I wrote a parser at one point but no longer have the code.

It's not that hard to draw heuristical rules to follow. In short, what you want to do is break the address down into "words". High probability that the first word is the street number. If there are more than 3 words, then check the last to see if a suite number or office number. If not, then the last word is the "Ave" or "St" and the middle words are the name of the street. If so, then the middle words are the street name, the second to last is the "Ave" or "St".

To trap standard abbreviations, you want to build a table of them and check against it.


>I'm sure someone has done this before. Looking for some quick code...
>
>Address: 12345 ABC ST
> 123 NBC AVE
> 1432 CBS DR
>
>Have to trun each in thre delimited fields.
>
>The syntax i've been working with goes something like"
>A = substr(property.address, 1," ")
>B = sbustr ( property.addresws, " ", " ")
>C = substr(property.address, (find where the streetname stops), (where the street suffix begins))
>
>Anybody got some code laying on in a drawer somewhere that does this?
>
>Thanks
>CMH.