def test_orthographic_density_none():
drop_caches()
# Lemmas are all lowercase.
for word in SubstitutionFeaturesMixin._orthographic_density():
assert word.islower()
# And it's computed right.
drop_caches()
with settings.file_override("CLEARPOND"):
with open(settings.CLEARPOND, "w") as f:
f.write("dog" + 5 * "\t" + "2" + 24 * "\t" + "3\n" "cat" + 5 * "\t" + "2" + 24 * "\t" + "3")
assert set(SubstitutionFeaturesMixin._orthographic_density()) == {"dog", "cat"}
def test_orthographic_density():
drop_caches()
assert SubstitutionFeaturesMixin._orthographic_density("time") == 13
assert np.isnan(SubstitutionFeaturesMixin._orthographic_density("wickiup"))
