def test_clean_fix_unicode_before_truncate():
# [adamhooper, 2019-12-13] I don't think we can actually test this in pure
# Python. We'd need a string that has invalid UTF-8 encoding, and I don't
# know how to generate one. The only thing I know how to generate is
# invalid _Unicode_ with surrogate pairs ... but the replacement character
# happens to have the same number of bytes as an erroneous surrogate.
#
# Oh well. Test that we can actually generate is_unicode_fixed+is_truncated,
# at least.
assert clean_colname("\ud800abcd",
settings=MockSettings(4)) == CleanColname(
"�a", is_unicode_fixed=True, is_truncated=True)
def test_clean_ascii_before_truncate():
assert clean_colname("ab\n\ncd", settings=MockSettings(3)) == CleanColname(
"abc", is_ascii_cleaned=True, is_truncated=True)
def test_clean_truncate_allow_full_unicode_character():
assert clean_colname("acé",
settings=MockSettings(4)) == CleanColname("acé")
def test_clean_truncate_nix_partial_unicode_character():
assert clean_colname("acé", settings=MockSettings(3)) == CleanColname(
"ac", is_truncated=True)
def test_clean_fix_unicode():
assert clean_colname("ab\ud800\udc00cd") == CleanColname(
"ab��cd", is_unicode_fixed=True)
def test_clean_ascii_control_characters():
assert clean_colname("ab\0\n\tcd") == CleanColname("abcd",
is_ascii_cleaned=True)
def test_clean_empty_str():
assert clean_colname("") == CleanColname("")