.delete_data_source("shared_source_name", project_a, commit=True)

# project_a should have no sources; project_b should be unaffected
sources_a = test_registry.list_data_sources(project_a)
sources_b = test_registry.list_data_sources(project_b)
assert len(sources_a) == 0, "Source should be deleted from project_a"
assert len(sources_b) == 1, "Source in project_b must not be deleted"
assert sources_b[0].path == "file://feast/project_b.parquet"

test_registry.teardown()


@pytest.mark.integration
@pytest.mark.parametrize(
"test_registry",
Expand Down