def test_cat_with_header():
table1 = (('bar', 'foo'),
('A', 1),
('B', 2))
table2 = (('bar', 'baz'),
('C', True),
('D', False))
actual = cat(table1, header=['A', 'foo', 'B', 'bar', 'C'])
expect = (('A', 'foo', 'B', 'bar', 'C'),
(None, 1, None, 'A', None),
(None, 2, None, 'B', None))
ieq(expect, actual)
ieq(expect, actual)
actual = cat(table1, table2, header=['A', 'foo', 'B', 'bar', 'C'])
expect = (('A', 'foo', 'B', 'bar', 'C'),
(None, 1, None, 'A', None),
(None, 2, None, 'B', None),
(None, None, None, 'C', None),
(None, None, None, 'D', None))
ieq(expect, actual)
ieq(expect, actual)
def test_cat_dupfields():
table1 = (('foo', 'foo'),
(1, 'A'),
(2,),
(3, 'B', True))
# these cases are pathological, including to confirm expected behaviour,
# but user needs to rename fields to get something sensible
actual = cat(table1)
expect = (('foo', 'foo'),
(1, 1),
(2, 2),
(3, 3))
ieq(expect, actual)
table2 = (('foo', 'foo', 'bar'),
(4, 'C', True),
(5, 'D', False))
actual = cat(table1, table2)
expect = (('foo', 'foo', 'bar'),
(1, 1, None),
(2, 2, None),
(3, 3, None),
(4, 4, True),
(5, 5, False))
ieq(expect, actual)
def test_cat_with_header():
table1 = (('bar', 'foo'),
('A', 1),
('B', 2))
table2 = (('bar', 'baz'),
('C', True),
('D', False))
actual = cat(table1, header=['A', 'foo', 'B', 'bar', 'C'])
expect = (('A', 'foo', 'B', 'bar', 'C'),
(None, 1, None, 'A', None),
(None, 2, None, 'B', None))
ieq(expect, actual)
ieq(expect, actual)
actual = cat(table1, table2, header=['A', 'foo', 'B', 'bar', 'C'])
expect = (('A', 'foo', 'B', 'bar', 'C'),
(None, 1, None, 'A', None),
(None, 2, None, 'B', None),
(None, None, None, 'C', None),
(None, None, None, 'D', None))
ieq(expect, actual)
ieq(expect, actual)
def test_cat():
table1 = (('foo', 'bar'), (1, 'A'), (2, 'B'))
table2 = (('bar', 'baz'), ('C', True), ('D', False))
cat1 = cat(table1, table2, missing=None)
expectation = (('foo', 'bar', 'baz'), (1, 'A', None), (2, 'B', None),
(None, 'C', True), (None, 'D', False))
ieq(expectation, cat1)
# how does cat cope with uneven rows?
table3 = (('foo', 'bar', 'baz'), ('A', 1, 2), ('B', '2', '3.4'),
(u'B', u'3', u'7.8', True), ('D', 'xyz', 9.0), ('E', None))
cat3 = cat(table3, missing=None)
expectation = (('foo', 'bar', 'baz'), ('A', 1, 2), ('B', '2', '3.4'),
(u'B', u'3', u'7.8'), ('D', 'xyz', 9.0), ('E', None, None))
ieq(expectation, cat3)
# cat more than two tables?
cat4 = cat(table1, table2, table3)
expectation = (('foo', 'bar', 'baz'), (1, 'A', None), (2, 'B', None),
(None, 'C', True), (None, 'D',
False), ('A', 1, 2), ('B', '2', '3.4'),
(u'B', u'3', u'7.8'), ('D', 'xyz', 9.0), ('E', None, None))
ieq(expectation, cat4)
def test_cat():
table1 = (('foo', 'bar'),
(1, 'A'),
(2, 'B'))
table2 = (('bar', 'baz'),
('C', True),
('D', False))
cat1 = cat(table1, table2, missing=None)
expectation = (('foo', 'bar', 'baz'),
(1, 'A', None),
(2, 'B', None),
(None, 'C', True),
(None, 'D', False))
ieq(expectation, cat1)
# how does cat cope with uneven rows?
table3 = (('foo', 'bar', 'baz'),
('A', 1, 2),
('B', '2', '3.4'),
(u'B', u'3', u'7.8', True),
('D', 'xyz', 9.0),
('E', None))
cat3 = cat(table3, missing=None)
expectation = (('foo', 'bar', 'baz'),
('A', 1, 2),
('B', '2', '3.4'),
(u'B', u'3', u'7.8'),
('D', 'xyz', 9.0),
('E', None, None))
ieq(expectation, cat3)
# cat more than two tables?
cat4 = cat(table1, table2, table3)
expectation = (('foo', 'bar', 'baz'),
(1, 'A', None),
(2, 'B', None),
(None, 'C', True),
(None, 'D', False),
('A', 1, 2),
('B', '2', '3.4'),
(u'B', u'3', u'7.8'),
('D', 'xyz', 9.0),
('E', None, None))
ieq(expectation, cat4)
def test_cat_dupfields():
table1 = (('foo', 'foo'), (1, 'A'), (2, ), (3, 'B', True))
# these cases are pathological, including to confirm expected behaviour,
# but user needs to rename fields to get something sensible
actual = cat(table1)
expect = (('foo', 'foo'), (1, 1), (2, 2), (3, 3))
ieq(expect, actual)
table2 = (('foo', 'foo', 'bar'), (4, 'C', True), (5, 'D', False))
actual = cat(table1, table2)
expect = (('foo', 'foo', 'bar'), (1, 1, None), (2, 2, None), (3, 3, None),
(4, 4, True), (5, 5, False))
ieq(expect, actual)
def fetch_people_table():
planet_fetcher = CachedPlanetFetcher()
first_page_response = _fetch_people_page(1).json()
total_count = first_page_response['count']
fetched_results = first_page_response['results']
fetched_count = len(fetched_results)
remaining_count = total_count - fetched_count
remaining_pages = math.ceil(remaining_count / fetched_count)
table = fromdicts(fetched_results, header=PEOPLE_HEADER)
with ThreadPoolExecutor(max_workers=8) as executor:
response_futures = [
executor.submit(_fetch_people_page, page_number)
for page_number in range(2, 2 + remaining_pages)
]
for future in as_completed(response_futures):
page_response = future.result().json()
table = cat(
table, fromdicts(page_response['results'],
header=PEOPLE_HEADER))
table = addfields(table, [('date', lambda rec: datetime.fromisoformat(rec[
'edited'].replace('Z', '+00:00')).date().isoformat())])
table = cutout(table, 'edited')
table = convert(
table, 'homeworld', lambda homeworld_url: planet_fetcher.fetch(
homeworld_url).json()['name'])
return table
def test_mergesort_3():
table1 = (('foo', 'bar'),
('A', 9),
('C', 2),
('D', 10),
('A', 6),
('F', 1))
table2 = (('foo', 'baz'),
('B', 3),
('D', 10),
('A', 10),
('F', 4))
# should be same as concatenate then sort (but more efficient, esp. when
# presorted)
expect = sort(cat(table1, table2), key='foo', reverse=True)
actual = mergesort(table1, table2, key='foo', reverse=True)
ieq(expect, actual)
ieq(expect, actual)
actual = mergesort(sort(table1, key='foo', reverse=True),
sort(table2, key='foo', reverse=True),
key='foo', reverse=True, presorted=True)
ieq(expect, actual)
ieq(expect, actual)
def test_cat_empty():
table1 = (('foo', 'bar'),
(1, 'A'),
(2, 'B'))
table2 = (('bar', 'baz'),)
expect = (('foo', 'bar', 'baz'),
(1, 'A', None),
(2, 'B', None))
actual = cat(table1, table2)
ieq(expect, actual)
def test_mergesort_4():
table1 = (("foo", "bar", "baz"), (1, "A", True), (2, "B", None), (4, "C", True))
table2 = (("bar", "baz", "quux"), ("A", True, 42.0), ("B", False, 79.3), ("C", False, 12.4))
expect = sort(cat(table1, table2), key="bar")
actual = mergesort(table1, table2, key="bar")
ieq(expect, actual)
ieq(expect, actual)
def test_cat_empty():
table1 = (('foo', 'bar'),
(1, 'A'),
(2, 'B'))
table2 = (('bar', 'baz'),)
expect = (('foo', 'bar', 'baz'),
(1, 'A', None),
(2, 'B', None))
actual = cat(table1, table2)
ieq(expect, actual)
def test_mergesort_4():
table1 = (('foo', 'bar', 'baz'), (1, 'A', True), (2, 'B', None), (4, 'C',
True))
table2 = (('bar', 'baz', 'quux'), ('A', True, 42.0), ('B', False, 79.3),
('C', False, 12.4))
expect = sort(cat(table1, table2), key='bar')
actual = mergesort(table1, table2, key='bar')
ieq(expect, actual)
ieq(expect, actual)
def test_mergesort_4():
table1 = (('foo', 'bar', 'baz'),
(1, 'A', True),
(2, 'B', None),
(4, 'C', True))
table2 = (('bar', 'baz', 'quux'),
('A', True, 42.0),
('B', False, 79.3),
('C', False, 12.4))
expect = sort(cat(table1, table2), key='bar')
actual = mergesort(table1, table2, key='bar')
ieq(expect, actual)
ieq(expect, actual)
def _append_to_avro_file(test_rows1,
test_rows2,
test_schema,
test_expect=None,
print_tables=True):
_show__expect_rows(test_rows1, print_tables)
_show__expect_rows(test_rows2, print_tables)
test_filename = _get_tempfile_path()
toavro(test_rows1, test_filename, schema=test_schema)
appendavro(test_rows2, test_filename, schema=test_schema)
test_actual = fromavro(test_filename)
if test_expect is not None:
test_expect2 = test_expect
else:
test_expect2 = cat(test_rows1, test_rows2)
_assert_rows_are_equals(test_expect2, test_actual, print_tables)
def test_mergesort_2():
table1 = (("foo", "bar"), ("A", 9), ("C", 2), ("D", 10), ("A", 6), ("F", 1))
table2 = (("foo", "baz"), ("B", 3), ("D", 10), ("A", 10), ("F", 4))
# should be same as concatenate then sort (but more efficient, esp. when
# presorted)
expect = sort(cat(table1, table2), key="foo")
actual = mergesort(table1, table2, key="foo")
ieq(expect, actual)
ieq(expect, actual)
actual = mergesort(sort(table1, key="foo"), sort(table2, key="foo"), key="foo", presorted=True)
ieq(expect, actual)
ieq(expect, actual)
def test_mergesort_1():
table1 = (('foo', 'bar'), ('A', 6), ('C', 2), ('D', 10), ('A', 9), ('F',
1))
table2 = (('foo', 'bar'), ('B', 3), ('D', 10), ('A', 10), ('F', 4))
# should be same as concatenate then sort (but more efficient, esp. when
# presorted)
expect = sort(cat(table1, table2))
actual = mergesort(table1, table2)
ieq(expect, actual)
ieq(expect, actual)
actual = mergesort(sort(table1), sort(table2), presorted=True)
ieq(expect, actual)
ieq(expect, actual)