By default 5.8 uses a merge sort, but you can revert to quicksort if that'd work better for your data. See perldoc -f sort and perldoc sort for more explanation.
| [reply]
[d/l]
[select] |
I wrote a program in C++ to sort web server cache logs by IP address. I discovered that a quicksort took forever to sort the data by the IP address, but a heapsort routine that I had in a programming manual was more efficient. If possible, see if you can utilize a binsort routine. If I recall my algorithms class correctly, binsort runs in O(n) time.
| [reply] |