Re: signature discrepancy between equal_range and upper_bound

From: Nat Linden Reply to: boost-users@lists.boost.org Date: Tue, 6 Dec 2011 08:58:18 -0500 Newsgroups: gmane.comp.lib.boost.user Followup to: newsgroup References: On Tue, Dec 6, 2011 at 5:34 AM, wrote:

...

The algorithm equal_range takes the Value by constant reference while the algorithm upper_bound takes Value by value. Why is that? =A0(Note that standard algorithms take a constant reference.)

Sorry to be obtuse, but... are you talking about a Boost library? If so, which one?

Re: [range] signature discrepancy between equal_range and upper_bound

From: Marshall Clow Reply to: boost-users@lists.boost.org Date: Tue, 6 Dec 2011 11:29:34 -0800 Newsgroups: gmane.comp.lib.boost.user Followup to: newsgroup References: On Dec 6, 2011, at 11:17 AM, giecrilj@stegny.2a.pl wrote:

I think what the OP is saying is that in the STL, we have: equal_range ( ForwardIterator first, ForwardIterator last, c onst T& value ); and upper_bound ( ForwardIterator first, ForwardIterator last, c onst T& value );

but in Boost.Range, we have: equal_range ( Range &r, const T& value ); and upper_bound ( Range &r, T value );

And he's wondering why boost::range::upper_bound takes its' parameter by value.

Is that correct?

I wonder why boost ::range ::upper_bound takes its parameter by value. Chris