Lars Gullik Bjønnes wrote:
>>> bool BranchList::remove(string const & s)
>>> {
>>> List::size_type const size = list.size();
>>> - list.remove_if(bind2nd(match(), s));
>>> + list.remove_if(bind(equal_to<string>(),
>>> + bind(&Branch::getBranch, _1),
>>> + s));
>>> return size != list.size();
>>> }
>>
> | Lars! Be nice to us! Add comments explaining what this stuff is
> | equivalent to. If you're not careful, your fellow developers will
> | start to view you as the evil genius who obfuscates their
> | perfectly reasonable code and then laughs at their confusion, "Mu
> | hah hah hah!"
>
> Where is the obfuscation?
>
> Actually I find the new version a lot clearer than the old one. Now
> you can actually see what "match" is doing: equal_to. std::equal_to
> takes to parameters, the elements to be compared, in this case:
> Branch::getBranch() and s.
>
> So: remove the element if equal_to is true when supplied with the
> parameters Branch::getBranc() (for each element of the list) and s.
Yes, I understand that, because I have used bind a lot myself.
However, a comment to this effect would be good too.
> | Would boost::lambda help here?
>
> The code would probably be even clearer with boost::lambda.
Try me. I've never actually used it, so would like to see a real-world
example
> PS. To get the more "obfuscated" stuff you should have picked the
> make_function_output_iterator stuff. Or the operator-> stuff...
I wasn't trying to start a competition on who can write the most
obfuscated code. I was complaining that you are in danger of being
viewed in such a light. Unfair to you and unfair to the rest of us.
--
Angus
