def test_temp_ssh_files():
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn)
scsv = into(Temp(SSH(CSV)), csv, hostname='localhost')
assert discover(csv) == discover(scsv)
assert isinstance(scsv, _Temp)
def test_pandas_read_supports_read_csv_kwargs():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds, usecols=['name'])
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice',), ('Bob',)]
def test_pandas_discover_on_gzipped_files():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02',
open=gzip.open,
extension='.csv.gz') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: string, when: datetime}')
assert discover(csv) == ds
def test_pandas_read_supports_read_csv_kwargs():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds, usecols=['name'])
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', ), ('Bob', )]
def test_pandas_loads_in_datetimes_naively():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: string, when: datetime}')
assert discover(csv) == ds
df = convert(pd.DataFrame, csv)
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read_supports_gzip():
with filetext('Alice,1\nBob,2', open=gzip.open, extension='.csv.gz') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_pandas_loads_in_datetimes_naively():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: ?string, when: ?datetime}')
assert discover(csv) == ds
df = convert(pd.DataFrame, csv)
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read_supports_datetimes():
with filetext('Alice,2014-01-02\nBob,2014-01-03') as fn:
ds = datashape.dshape('var * {name: string, when: date}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert list(df.columns) == ['name', 'when']
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read_supports_missing_integers():
with filetext('Alice,1\nBob,') as fn:
ds = datashape.dshape('var * {name: string, val: ?int32}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert list(df.columns) == ['name', 'val']
assert df.dtypes['val'] == 'f4'
def test_unused_datetime_columns():
ds = datashape.dshape('var * {val: string, when: datetime}')
with filetext("val,when\na,2000-01-01\nb,2000-02-02") as fn:
csv = CSV(fn, has_header=True)
assert convert(
list,
csv_to_DataFrame(csv, usecols=['val'], squeeze=True,
dshape=ds)) == ['a', 'b']
def test_pandas_read():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_header_disagrees_with_dshape():
ds = datashape.dshape('var * {name: string, bal: int64}')
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn, header=True)
assert convert(list, csv) == [('Alice', 100), ('Bob', 200)]
assert list(convert(pd.DataFrame, csv).columns) == ['name', 'val']
assert list(convert(pd.DataFrame, csv, dshape=ds).columns) == ['name', 'bal']
def test_pandas_read_supports_datetimes():
with filetext('Alice,2014-01-02\nBob,2014-01-03') as fn:
ds = datashape.dshape('var * {name: string, when: date}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert list(df.columns) == ['name', 'when']
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read_supports_missing_integers():
with filetext('Alice,1\nBob,') as fn:
ds = datashape.dshape('var * {name: string, val: ?int32}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert list(df.columns) == ['name', 'val']
assert df.dtypes['val'] == 'f4'
def test_into_double_string():
with filetext('alice,1\nbob,2', extension='.csv') as source:
assert into(list, source) == [('alice', 1), ('bob', 2)]
with tmpfile('.csv') as target:
csv = into(target, source)
assert isinstance(csv, CSV)
with open(target) as f:
assert 'alice' in f.read()
def test_into_double_string():
with filetext('alice,1\nbob,2', extension='.csv') as source:
assert into(list, source) == [('alice', 1), ('bob', 2)]
with tmpfile('.csv') as target:
csv = into(target, source)
assert isinstance(csv, CSV)
with open(target) as f:
assert 'alice' in f.read()
def test_header_disagrees_with_dshape():
ds = datashape.dshape('var * {name: string, bal: int64}')
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn, header=True)
assert convert(list, csv) == [('Alice', 100), ('Bob', 200)]
assert list(convert(pd.DataFrame, csv).columns) == ['name', 'val']
assert list(convert(pd.DataFrame, csv,
dshape=ds).columns) == ['name', 'bal']
def test_pandas_read_supports_gzip():
with filetext('Alice,1\nBob,2',
open=gzip.open,
mode='wt',
extension='.csv.gz') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_DataFrame(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_copy_remote_csv():
with tmpfile('csv') as target:
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
csv = resource(fn)
scsv = into('ssh://localhost:foo.csv', csv)
assert isinstance(scsv, SSH(CSV))
assert discover(scsv) == discover(csv)
# Round trip
csv2 = into(target, scsv)
assert into(list, csv) == into(list, csv2)
def test_copy_remote_csv():
with tmpfile('csv') as target:
with filetext('name,balance\nAlice,100\nBob,200',
extension='csv') as fn:
csv = resource(fn)
scsv = into('ssh://localhost:foo.csv', csv)
assert isinstance(scsv, SSH(CSV))
assert discover(scsv) == discover(csv)
# Round trip
csv2 = into(target, scsv)
assert into(list, csv) == into(list, csv2)
def test_convert_through_temporary_local_storage():
with filetext('name,quantity\nAlice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn)
df = into(pd.DataFrame, csv)
scsv = into(Temp(SSH(CSV)), csv, hostname='localhost')
assert into(list, csv) == into(list, scsv)
scsv2 = into(Temp(SSH(CSV)), df, hostname='localhost')
assert into(list, scsv2) == into(list, df)
sjson = into(Temp(SSH(JSONLines)), df, hostname='localhost')
assert (into(np.ndarray, sjson) == into(np.ndarray, df)).all()
def test_drop():
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
with tmpfile('csv') as target:
scsv = SSH(CSV)(target, hostname='localhost')
assert not os.path.exists(target)
conn = sftp(**scsv.auth)
conn.put(fn, target)
assert os.path.exists(target)
drop(scsv)
assert not os.path.exists(target)
def test_drop():
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
with tmpfile('csv') as target:
csv = CSV(fn)
scsv = SSH(CSV)(target, hostname='localhost')
assert not os.path.exists(target)
with sftp(**scsv.auth) as conn:
conn.put(fn, target)
assert os.path.exists(target)
drop(scsv)
assert not os.path.exists(target)
def test_discover():
with filetext('name,balance\nAlice,100\nBob,200') as fn:
local = CSV(fn)
remote = SSH(CSV)(fn, hostname='localhost')
assert discover(local) == discover(remote)
def test_discover_csv_without_columns():
with filetext('Alice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn)
ds = discover(csv)
assert '100' not in str(ds)
def test_unicode_column_names():
with filetext('foo\xc4\x87,a\n1,2\n3,4', extension='csv') as fn:
csv = CSV(fn, has_header=True)
df = into(pd.DataFrame, csv)
def test_discover_csv_with_spaces_in_header():
with filetext(' name, val\nAlice,100\nBob,200', extension='csv') as fn:
ds = discover(CSV(fn, has_header=True))
assert ds.measure.names == ['name', 'val']
def test_unused_datetime_columns():
ds = datashape.dshape('var * {val: string, when: datetime}')
with filetext("val,when\na,2000-01-01\nb,2000-02-02") as fn:
csv = CSV(fn, has_header=True)
assert convert(list, csv_to_DataFrame(csv, usecols=['val'],
squeeze=True, dshape=ds)) == ['a', 'b']
def test_csv_missing_values():
with filetext('name,val\nAlice,100\nNA,200', extension='csv') as fn:
csv = CSV(fn)
assert discover(csv).measure.dict['name'] == Option(string)
def test_into_string_on_right():
with filetext('alice,1\nbob,2', extension='.csv') as source:
assert into([], source) == [('alice', 1), ('bob', 2)]
def test_empty_dataframe():
with filetext('name,val', extension='csv') as fn:
csv = CSV(fn, has_header=True)
df = convert(pd.DataFrame, csv)
assert isinstance(df, pd.DataFrame)
def test_header_argument_set_with_or_without_header():
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
assert into(list, fn) == [('Alice', 100), ('Bob', 200)]
with filetext('Alice,100\nBob,200', extension='csv') as fn:
assert into(list, fn) == [('Alice', 100), ('Bob', 200)]
def test_drop():
with filetext('hello\nworld') as fn:
t = TextFile(fn)
assert os.path.exists(fn)
drop(t)
assert not os.path.exists(fn)
def test_discover_csv_files_without_header():
with filetext('Alice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=False)
df = convert(pd.DataFrame, csv)
assert len(df) == 2
assert 'Alice' not in list(df.columns)
def test_pandas_discover_on_gzipped_files():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02',
open=gzip.open, extension='.csv.gz') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: string, when: datetime}')
assert discover(csv) == ds
def test_discover_from_resource():
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
local = CSV(fn)
remote = resource('ssh://localhost:' + fn)
assert discover(local) == discover(remote)
def test_header_argument_set_with_or_without_header():
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
assert into(list, fn) == [('Alice', 100), ('Bob', 200)]
with filetext('Alice,100\nBob,200', extension='csv') as fn:
assert into(list, fn) == [('Alice', 100), ('Bob', 200)]
def test_into_string_on_right():
with filetext('alice,1\nbob,2', extension='.csv') as source:
assert into([], source) == [('alice', 1), ('bob', 2)]
def test_discover_csv_with_spaces_in_header():
with filetext(' name, val\nAlice,100\nBob,200', extension='csv') as fn:
ds = discover(CSV(fn, has_header=True))
assert ds.measure.names == ['name', 'val']
def test_csv_into_list():
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
L = into(list, fn)
assert L == [('Alice', 100), ('Bob', 200)]
def test_raise_errors_quickly_on_into_chunks_dataframe():
with filetext('name,val\nAlice,100\nBob,foo', extension='csv') as fn:
ds = datashape.dshape('var * {name: string, val: int}')
csv = CSV(fn, header=True)
assert raises(Exception,
lambda: CSV_to_chunks_of_dataframes(csv, dshape=ds))
def test_discover_csv_yields_string_on_totally_empty_columns():
expected = dshape('var * {a: int64, b: string, c: int64}')
with filetext('a,b,c\n1,,3\n4,,6\n7,,9') as fn:
csv = CSV(fn, has_header=True)
assert discover(csv) == expected
def test_empty_dataframe():
with filetext('name,val', extension='csv') as fn:
csv = CSV(fn, has_header=True)
df = convert(pd.DataFrame, csv)
assert isinstance(df, pd.DataFrame)
def test_discover():
with filetext('name,balance\nAlice,100\nBob,200') as fn:
local = CSV(fn)
remote = SSH(CSV)(fn, hostname='localhost')
assert discover(local) == discover(remote)
def test_csv_separator_header():
with filetext('a|b|c\n1|2|3\n4|5|6', extension='csv') as fn:
csv = CSV(fn, delimiter='|', has_header=True)
assert convert(list, csv) == [(1, 2, 3), (4, 5, 6)]
def test_discover_csv_files_without_header():
with filetext('Alice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=False)
df = convert(pd.DataFrame, csv)
assert len(df) == 2
assert 'Alice' not in list(df.columns)
def test_raise_errors_quickly_on_into_chunks_dataframe():
with filetext('name,val\nAlice,100\nBob,foo', extension='csv') as fn:
ds = datashape.dshape('var * {name: string, val: int}')
csv = CSV(fn, header=True)
assert raises(Exception,
lambda: CSV_to_chunks_of_dataframes(csv, dshape=ds))
def test_discover_csv_yields_string_on_totally_empty_columns():
expected = dshape('var * {a: int64, b: ?string, c: int64}')
with filetext('a,b,c\n1,,3\n4,,6\n7,,9') as fn:
csv = CSV(fn, has_header=True)
assert discover(csv) == expected
def test_csv_into_list():
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
L = into(list, fn)
assert L == [('Alice', 100), ('Bob', 200)]
def test_discover_from_resource():
with filetext('name,balance\nAlice,100\nBob,200', extension='csv') as fn:
local = CSV(fn)
remote = resource('ssh://localhost:' + fn)
assert discover(local) == discover(remote)
def test_convert():
with filetext('Hello\nWorld') as fn:
assert convert(list, TextFile(fn)) == ['Hello\n', 'World']
def test_discover_csv_without_columns():
with filetext('Alice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn)
ds = discover(csv)
assert '100' not in str(ds)