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.

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

Fix circular relationships validation #2204

Merged
merged 2 commits into from
Sep 9, 2024
Merged
Show file tree
Hide file tree
Changes from 1 commit
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
2 changes: 1 addition & 1 deletion sdv/metadata/multi_table.py
Original file line number Diff line number Diff line change
Expand Up @@ -138,7 +138,7 @@ def _validate_circular_relationships(

for child in children:
if child in parents:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

minor: can we rename parents to visited? I was confused reading this thinking that we were checking the parents of only the current node, when this list actually keeps growing.

break
continue

parents.add(child)
self._validate_circular_relationships(
Expand Down
96 changes: 96 additions & 0 deletions tests/unit/metadata/test_multi_table.py
Original file line number Diff line number Diff line change
Expand Up @@ -1418,6 +1418,102 @@ def test_validate_data_datetime_warning(self):
with pytest.warns(UserWarning, match=warning_msg):
metadata.validate_data(data)

def test_add_relationship_circular_graph(self):
"""Test that an error is raised when a circular relationship is detected.

The graph has the cycle B->C->D->B.
Besides the cycle, the other relationships are: B->A, C->A, D->A.
"""
# Setup
metadata = MultiTableMetadata()
metadata.add_table('A')
metadata.add_column('A', 'id', sdtype='id')
metadata.add_column('A', 'fk', sdtype='id')
metadata.set_primary_key('A', 'id')

metadata.add_table('B')
metadata.add_column('B', 'id', sdtype='id')
metadata.add_column('B', 'fk', sdtype='id')
metadata.set_primary_key('B', 'id')

metadata.add_table('C')
metadata.add_column('C', 'id', sdtype='id')
metadata.add_column('C', 'fk', sdtype='id')
metadata.set_primary_key('C', 'id')

metadata.add_table('D')
metadata.add_column('D', 'id', sdtype='id')
metadata.add_column('D', 'fk', sdtype='id')
metadata.set_primary_key('D', 'id')

metadata.add_relationship('B', 'C', 'id', 'fk')
metadata.add_relationship('B', 'A', 'id', 'fk')

metadata.add_relationship('C', 'D', 'id', 'fk')
metadata.add_relationship('C', 'A', 'id', 'fk')

metadata.add_relationship('D', 'A', 'id', 'fk')

# Run and Assert
error_msg = re.escape(
'The relationships in the dataset describe a '
"circular dependency between tables ['B', 'C', 'D']."
)
with pytest.raises(InvalidMetadataError, match=error_msg):
metadata.add_relationship('D', 'B', 'id', 'fk')

def test_add_relationship_circular_graph_complex(self):
"""Test that an error is raised when a circular relationship is detected.

The graph has the cycle C->E->D->C.
Besides the cycle, the other relationships are: C->B, D->B, E->B, E->A, A->B.
"""
# Setup
metadata = MultiTableMetadata()
metadata.add_table('A')
metadata.add_column('A', 'id', sdtype='id')
metadata.add_column('A', 'fk', sdtype='id')
metadata.set_primary_key('A', 'id')

metadata.add_table('B')
metadata.add_column('B', 'id', sdtype='id')
metadata.add_column('B', 'fk', sdtype='id')
metadata.set_primary_key('B', 'id')

metadata.add_table('C')
metadata.add_column('C', 'id', sdtype='id')
metadata.add_column('C', 'fk', sdtype='id')
metadata.set_primary_key('C', 'id')

metadata.add_table('D')
metadata.add_column('D', 'id', sdtype='id')
metadata.add_column('D', 'fk', sdtype='id')
metadata.set_primary_key('D', 'id')

metadata.add_table('E')
metadata.add_column('E', 'id', sdtype='id')
metadata.add_column('E', 'fk', sdtype='id')
metadata.set_primary_key('E', 'id')

metadata.add_relationship('C', 'B', 'id', 'fk')
metadata.add_relationship('C', 'E', 'id', 'fk')

metadata.add_relationship('D', 'B', 'id', 'fk')
metadata.add_relationship('D', 'C', 'id', 'fk')

metadata.add_relationship('A', 'B', 'id', 'fk')

metadata.add_relationship('E', 'A', 'id', 'fk')
metadata.add_relationship('E', 'B', 'id', 'fk')

# Run and Assert
error_msg = re.escape(
'The relationships in the dataset describe a '
"circular dependency between tables ['C', 'D', 'E']."
)
with pytest.raises(InvalidMetadataError, match=error_msg):
metadata.add_relationship('E', 'D', 'id', 'fk')

@patch('sdv.metadata.multi_table.SingleTableMetadata')
def test_add_table(self, table_metadata_mock):
"""Test that the method adds the table name to ``instance.tables``."""
Expand Down
Loading