def test_normalize_diacritics_file(self):
nfd_filepath = datapath('nfd.txt')
reference_nfc_filepath = datapath('nfc.txt')
processed_nfc_filepath = datapath('processed_nfc.txt')
self.assertTrue(preprocessing.normalize_diacritics_file(nfd_filepath, processed_nfc_filepath))
self.assertFalse(filecmp.cmp(nfd_filepath, processed_nfc_filepath)) # src & processed are different
self.assertTrue(filecmp.cmp(reference_nfc_filepath, processed_nfc_filepath)) # processed matches reference
def test_strip_accents_file(self):
src_filepath = datapath('src_file.txt')
reference_stripped_filepath = datapath('ref_proccessed_file.txt')
processed_stripped_filepath = datapath('processed_file.txt')
self.assertTrue(preprocessing.strip_accents_file(src_filepath, processed_stripped_filepath))
self.assertFalse(filecmp.cmp(src_filepath, processed_stripped_filepath))
self.assertTrue(filecmp.cmp(reference_stripped_filepath, processed_stripped_filepath))
def test_second_too_short():
with temporary_path("truncated.2.fastq") as trunc2:
# Create a truncated file in which the last read is missing
with open(datapath('paired.2.fastq')) as f:
lines = f.readlines()
lines = lines[:-4]
with open(trunc2, 'w') as f:
f.writelines(lines)
with raises(SystemExit), redirect_stderr():
execute_cli('-a XX --paired-output out.fastq'.split() +
[datapath('paired.1.fastq'), trunc2])
def test_load_corpus_from_directory(self):
direc = datapath('dirdata')
invalid_dir = datapath('test_data')
multi_dir = datapath()
path = Path(direc).glob('*')
dir_corpus = self.directory_loader(path=direc)
self.assertEqual(len(dir_corpus.data), len(list(path)))
with self.assertRaises(NotADirectoryError):
self.directory_loader(path=invalid_dir)
multi_corp = self.directory_loader(path=multi_dir)
multi_path = Path(multi_dir).glob('**/*')
self.assertEqual(len(multi_corp.data), len(list(multi_path)) - 1)
def test_explicit_format_with_paired():
# Use --format=fastq with input files whose extension is .txt
with temporary_path("paired.1.txt") as txt1:
with temporary_path("paired.2.txt") as txt2:
shutil.copyfile(datapath("paired.1.fastq"), txt1)
shutil.copyfile(datapath("paired.2.fastq"), txt2)
run_paired(
'--format=fastq -a TTAGACATAT -m 14',
in1=txt1,
in2=txt2,
expected1='paired.m14.1.fastq',
expected2='paired.m14.2.fastq',
)
def test_no_trimming_legacy():
# make sure that this doesn't divide by zero
execute_cli([
'-a',
'XXXXX',
'-o',
'/dev/null',
'-p',
'/dev/null',
'-pe1',
datapath('paired.1.fastq'),
'-pe2',
datapath('paired.2.fastq'),
])
def test_unmatched_read_names():
with temporary_path("swapped.1.fastq") as swapped:
try:
# Create a file in which reads 2 and are swapped
with open(datapath('paired.1.fastq')) as f:
lines = f.readlines()
lines = lines[0:4] + lines[8:12] + lines[4:8] + lines[12:]
with open(swapped, 'w') as f:
f.writelines(lines)
with redirect_stderr():
command = get_command('trim')
result = command.execute(
'-a XX -o out1.fastq -p out2.fastq'.split() +
['-pe1', swapped, '-pe2',
datapath('paired.2.fastq')])
assert isinstance(result, tuple)
assert len(result) == 2
assert result[0] != 0
finally:
os.remove('out1.fastq')
os.remove('out2.fastq')
def run_paired(
params,
in1,
in2,
expected1,
expected2,
aligners=('adapter', ),
callback=None,
assert_files_equal=True,
error_on_rc=True,
):
if type(params) is str:
params = params.split()
for aligner in aligners:
with temporary_path('tmp1-' + expected1.format(aligner=aligner)) as p1:
with temporary_path('tmp2-' +
expected2.format(aligner=aligner)) as p2:
p = params.copy()
p += ['--aligner', aligner, '-o', p1, '-p', p2]
infiles = [
datapath(i.format(aligner=aligner)) for i in (in1, in2)
]
for infile_args in zip(('-pe1', '-pe2'), infiles):
p.extend(infile_args)
command = get_command('trim')
result = command.execute(p)
assert isinstance(result, tuple)
assert len(result) == 2
if error_on_rc:
err = result[1]['exception'] if result[
1] and 'exception' in result[1] else None
# from traceback import format_exception
if result[0] != 0:
if err is None:
raise AssertionError("Return code {} != 0".format(
result[0]))
else:
raise AssertionError("Return code {} != 0".format(
result[0])) from err['details'][1]
if assert_files_equal:
assert files_equal(
cutpath(expected1.format(aligner=aligner)), p1)
assert files_equal(
cutpath(expected2.format(aligner=aligner)), p2)
if callback:
callback(aligner, infiles, (p1, p2), result)
def run_interleaved(params, inpath, expected, aligners=('adapter', )):
if type(params) is str:
params = params.split()
for aligner in aligners:
with temporary_path(expected.format(aligner=aligner)) as tmp:
p = params.copy()
p += [
'--aligner',
aligner,
'-l',
datapath(inpath.format(aligner=aligner)),
'-L',
tmp,
]
command = get_command('trim')
result = command.execute(p)
assert isinstance(result, tuple)
assert len(result) == 2
assert files_equal(cutpath(expected.format(aligner=aligner)), tmp)
def test_file_info(self):
reference_nfc_filepath = datapath('nfc.txt')
utils.file_info(reference_nfc_filepath)
def test_is_file_nfc(self):
src_filepath_pass = datapath('nfc.txt')
src_filepath_fail = datapath('nfc_fail.txt')
self.assertTrue(utils.is_file_nfc(src_filepath_pass))
self.assertFalse(utils.is_file_nfc(src_filepath_fail))
def test_load_corpus_from_text(self):
text = open(datapath('owe_pass')).read()
corpus = self.corpus_class(text=text)
self.assertEqual(len(corpus), 420)
def test_load_corpus_from_path_stream(self):
path = datapath('owe_pass')
corpus = self.corpus_class(path=path,
fformat=self.txt_extension,
stream=True)
self.assertEqual(len(corpus), 420)
def test_missing_file():
with raises(SystemExit), redirect_stderr():
execute_cli([
'-a', 'XX', '--paired-output', 'out.fastq',
datapath('paired.1.fastq')
])
def test_is_valid_owe_format(self):
fail_path = datapath('nfc.txt')
self.assertFalse(preprocessing.is_valid_owé_format(fail_path))
