On 25 Mar 2015, at 17:02, Dimitre Novatchev dnovatchev@xxxxxxxxx
<xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:

> And if you just need the boolean  (true()/false()) indicating whether
> a sequence has duplicate values -- not needing the values themselves
> -- then use the following, slightly adjusted and more efficient
> expression:
>
> exists($vSeq[index-of($vSeq,.)[2]][1] )
>

I think that if there are no duplicates, this is O(n^2), whereas the
distinct-values solution is O(n log n). Harder to judge how they compare if
duplicates are more probable: I think this is O(m*n) where n is the size of
the sequence and m is the expected number of items between two duplicates,
i.e. m=1/p where p is the probability of an item being a duplicate.

Michael Kay
Saxonica