def test_miner_to_dict():
miner = AdeftMiner('INDRA')
miner.process_texts(
[example_text1, example_text2, example_text3, example_text4])
miner_dict = miner.to_dict()
miner2 = load_adeft_miner_from_dict(miner_dict)
assert miner.top() == miner2.top()
assert miner.get_longforms(use_alignment_based_scoring=False) == \
miner2.get_longforms(use_alignment_based_scoring=False)
miner.compute_alignment_scores()
assert miner.get_longforms() == miner2.get_longforms()
def test_serialize_adeft_miner():
miner = AdeftMiner('INDRA')
miner.process_texts(
[example_text1, example_text2, example_text3, example_text4])
temp_filename = os.path.join(SCRATCH_PATH, uuid.uuid4().hex)
with open(temp_filename, 'w') as f:
miner.dump(f)
with open(temp_filename) as f:
miner2 = load_adeft_miner(f)
assert miner.top() == miner2.top()
assert miner.get_longforms() == miner2.get_longforms()
def test_get_longforms():
"""Test breadth first search algorithm to extract longforms
"""
miner = AdeftMiner('INDRA')
# ensure list of longforms is initialized correctly
assert miner.top() == []
miner.process_texts(
[example_text1, example_text2, example_text3, example_text4])
longforms = miner.get_longforms(cutoff=0.5)
assert (len(longforms) == 2)
assert longforms[0] == ('indonesian debt restructuring agency', 1.0)
assert longforms[1] == ('integrated network and dynamical'
' reasoning assembler', 1.0)