Don't parse dates with more than 4 digits for the year

[heinrich5991]

The regex was broken before, using (?!…) instead of (?<=…).

[MasterofJOKers]

Why do we need to lookbehind and lookahead? Can't we get away with something like this?

r = re.compile(r'(?:\b|[_-])('\
               r'(?:[0-9]{1,2}[./-][0-9]{1,2}[./-](?:[0-9]{4}|[0-9]{2}))|'\
               r'(?:(?:[0-9]{4}|[0-9]{2})[./-][0-9]{1,2}[./-][0-9]{1,2})|'\
               r'(?:[0-9]{1,2}\. +[^\W\d_]{3,9} (?:[0-9]{4}|[0-9]{2}))|'\
               r'(?:[^\W\d_]{3,9}(?: [0-9]{1,2},)? [0-9]{4})'\
               r')(?:\b|[_-])')

In some manual testing, it seems to match everything matched in the unit tests. We can then use m.group(1) for the date-part of the matched string.

👍 for the additional tests.

[heinrich5991]

Updated with the suggestion to not use lookahead/lookbehind.

[heinrich5991]

Removed all the superfluous (?:).

[MasterofJOKers]

Removed all the superfluous (?:).

Great, could you also remove the superfluous \ in the [], while you're at it?

[heinrich5991]

Removed the superfluous backslashes in the regex.