inconsistency? Re: Boy do REs ever suck sometimes...

[Tim Peters]

[Roman Suzi]
> I think the following behaviour of re module is incorrect:
>
> >>> re.match("^(123)($)", "123\n").group(1, 2)
> ('123', '')
>
> If $ matches \n, then why doesn't it include it to the
> group?

Because the antecedent is incorrect:  $ is a "zero-width assertion", like \b
and ^:  if it matches it all, it matches a null string.  The only useful
behavior of zero-width assertions is to make a path of the match attempt
fail if the asserted condition doesn't hold.