Case 1: rel #326
If the _parse_numeric_token method exits without having used the token, the token does not get handled any further but also does not get added to the skipped_idxs list.
Case 2: rel #271 #1062
When parsing a string with multiple HH:MM times, they get overwritten each time a new time is encountered, without the overwritten values being added to the token list. For example:
>>> parse("from 13:45 to 17:15", fuzzy_with_tokens=True)
(datetime.datetime(2020, 6, 27, 17, 15), ('from ', ' to '))While case 1 might be solved by fixing the underlying bug with how colons are handled, there might be more edge cases which could trigger this issue. I also do not think that case 2 would be the expected behavior. There should probably be a check whether a value has already been set here, which could perhaps be an issue in other places as well.
Case 1: rel #326
If the
_parse_numeric_tokenmethod exits without having used the token, the token does not get handled any further but also does not get added to theskipped_idxslist.Case 2: rel #271 #1062
When parsing a string with multiple
HH:MMtimes, they get overwritten each time a new time is encountered, without the overwritten values being added to the token list. For example:While case 1 might be solved by fixing the underlying bug with how colons are handled, there might be more edge cases which could trigger this issue. I also do not think that case 2 would be the expected behavior. There should probably be a check whether a value has already been set here, which could perhaps be an issue in other places as well.