Trouble parsing string

Question

I need to parse strings in the format "City State
Abbreviation Zipcode" (ie "Glenview, IL 60062") to seperate them as
their own variables. Problem is that sometimes there is a comma
after city, sometimes not, so I've resorted to REfind to seperate
the string. Here's the snippet, "city" is the entire string I
mentioned above. The problem is that the refind I use seems to be
returning 0. I'm trying to find the two capital letters that
designate the state abbeviation.
<cfif city neq ''>
<cfset crpos = refind("[:upper:][:upper:]",city) >
<cfset zip = trim(right(city,len(city)-crpos))>
<cfset citystate = trim(left(city,crpos)) >
<cfset newpos = find("#chr(32)#",citystate) >
<cfset state =
trim(right(citystate,len(citystate)-newpos)) >
<cfset actualcity =
trim(left(citystate,len(citystate)-newPos)) >
</cfif>

Answer 1

I probably should mention some explaination about what the
regular expression is doing:
Note: Groups are RegExp statements surrounded by ()
Group 1: Combination of Letters and Spaces (e.g. City Name)
optional comma and (required) space
Group 2: 2 Character upper case state code (State Code) (note
- depending on your source, state codes may not always be upper
case)
(required) space
Group 3: 5 digit string (e.g. Zip Code) (note - again,
depending on your source, you may be getting 5 digit zip + 4 or
even non-us zip codes that may involve alpha characters.)
The replace function is using back references to refer to the
text matched by group 1,2 and 3.