def test_detects_many_to_one():
adset_id_attr = FdqModelAttribute(data_reference='"adset_id"', field_map=['adset_id'])
ad_id_attr = FdqModelAttribute(data_reference='"ad_id"', field_map=['ad_id'])
spend_attr = FdqModelAttribute(data_reference='"spend"', field_map=['spend'])
impressions_attr = FdqModelAttribute(data_reference='"impressions"', field_map=['impressions'])
adset_model = create_temp_internal_from_api_model(
name='model_1',
data_source='source.db.schema.table',
fields=[adset_id_attr, spend_attr],
identifiers=['adset_id'],
)
ad_model = create_temp_internal_from_api_model(
name='model_2',
data_source='source.db.schema.second_table',
fields=[ad_id_attr, adset_id_attr, spend_attr, impressions_attr],
identifiers=['adset_id', 'ad_id'],
)
models = [adset_model, ad_model]
result = detect_joins(models)
assert result == {
ad_model.model_name: [
FdqModelJoin(
to_model=adset_model.model_name,
relationship=FdqModelJoinRelationship.many_to_one,
fields=['adset_id'],
join_type=JoinType.left,
).dict(by_alias=True)
]
}
def test_does_not_find_joins_with_no_matching_ids():
some_id_attr = FdqModelAttribute(data_reference='"some_id"', field_map=['some_id'])
another_id_attr = FdqModelAttribute(data_reference='"another_id"', field_map=['another_id'])
spend_attr = FdqModelAttribute(data_reference='"spend"', field_map=['spend'])
model_one = create_temp_internal_from_api_model(
name='model_1', data_source='source.db.schema.table', fields=[some_id_attr, spend_attr], identifiers=['some_id']
)
model_two = create_temp_internal_from_api_model(
name='model_2',
data_source='source.db.schema.table',
fields=[another_id_attr, spend_attr],
identifiers=['another_id'],
)
assert detect_joins(models=[model_one, model_two]) == {}
def from_internal(
cls,
transformation: str,
model_attributes: List[ModelAttribute],
virtual_data_source: str,
) -> FdqModelAttribute:
"""
Initialize model attribute for API model from Husky model attribute
and cutoff virtual data source from taxon slug, if needed
"""
# type is same across all attributes
return FdqModelAttribute.construct(
field_map=[
remove_virtual_data_source_prefix(virtual_data_source,
attr.taxon)
for attr in model_attributes
],
# for backward compatibility, we reuse column_name for a bit
data_reference=transformation,
)
import pytest
from panoramic.cli.husky.core.federated.model.models import FdqModelAttribute
from panoramic.cli.husky.core.federated.model.tel.data_structures import (
AttributeValidationTelVisitorParams, )
from panoramic.cli.husky.core.federated.model.tel.visitor import (
AttributeValidationTelVisitor, )
from panoramic.cli.husky.core.tel.exceptions import TelExpressionException
from panoramic.cli.husky.core.tel.tel_dialect import ModelTelDialect
_TMP_MODEL_ATTRIBUTES = [
FdqModelAttribute(data_reference='"column_0"', field_map=['source_taxon']),
FdqModelAttribute(data_reference='"column_1"', field_map=['gender']),
FdqModelAttribute(data_reference='"column_2"', field_map=['ad_id']),
FdqModelAttribute(data_reference='"column_3"', field_map=['spend']),
FdqModelAttribute(data_reference='spend + 10', field_map=['spend_2']),
]
@pytest.mark.parametrize(
['inp', 'exp', 'exp_used_taxons', 'exp_used_columns'],
[
('1+10', '1+10', set(), set()),
('1+10/spend', '1+10/"column_3"', {'spend'}, {'column_3'}),
('1+10/spend_2', '1+10/"column_3"+10', {'spend_2', 'spend'
}, {'column_3'}),
(
'1+10/spend_2 * unknown_taxon',
'1+10/"column_3"+10*unknown_taxon',
{'spend_2', 'spend', 'unknown_taxon'},
{'column_3'},
_VIRTUAL_DATA_SOURCE)
assert attr.dict(by_alias=True) == {
'data_reference':
transformation,
'field_map': [
remove_virtual_data_source_prefix(_VIRTUAL_DATA_SOURCE,
model_attrs[0].taxon)
],
}
@pytest.mark.parametrize(
'model_attr',
[
FdqModelAttribute(data_reference='"first_col"',
field_map=['taxon_slug']),
FdqModelAttribute(data_reference='"col_name"',
field_map=['taxon_slug_2']),
],
)
def test_api_model_attribute_to_internal(model_attr):
identifiers = {'taxon_slug'}
attrs = FdqModelAttributeMapper.to_internal(model_attr,
_VIRTUAL_DATA_SOURCE,
identifiers)
assert [attr.to_primitive() for attr in attrs] == [{
'tel_transformation':
model_attr.data_reference,
'taxon':
prefix_with_virtual_data_source(_VIRTUAL_DATA_SOURCE,