def test_fancy(self):
new = TypedDfs.typed("a class", doc="A doc").build()
assert new.__name__ == "a class"
assert new.__doc__ == "A doc"
df = new.convert(pd.DataFrame(sample_data()))
assert isinstance(df, TypedDf)
assert df.__class__.__name__ == "a class"
def test_fancy_with_no_index(self):
new = TypedDfs.typed("a class").require("abc", "123", "xyz", index=False).build()
df = new.convert(pd.DataFrame(sample_data()))
assert df.index_names() == []
assert df.column_names() == ["abc", "123", "xyz"]
assert isinstance(df, TypedDf)
assert df.__class__.__name__ == "a class"
def test_matrix(self):
clazz = (TypedDfs.matrix("pretty bird matrix").doc(
"A table of species of pretty birds and whether they like each other."
).dtype(bool)).build()
z = DfCliHelp.help(clazz)
#
# test long text
#
txt = z.get_long_text()
lines = txt.splitlines()
expected_lines = 3 + 1 + len(FileFormat)
assert len(
lines
) == expected_lines, f"{len(lines)} != {expected_lines}: {txt}"
assert "whether they like each other" in lines[1]
assert "cast to bool" in lines[2]
#
# test short text
#
txt = z.get_short_text(recommended_only=True)
lines = txt.splitlines()
expected_lines = 4
assert len(
lines
) == expected_lines, f"{len(lines)} != {expected_lines}: {txt}"
assert "Boolean (bool)" in lines[2]
def test_wrap_multilayer(self):
# not fully supported yet, but let's check that it's reasonable
rows = ["yes", "no", "maybe"]
cols = [("animal", "cat"), ("animal", "armadillo", ("person", "matt"))]
cols = pd.MultiIndex.from_tuples(cols)
df = pd.DataFrame(np.zeros((3, 2)), rows, cols)
df = TypedDfs.wrap(df)
assert df.column_names() == [
("animal", "cat", np.nan),
("animal", "armadillo", ("person", "matt")),
]
def test_typed(self):
clazz = (TypedDfs.typed("pretty bird table").doc(
"A table of species of pretty birds and their characteristics.").
require("species",
dtype=str).require("prettiness", dtype=float).require(
"cuteness",
dtype=np.int16).reserve("lifespan",
dtype=Period)).build()
z = DfCliHelp.help(clazz)
#
# test long text
#
txt = z.get_long_text(nl="\n")
lines = txt.splitlines()
expected_lines = 1 + 1 # header and docstring
expected_lines += 1 + 3 # required columns
expected_lines += 1 + 1 + 1 # optional columns, inc. "any others allowed" (e.g.)
expected_lines += 1 + len(FileFormat) # file formats
assert len(
lines
) == expected_lines, f"{len(lines)} != {expected_lines}: " + txt
assert "pretty bird table" in lines[0]
assert "table of species" in lines[1]
assert "columns" in lines[2].lower()
assert "- species (string)" in lines[3]
assert "- prettiness (floating-point)" in txt
assert "- cuteness (integer)" in txt
assert "- lifespan (time period)" in txt
assert "[avoid]" in txt or "[discouraged]" in txt or "[not recommended]" in txt
#
# test short text
#
txt = z.get_short_text(recommended_only=True, nl="\n")
lines = txt.splitlines()
expected_lines = 4 # 1 for header, 1 for doc, 1 for formats, 1 for typing
assert len(
lines
) == expected_lines, f"{len(lines)} != {expected_lines}: {txt}"
assert "pretty bird table" in lines[0]
assert "table of species" in lines[1]
assert "species (str)" in lines[2]
assert "prettiness (float)" in lines[2]
assert "lifespan (period)" in lines[2]
assert "csv" in lines[3]
def test_no_name_fancy(self):
with pytest.raises(TypeError):
# noinspection PyTypeChecker
TypedDfs.typed(None).build()
def test_no_name_simple(self):
with pytest.raises(TypeError):
# noinspection PyTypeChecker
TypedDfs.untyped(None)
def test_empty_simple(self):
new = TypedDfs.untyped("a class")
df = new.convert(pd.DataFrame())
assert list(df.columns) == []
def test_wrap(self):
df = pd.DataFrame({})
df2 = TypedDfs.wrap(df)
assert not isinstance(df, AbsDf)
assert isinstance(df2, BaseDf)
def test_missing(self):
new = TypedDfs.typed("a class").require("qqq", index=False).strict().build()
df = pd.DataFrame(sample_data())
with pytest.raises(TypedDfs.MissingColumnError):
new.convert(df)
def test_extra_index(self):
new = TypedDfs.typed("a class").require("xyz", index=False).strict().build()
df = pd.DataFrame(sample_data())
with pytest.raises(TypedDfs.UnexpectedColumnError):
new.convert(df)