def node_visible(key: str) -> bool:
k: Kind = graph.nodes[key]["data"]
if hide and exist(lambda r: r.fullmatch(k.fqn), hide):
return False
if show is None:
return True
else:
return exist(lambda r: r.fullmatch(k.fqn), show)
def current_step_done(self) -> bool:
"""
This step is done, when the event it is waiting for has arrived.
"""
return self.timed_out or exist(
lambda x: isinstance(x, Event) and x.message_type == self.perform.
wait_for_message_type,
self.instance.received_messages,
)
async def query_cost(graph_db: Any, model: QueryModel,
with_edges: bool) -> EstimatedSearchCost:
q_string, bind = to_query(graph_db, model, with_edges=with_edges)
nr_nodes = await graph_db.db.count(graph_db.vertex_name)
plan = await graph_db.db.explain(query=q_string, bind_vars=bind)
full_collection_scan = exist(
lambda node: node["type"] == "EnumerateCollectionNode", plan["nodes"])
estimated_cost = int(plan["estimatedCost"])
estimated_items = int(plan["estimatedNrItems"])
# If the number of returned items is small, most of the computation happens on the db side
# A higher factor (==estimated cost) is acceptable in this case.
factor = 20 if estimated_items < 3 else 2.1
# max upper bound, if the number of nodes is very small
ratio = estimated_cost / max(10000, nr_nodes * factor)
# the best rating is complex, if a full collection scan is required.
best = Rating.complex if full_collection_scan else Rating.simple
rating = best if ratio < 0.2 else (
Rating.complex if ratio < 1 else Rating.bad)
return EstimatedSearchCost(estimated_cost, estimated_items, nr_nodes,
full_collection_scan, rating)
def not_hidden(key: str) -> bool:
k: Kind = graph.nodes[key]["data"]
return not (hide and exist(lambda r: r.fullmatch(k.fqn), hide))