Why are the predicate version of lower_bound and upper_bound passing the iterator value inconsistently?

Question

In upper_bound 's binary predicate, the iterator value is passed as the second argument while in lower_bound the iterator value is passed as the first argument.

What is the reasoning here? And does it matter if I remember this detail or not when writing my own binary predicates?

---

NOTE my reference is www.cplusplus.com (which I am told may not be the best reference) and I confirmed it by looking at the implementation in the stl library shipped with VC++ .

Answer 1

This is intentional. The reasoning is so you can use the same comparator for both algorithms. Consider the descriptions. lower_bound :

Returns an iterator pointing to the first element in the range [first, last) that is not less than (ie greater or equal to) value.

and upper_bound :

Returns an iterator pointing to the first element in the range [first, last) that is greater than value.

Consider that the standard comparator is < . In order to implement both algorithms with only < would require !(elem < value) for one and value < elem for the other. The inversion of the order of arguments just follows directly from that.

This shouldn't affect how you implement your Comparator though.