def test_identities(events, result_sets, db, es):
# NOTE: We assume that only on identifier scheme being used so just using
# identifier values is enough when comparing sets.
for ev in generate_payloads(events):
EventAPI.handle_event(ev)
for rs in result_sets:
for v in rs:
id_ = Identifier.query.filter_by(value=v).one()
ids = set(i.value for i in id_.get_identities())
assert ids == rs
def test_simple_citations(test_case_name, events, results, db, es):
"""Test simple citation queries."""
for ev in generate_payloads(events):
EventAPI.handle_event(ev)
es.indices.refresh()
for cited_id_value, (citation_result, relation_result) in results.items():
cited_id = (Identifier.query
.filter_by(value=cited_id_value).one())
citations = set()
relations = set()
for citing_id, relation in RelationshipAPI.get_citations(cited_id):
citations |= {i.value for i in citing_id}
relations |= {(r.source.value, r.target.value) for r in relation}
assert citations == citation_result
assert relations == relation_result
def generate_full_graph():
# N_nodes = (N_parents * (N_children + 1)) * (N_ids + 1)
# N_HasVersion = N_parents * N_children
# N_IsIdenticalTo = N_nodes
# N_Cites = ~ N_parents * N_citations
N_parents = 10000
N_children = 5
N_ids = 2
N_citations = 20
relations = []
parents = ['P_' + str(i) for i in range(N_parents)]
par_group = defaultdict(lambda: [])
child_lut = {}
for p in parents:
for id_idx in range(N_ids):
p_iden = p + "_ID_" + str(id_idx)
par_group[p].append(p_iden)
relations.append(['C', p, 'IsIdenticalTo', p_iden, '2018-01-01'])
for c_idx in range(N_children):
c = p + "_C_" + str(c_idx)
par_group[p].append(p)
relations.append(['C', p, 'HasVersion', c, '2018-01-01'])
for id_idx in range(N_ids):
c_iden = c + "_ID_" + str(id_idx)
par_group[p].append(c_iden)
relations.append(['C', c, 'IsIdenticalTo', c_iden, '2018-01-01'])
for idx, (par, grp) in enumerate(par_group.items()):
if (idx + 1) <= N_parents - 1:
n_samples = min(N_citations, N_parents - idx - 1)
next_parents = random.sample(range(idx + 1, N_parents), n_samples)
for next_par_idx in next_parents:
citeA = random.choice(grp)
next_par = parents[next_par_idx]
citeB = random.choice(par_group[next_par])
relations.append(['C', citeA, 'Cites', citeB, '2018-01-01'])
res = generate_payloads([relations])
print(json.dumps(res, indent=2))
def test_grouping_query(test_case_name, events, results, db, es):
for ev in generate_payloads(events):
EventAPI.handle_event(ev)
for cited_id_value, _ in results.items():
pass
def _handle_events(evtsrc):
events = generate_payloads(evtsrc)
for ev in events:
EventAPI.handle_event(ev)