def test_name():
filename = "name.txt"
name_file = 'tests/test_files/' + filename
output_dir = 'tests/test_files/redacted/'
main.init_stats(name_file, 0, None)
# Get test file
content = main.get_file_contents(name_file)
# Used to split the file for POS analysis
word_punct_tokenizer = WordPunctTokenizer()
tagged_content = nltk.pos_tag(word_punct_tokenizer.tokenize(content))
# Redacte
content = main.redact_names(content, tagged_content, name_file)
# X nameed words in file
assert (main.num_names[name_file] == 22)
# Create path
if (not os.path.isdir(output_dir)):
sys.stderr.write("Output directory did not exist...creating " +
output_dir + "/\n")
os.makedirs(output_dir)
# Write out the redacted test file for reference
main.write_redacted(content, name_file, output_dir)
def test_concept():
filename = "concept.txt"
concept_file = 'tests/test_files/' + filename
output_dir = 'tests/test_files/redacted/'
main.init_stats(concept_file, 0, None)
# Get test file
content = main.get_file_contents(concept_file)
# Used to split the file for POS analysis
word_punct_tokenizer = WordPunctTokenizer()
tagged_content = nltk.pos_tag(word_punct_tokenizer.tokenize(content))
# Make required dot structure.
# See https://stackoverflow.com/questions/2352181/how-to-use-a-dot-to-access-members-of-dictionary
arg = {"concept": ["child"]}
args = temp(arg)
# Redacte
content = main.redact_concept(content, concept_file, args)
# X concept words in file
assert (main.num_concept[concept_file] == 12)
# Create path
if (not os.path.isdir(output_dir)):
sys.stderr.write("Output directory did not exist...creating " +
output_dir + "/\n")
os.makedirs(output_dir)
# Write out the redacted test file for reference
main.write_redacted(content, concept_file, output_dir)
def test_address():
filename = "addresses.txt"
address_file = 'tests/test_files/' + filename
output_dir = 'tests/test_files/redacted/'
main.init_stats(address_file, 0, None)
# Get test file
content = main.get_file_contents(address_file)
# Redacte
content = main.redact_addresses(content, address_file)
# Three addressses in file
assert (main.num_addresses[address_file] == 3)
# Create path
if (not os.path.isdir(output_dir)):
sys.stderr.write("Output directory did not exist...creating " +
output_dir + "/\n")
os.makedirs(output_dir)
# Write out the redacted test file for reference
main.write_redacted(content, address_file, output_dir)