On Fri, 2014-06-20 at 23:59 +0000, bearophile via Digitalmars-d wrote: > Walter Bright: > > > If you could submit a bugzilla issue and include the faster > > version, that would be great! > > My "faster" version is also doing lot of memory swaps (so it's > not faster in all cases), and it's not generic enough for the > current ranges, and it doesn't switch to a safe O(n ln n) sort > like a HeapSort when the ordering of the input data is bad, like > a safe introsort of a standard library has to do. > > So while I sometimes use it, I don't think it will be an > improvement in the general case.
Both Python (since that is the original) and Java now use modified Timsort as the standard sort. Java also has some specialization but not to the extent of using Radix Sort, which would be a good idea for pure integer data. Per Nordlöw is offering a Radix Sort implementation. I had promised to do a thorough review of his code, but have so far failed to do more than a surface review. I clearly must try harder. -- Russel. ============================================================================= Dr Russel Winder t: +44 20 7585 2200 voip: sip:[email protected] 41 Buckmaster Road m: +44 7770 465 077 xmpp: [email protected] London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder
