def test_run_corpus_failure() -> None:
fail_corpus = [
('never produced',
'2015-12-12T12:30',
('today', 'heute'))
]
with pytest.raises(Exception):
run_corpus(fail_corpus)
def test_run_corpus() -> None:
"""The corpus passes if ctparse generates the desired
solution for each test at least once. Otherwise it fails.
"""
X, y = run_corpus(corpus)
assert isinstance(y[0], bool)
assert isinstance(X[0][0], str)
def main():
args = parse_args()
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s %(levelname)s [%(name)s] %(message)s")
X_combined = []
y_combined = []
if args.legacy:
logger.info("Loading legacy dataset")
X, y = run_corpus(corpus.corpus + auto_corpus.corpus)
X_combined.extend(X)
y_combined.extend(y)
if args.dataset:
logger.info("Loading dataset {}".format(args.dataset))
entries = load_timeparse_corpus(args.dataset)
X, y = zip(*make_partial_rule_dataset(
entries,
scorer=DummyScorer(),
timeout=30,
max_stack_depth=100,
progress=True,
))
X_combined.extend(X)
y_combined.extend(y)
if len(X) == 0:
raise ValueError("Need to specify at least a dataset for training")
mdl = train_naive_bayes(X_combined, y_combined)
save_naive_bayes(mdl, DEFAULT_MODEL_FILE)
def test_run_corpus_failure() -> None:
fail_corpus = [("never produced", "2015-12-12T12:30", ("today", "heute"))]
with pytest.raises(Exception):
run_corpus(fail_corpus)