def test_load_pickle(self, tmp_path):
"""
Tests that the corpus can properly load from a pickle file, while retaining
all of the relevant information
:param tmp_path: a temporary directory created by pytest that will be used to store
a pickle file from the test
"""
pickle_path = tmp_path / 'pickle.pgz'
original_corpus = Corpus(common.TEST_CORPUS_PATH,
csv_path=common.SMALL_TEST_CORPUS_CSV,
name='test_corpus',
pickle_on_load=pickle_path,
ignore_warnings=True)
# first make sure the small corpus is correct
assert len(original_corpus) == 10
assert type(original_corpus.documents) == list
assert original_corpus.name == 'test_corpus'
# next load the pickle file to make sure data was copied correctly
pickle_corpus = Corpus(pickle_path, name='test_corpus')
assert len(pickle_corpus) == 10
assert type(original_corpus.documents) == list
assert pickle_corpus.name == 'test_corpus'
# Make sure the corpora are equal
assert original_corpus == pickle_corpus
def test_plot_gender_breakdown_different_file_constructions(self):
c = Corpus(
common.TEST_CORPUS_PATH,
csv_path=common.LARGE_TEST_CORPUS_CSV,
name='test_corpus',
)
default_save_name = 'gender_breakdown_for_' + c.name.replace(
' ', '_') + '.png'
test_file_1_name = "testing_file1.png"
default_save_path = OUTPUT_DIRECTORY_PATH / default_save_name
test_file_save_path = OUTPUT_DIRECTORY_PATH / test_file_1_name
test_file_paths = []
plot_gender_breakdown(c, OUTPUT_DIRECTORY_PATH)
assert Path.is_file(default_save_path)
test_file_paths.append(default_save_path)
plot_gender_breakdown(c, OUTPUT_DIRECTORY_PATH, "testing file1")
assert Path.is_file(test_file_save_path)
test_file_paths.append(test_file_save_path)
for file_created1 in test_file_paths:
for file_created2 in test_file_paths:
assert filecmp.cmp(file_created1, file_created2)
for file_created in test_file_paths:
Path.unlink(file_created)
def test_load_without_csv(self):
"""
Tests that the corpus properly loads when not provided metadata
"""
c = Corpus(common.TEST_CORPUS_PATH)
assert len(c) == 99
assert type(c.documents) == list
assert c.name is None
def test_load_with_csv(self):
"""
Test that the corpus properly loads when provided a metadata csv
"""
c = Corpus(
common.TEST_CORPUS_PATH,
csv_path=common.LARGE_TEST_CORPUS_CSV,
name='test_corpus',
)
assert len(c) == 99
assert type(c.documents) == list
assert c.name == 'test_corpus'
def test_create_all_visualizations_but_with_no_corpus_name(self):
c = Corpus(common.TEST_CORPUS_PATH,
csv_path=common.LARGE_TEST_CORPUS_CSV)
default_gender_breakdown = 'gender_breakdown_for_corpus.png'
default_metadata_pie = 'percentage_acquired_metadata_for_corpus.png'
default_country_pub = 'country_of_pub_for_corpus.png'
default_pub_date = 'date_of_pub_for_corpus.png'
create_corpus_summary_visualizations(c, OUTPUT_DIRECTORY_PATH)
assert Path.is_file(OUTPUT_DIRECTORY_PATH / default_gender_breakdown)
assert Path.is_file(OUTPUT_DIRECTORY_PATH / default_pub_date)
assert Path.is_file(OUTPUT_DIRECTORY_PATH / default_country_pub)
assert Path.is_file(OUTPUT_DIRECTORY_PATH / default_metadata_pie)