[ldolse]

Originally Posted by kiwidude

Thanks for the heads up on the decode function. Actually right now I don't do any conversion of those values, I just ignore them! So it was purely by coincidence (that i and e have the same soundex values and were following each other) that it worked for you in this case.

However your suggestion to use this for all comparisons I think is an inspired idea. Not I think on the "identical" matches, but on every other kind of match I think it makes sense? Obviously I don't know the downside if any to that function - is there any reason not to use it everywhere other than performance which isn't really much of a concern imho versus the improved matching?

Agree it doesn't make sense for the 'identical' algorithm. I'm not sure what the performance hit would be - it looks like you can pass whole strings over to it, and it does a decode test there, filenames.py has an example. I suspect it wouldn't be much.

Originally Posted by kiwidude

The "last set" wonkiness sounds very odd and I haven't been able to replicate it. Was that showing one group at a time or all at once? You said finishing "all/most" - which was it? When did it go wonky - how many sets were remaining?

It was late when I ran into it, and my vagueness was because I didn't go back and try to reproduce with notes. Will try again and give you more specifics since you couldn't easily reproduce it yourself.

Thanks for the other tips - glad that's all already trivial to do.