def test_merge_omd():
target = [{'a': 'A'}, {'a': 'aleph'}]
assert glom(target, Merge(init=OMD)) == OMD({'a': 'aleph'})
assert glom(target, Merge(init=OMD,
op='update_extend')) == OMD([('a', 'A'),
('a', 'aleph')])
def test_merge():
target = [{'a': 'A'}, {'b': 'B'}]
assert glom(target, Merge()) == {'a': 'A', 'b': 'B'}
assert glom(target, Merge(op=dict.update)) == {'a': 'A', 'b': 'B'}
with pytest.raises(ValueError):
Merge(init=list) # has no .update()
with pytest.raises(ValueError):
Merge(op='update_extend') # doesn't work on base dict, the default init
def fuzzy_match(
df: pd.DataFrame,
source_column: Union[str, int],
target: Iterable[str],
scorer: Callable = textdistance.jaro_winkler.distance,
key: Callable = lambda x: x,
maximize: bool = False,
debug=False,
):
source = df[source_column].astype("str").unique()
source_diff = list(set(source) - set(target))
target_diff = list(set(target) - set(source))
matches = fuzzy_matching_best(
source=source_diff,
target=target_diff,
key=key,
scorer=scorer,
maximize=maximize,
)
replacements_spec = Merge([{T["source"]: "target"}])
replacements_dict = glom(matches, replacements_spec)
distances_spec = ([{T["source"]: "distance"}], Merge())
distances_dict = glom(matches, distances_spec)
if debug:
debug_col_name = f"{source_column}_match_from_target"
return df.pipe(
lambda df: df.assign(
**{
debug_col_name: df[source_column].replace(replacements_dict),
"distance": df[source_column]
.replace(distances_dict)
.replace(r"\D+", 0, regex=True),
}
)
).set_index([source_column, debug_col_name, "distance"])
else:
return df.pipe(
lambda df: df.assign(
**{source_column: df[source_column].replace(replacements_dict)}
)
)
def test_merge():
target = [{'a': 'A'}, {'b': 'B'}]
assert glom(target, Merge()) == {'a': 'A', 'b': 'B'}