
Boost : 
From: Jeremy Siek (jsiek_at_[hidden])
Date: 20001226 12:46:08
I like this formulation of the requirements... it is certainly much
simpler. I've also been thinking in this direction, but have a
couple worries:
One thing that concerns me is that as a result (I think) we get a
disconect between what it means for something to be sorted, and for
something to be binarysearchable.
To put it another way, we may need to have some notion to describe that a
range is in such a state so that the binary search will work for *any*
value object that is used with the comparison functor or operator <.
Normally, the definition of "sorted" ensures this. Related to this, I'm
worried that requirements based on a comparison "expression" hide the
separation between the comparison function and the value object.
For a particular single call to binary search this may not be an issue,
but how would you specify the requirements for a sorted_vector class?
Cheers,
Jeremy
On Sun, 24 Dec 2000, David Abrahams wrote:
abraha> A small correction. I wrote:
abraha>
abraha> > Add the following paragraph after 25.3 [lib.alg.sorting] paragraph 5:
abraha> >
abraha> > 6 A sequence [begin, end) is partitioned with respect to an
abraha> > expression f(e) if there exists a nonnegative integer n such
abraha> > that for all 0 <= i < distance(begin, end), f(j) is true if and
abraha> > only if j < n.
abraha>
abraha> Please imagine that I wrote the following instead:
abraha>
abraha> 6 A sequence [start, finish) is partitioned with respect to an
abraha> expression f(e) if there exists a nonnegative integer n such that
abraha> for all 0 <= i < distance(start, finish), f(*(begin+i)) is true if
abraha> and only if i < n.
abraha>
abraha> Thanks,
abraha> Dave
abraha>
abraha>
abraha>
abraha>
abraha>

Jeremy Siek www: http://www.lsc.nd.edu/~jsiek/
Ph.D. Candidate email: jsiek_at_[hidden]
Univ. of Notre Dame work phone: (219) 6313906

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk