def _filter_schemas(self, env: Environment, query: Select) -> Select:
if not self._filters.schema_keys:
return query
return query.filter(
DataBlockMetadata.nominal_schema_key.in_(
[d.key for d in self.get_schemas(env)]) # type: ignore
)
def _filter_unprocessed(
self,
env: Environment,
query: Select,
) -> Select:
if not self._filters.unprocessed_by_node_key:
return query
if self._filters.allow_cycle:
# Only exclude blocks processed as INPUT
filter_clause = and_(
DataBlockLog.direction == Direction.INPUT,
DataFunctionLog.node_key ==
self._filters.unprocessed_by_node_key,
)
else:
# No block cycles allowed
# Exclude blocks processed as INPUT and blocks outputted
filter_clause = (DataFunctionLog.node_key ==
self._filters.unprocessed_by_node_key)
already_processed_drs = (
Query(DataBlockLog.data_block_id).join(
DataFunctionLog).filter(filter_clause).filter(
DataBlockLog.invalidated == False) # noqa
.distinct())
return query.filter(
not_(DataBlockMetadata.id.in_(already_processed_drs)))
def _ignore_domains_filter(query: Select) -> Select:
"""Add a filter to ignore domains we do not fetch history for."""
return query.filter(
and_(*[
~States.entity_id.like(entity_domain)
for entity_domain in IGNORE_DOMAINS_ENTITY_ID_LIKE
]))
def _filter_inputs(
self,
env: Environment,
query: Select,
) -> Select:
if not self._filters.node_keys:
return query
eligible_input_drs = (
Query(DataBlockLog.data_block_id).join(SnapLog).filter(
DataBlockLog.direction == Direction.OUTPUT,
SnapLog.node_key.in_(self._filters.node_keys),
).filter(DataBlockLog.invalidated == False) # noqa
.distinct())
return query.filter(DataBlockMetadata.id.in_(eligible_input_drs))
def _filter_data_block(self, env: Environment, query: Select) -> Select:
if not self._filters.data_block_id:
return query
return query.filter(
DataBlockMetadata.id == self._filters.data_block_id)
