Hi Gerard,

just a hint from using Levensthein on laaaaarge data sets: vfp is NOT a language well fitted for implementing Levenstein in.
The speed-up my vfp version creates pales in comparison to an implementation in C - think about tenfold - but I was not allowed to put the C version on wiki, as the whole project was done on contract and only by using a loophole in contract semantics I was able to post my speed-up hints ;-)

Think about the mentioned letter normalitation - in our use case it adds to Levenshtein, but might not do in all...

HTH

thomas

>Hi Thomas . Thanks for your reply
>I'll check LevenshTein and VFP
>Regards,
>Gerard
>
>
>
>>>Apparently, there is a 'Fuzzy Search' option that can be installed for Excel 2012 (I have Excel 2007)
>>> which compares 2 columns and gives a % Match result (e.g. 100%=Exact Match 90%= Near exact etc)
>>>
>>>Anybody know if:
>>>1. This can be done programatically (To exactly replicate what Excle does
>>
>>no idea on Excel implementation
>>
>>>2. Can it be done in Foxpro or DotNet
>>
>>Google Levenshtein and vfp - you will find a ready made routine (IIRC in vfp wiki) I made a bit faster...
>>Fields should be letter-normalized perhaps - replacing all labials for a common one for instance.
>>
>>HTH
>>
>>thomas