Natural sort comparison
- Posted by Spock Jul 19, 2017
- 2205 views
Forked from Re: sort constants confusing
A demo of the new sorting include is: http://openeuphoria.org/pastey/293.wc
It provides stable sorting, column sorting, index sorting, along with number and natural order options. All sorts are in the up direction.
After changing a few identifiers this works on OE and Phix.
_tom
Thanks for merging that. I have some questions about the Natural Sort comparator..
Because this is implemented at the compare level there would be quite a lot of overhead at each iteration computing 'natural' values for each target pair prior to the actual comparison. This has got to be quite costly but I'm not sure what the best way of solving it is. I suppose if the data set were always smallish, eg, n < 200 then it doesn't really matter.
Another question is: What would happen if we were sorting a mixture of atoms and strings? The code presented would fail in its fundamental purpose as atoms and sequences would end up being clumped (albeit in nicely sorted clumps).
Lastly, what about things like whitespace, decimal points etc.. the definition of what constitues 'naturalness' is a bit fuzzy. I mean, shouldn't we be also monocasing the text as well?
Spock
