def test_archive_fileexistserror(archive_params, archive_dir):
filename, schema = archive_params
context = Context()
context.archive = str(archive_dir)
context.data = os.path.join(get_data_path(), filename)
context.overwrite = False
context.schema = schema
init(context)
expected_df = DataFrame(read_csv(context.data, dtype=str))
expected_paths = _get_expected_paths(context.archive, schema, expected_df,
filename)
for e in expected_paths:
os.makedirs(os.path.dirname(e), exist_ok=True)
with open(e, mode='w') as f:
f.write('content')
with pytest.raises(FileExistsError):
archive(context)
try:
os.remove(os.path.join(get_data_path(), '#lock'))
except OSError:
raise
def test_build_iris_no_schema(self, archive_dir, cache_file, overwrite):
try:
TestBuild._delete_cache(str(cache_file))
except FileNotFoundError:
pass
except OSError:
raise
context = Context()
context.archive = str(archive_dir)
context.cache = str(cache_file)
context.data = os.path.join(get_data_path(), 'iris.csv')
context.overwrite = overwrite
context.schema = SortedDict()
archive(context)
assert not os.path.exists(os.path.join(get_data_path(), '#lock'))
expected_frame = DataFrame(read_csv(context.data, dtype=str))
if context.overwrite:
with open(context.cache, mode='w') as f:
f.write('content')
build(context)
actual_frame = DataFrame(read_csv(context.cache, dtype=str))
assert_frame_equal(expected_frame, actual_frame, check_like=True)
def test_archive(archive_params, archive_dir, overwrite):
filename, schema = archive_params
context = Context()
context.archive = str(archive_dir)
context.data = os.path.join(get_data_path(), filename)
context.overwrite = overwrite
context.schema = schema
init(context)
expected_df = DataFrame(read_csv(context.data, dtype=str))
expected_df.sort_values(list(expected_df.columns), inplace=True)
expected_df.reset_index(drop=True, inplace=True)
expected_paths = _get_expected_paths(
context.archive,
schema,
expected_df,
filename
)
if context.overwrite:
for e in expected_paths:
os.makedirs(os.path.dirname(e), exist_ok=True)
with open(e, mode='w') as f:
f.write('content')
archive(context)
assert not os.path.exists(os.path.join(get_data_path(), '#lock'))
actual_frame = DataFrame()
actual_paths = SortedList()
for root, _, files in os.walk(context.archive):
for f in files:
if '.csv' in f:
filepath = os.path.join(root, f)
actual_paths.add(filepath)
actual_frame = concat([
actual_frame,
DataFrame(read_csv(filepath, dtype=str))
])
actual_frame.sort_values(list(actual_frame.columns), inplace=True)
actual_frame.reset_index(drop=True, inplace=True)
assert expected_paths == actual_paths
assert_frame_equal(expected_df, actual_frame)
def test_archive_no_schema(archive_params, archive_dir, overwrite):
filename, _ = archive_params
context = Context()
context.archive = str(archive_dir)
context.data = os.path.join(get_data_path(), filename)
context.overwrite = overwrite
context.schema = SortedDict()
expected_df = DataFrame(read_csv(context.data, dtype=str))
expected_df.sort_values(list(expected_df.columns), inplace=True)
expected_df.reset_index(drop=True, inplace=True)
expected_paths = SortedList([os.path.join(context.archive, filename)])
if context.overwrite:
for e in expected_paths:
path = archive_dir.new()
path.mkdir(os.path.basename(os.path.dirname(e)))
with open(e, mode='w') as f:
f.write('content')
archive(context)
actual_frame = DataFrame()
actual_paths = SortedList()
for root, _, files in os.walk(context.archive):
for f in files:
if '.csv' in f:
filepath = os.path.join(root, f)
actual_paths.add(filepath)
actual_frame = concat(
[actual_frame,
DataFrame(read_csv(filepath, dtype=str))])
actual_frame.sort_values(list(actual_frame.columns), inplace=True)
actual_frame.reset_index(drop=True, inplace=True)
assert expected_paths == actual_paths
assert_frame_equal(expected_df, actual_frame)
def test_build_fileexistserror(self, archive_dir, cache_file):
try:
TestBuild._delete_cache(str(cache_file))
except FileNotFoundError:
pass
except OSError:
raise
context = Context()
context.archive = str(archive_dir)
context.cache = str(cache_file)
context.data = os.path.join(get_data_path(), 'iris.csv')
context.overwrite = False
context.schema = SortedDict()
archive(context)
assert not os.path.exists(os.path.join(get_data_path(), '#lock'))
with open(context.cache, mode='w') as f:
f.write('content')
with pytest.raises(FileExistsError):
build(context)
def test_build_iris(self, archive_dir, cache_file, overwrite):
try:
TestBuild._delete_cache(str(cache_file))
except FileNotFoundError:
pass
except OSError:
raise
context = Context()
context.archive = str(archive_dir)
context.cache = str(cache_file)
context.data = os.path.join(get_data_path(), 'iris.csv')
context.overwrite = overwrite
context.schema = SortedDict({'0': 'Name'})
context.verbose = True
init(context)
archive(context)
assert not os.path.exists(os.path.join(get_data_path(), '#lock'))
expected_frame = DataFrame(read_csv(context.data, index_col='Index'))
expected_frame.sort_index(inplace=True)
if context.overwrite:
with open(context.cache, mode='w') as f:
f.write('content')
build(context)
actual_frame = DataFrame(read_csv(context.cache, index_col='Index'))
actual_frame.sort_index(inplace=True)
equal_shapes = expected_frame.shape == actual_frame.shape
equal_indices = (expected_frame.index).equals(actual_frame.index)
equal_columns = (expected_frame.columns).equals(actual_frame.columns)
if not equal_shapes:
print('unequal shapes: expected {0} found {1}'.format(
expected_frame.shape, actual_frame.shape))
if not equal_indices:
print('unequal indices: expected {0} found {1}'.format(
expected_frame.index, actual_frame.index))
if not equal_columns:
print('unequal columns: expected {0} found {1}'.format(
expected_frame.columns, actual_frame.columns))
for i, col in enumerate(expected_frame.columns):
if col in actual_frame:
expected_col = expected_frame.iloc[:, i]
actual_col = actual_frame.iloc[:, i]
equal_names = expected_col.name == actual_col.name
equal_lengths = len(expected_col) == len(actual_col)
equal_indices = (expected_col.index).equals(actual_col.index)
if not equal_names:
print(' col {0}: unequal names: expected {1} found {2}'.
format(i, expected_col.name, actual_col.name))
if not equal_lengths:
print(' col {0}: unequal lengths: expected {1} found {2}'.
format(i, len(expected_col), len(actual_col)))
if not equal_indices:
print(' col {0}: unequal indices: expected {1} found {3}'.
format(i, expected_col.index, actual_col.index))
if equal_lengths:
for j in range(expected_col.size):
if expected_col[j] != actual_col[j]:
msg = ' col {0}: unequal @ row {1}: '
msg += 'expected {2} found {3}'
print(
msg.format(i, j, expected_col[j],
actual_col[j]))
assert_frame_equal(expected_frame, actual_frame, check_exact=True)
