async def test_prefix_not_specified():
"""
Test that if we get a category with no preferred_prefixes
we make no changes.
"""
portal = KnowledgePortal()
preferred_prefixes = {}
query_graph = {
"nodes": {
"n0": {"ids": ["DOID:9352"]},
},
"edges": {},
}
msg = Message.parse_obj({"query_graph": query_graph})
await portal.map_prefixes(
msg,
preferred_prefixes,
)
# n0 should be unchanged
assert msg.dict()["query_graph"]["nodes"]["n0"]["ids"] == ["DOID:9352"]
async def test_unknown_prefix():
"""
Test that if passed an unknown prefix we
assume it doesn't need to be changed.
"""
portal = KnowledgePortal()
preferred_prefixes = {"biolink:Disease": ["MONDO"]}
query_graph = {
"nodes": {
"n0": {"ids": ["UNKNOWN:000000"]},
},
"edges": {},
}
msg = Message.parse_obj({"query_graph": query_graph})
await portal.map_prefixes(
msg,
preferred_prefixes,
)
# n0 should be unchanged
assert msg.dict()["query_graph"]["nodes"]["n0"]["ids"] == ["UNKNOWN:000000"]
async def test_map_prefixes_small_example():
"""
Test that prefixes are mapped properly and that already
mapped prefixes are unchanged.
"""
portal = KnowledgePortal()
preferred_prefixes = {"biolink:Disease": ["MONDO"]}
query_graph = {
"nodes": {
"n0": {"ids": ["DOID:9352"]},
"n1": {"ids": ["MONDO:0005148"]},
},
"edges": {},
}
msg = Message.parse_obj({"query_graph": query_graph})
await portal.map_prefixes(
msg,
preferred_prefixes,
)
msg = msg.dict()
# n0 should be converted to the correct prefix
assert msg["query_graph"]["nodes"]["n0"]["ids"] == ["MONDO:0005148"]
# There should be no change to n1
assert msg["query_graph"]["nodes"]["n1"]["ids"] == ["MONDO:0005148"]
async def test_normalizer_not_reachable(caplog):
"""
Test that if the normalizer is completely unavailable we make
no changes to the query graph and continue on while adding a
a warning to the log
"""
portal = KnowledgePortal()
preferred_prefixes = {"biolink:Disease": ["MONDO"]}
query_graph = {
"nodes": {
"n0": {"ids": ["DOID:9352"]},
},
"edges": {},
}
msg = Message.parse_obj({"query_graph": query_graph})
await portal.map_prefixes(
msg,
preferred_prefixes,
)
# n0 should be unchanged
assert msg.dict()["query_graph"]["nodes"]["n0"]["ids"] == ["DOID:9352"]
assert "RequestError contacting normalizer" in caplog.text
async def test_normalizer_no_synonyms_available(caplog):
"""
Test that if we send a node with no synonyms
to the normalizer that we continue working and add
a warning to the log
"""
portal = KnowledgePortal()
preferred_prefixes = {"biolink:Disease": ["MONDO"]}
query_graph = {
"nodes": {
"n0": {"ids": ["DOID:9352"]},
},
"edges": {},
}
msg = Message.parse_obj({"query_graph": query_graph})
await portal.map_prefixes(
msg,
preferred_prefixes,
)
# n0 should be unchanged
assert msg.dict()["query_graph"]["nodes"]["n0"]["ids"] == ["DOID:9352"]
# The error we recieved from the normalizer should be in the logs
assert normalizer_error_no_matches in caplog.text
def generate_message(spec) -> Message:
"""
Generate a message using a specification. Example for the specification format:
{
"random_seed" : 4,
"knowledge_graph" : {
"nodes" : {
"count" : 100,
"attributes" : {
"count" : 100,
"spec": {
"subattribute_count": 5,
"value_type": "list",
"value_count": 10
}
},
"categories_count" : 1
},
"edges" : {
"count" : 100,
"attributes" : {
"count" : 100,
"spec": {
"subattribute_count": 5,
"value_type": "dict",
"value_count": 10
}
}
}
},
"results" : {
"count" : 100,
"node_bindings" : {
"count_per_node" : 1
},
"edge_bindings" : {
"count_per_edge" : 1,
"attributes" : {
"count" : 100,
"spec": {
"subattribute_count": 5,
"value_type": "string"
}
}
}
}
}
"""
randomInstance = random.Random(spec.get("random_seed", None))
get_random = lambda: "".join(
randomInstance.choice(string.ascii_letters) for _ in range(10)
)
kg_node_ids = [
f"biolink:{get_random()}"
for _ in range(spec["knowledge_graph"]["nodes"]["count"])
]
kg_edge_ids = [
get_random() for _ in range(spec["knowledge_graph"]["nodes"]["count"])
]
return Message(
query_graph=QueryGraph(nodes={}, edges={}),
knowledge_graph=KnowledgeGraph(
nodes={
kgnid: Node(
attributes=[
generate_attribute(
spec["knowledge_graph"]["nodes"]["attributes"]["spec"],
get_random,
)
for _ in range(
spec["knowledge_graph"]["nodes"]["attributes"]["count"]
)
],
categories=[
f"biolink:Category{get_random()}"
for _ in range(
spec["knowledge_graph"]["nodes"]["categories_count"]
)
],
)
for kgnid in kg_node_ids
},
edges={
kgeid: Edge(
attributes=[
generate_attribute(
spec["knowledge_graph"]["edges"]["attributes"]["spec"],
get_random,
)
for _ in range(
spec["knowledge_graph"]["edges"]["attributes"]["count"]
)
],
subject=random.choice(kg_node_ids),
predicate=f"biolink:{get_random().lower()}",
object=random.choice(kg_node_ids),
)
for kgeid in kg_edge_ids
},
),
results=[
Result(
node_bindings={
f"QGraphNode:{get_random()}": [
NodeBinding(id=kgnid)
for _ in range(
spec["results"]["node_bindings"]["count_per_node"]
)
]
for kgnid in kg_node_ids
},
edge_bindings={
f"QGraphEdge:{get_random()}": [
EdgeBinding(
id=kgeid,
attributes=[
generate_attribute(
spec["results"]["edge_bindings"]["attributes"][
"spec"
],
get_random,
)
for _ in range(
spec["results"]["edge_bindings"]["attributes"][
"count"
],
)
],
)
for _ in range(
spec["results"]["edge_bindings"]["count_per_edge"]
)
]
for kgeid in kg_edge_ids
},
)
for _ in range(spec["results"]["count"])
],
)