zipfinder.cfc cutting off the first digit in some zip codes

This is how I am importing the results into my application:

<cfinvoke component="zipfinder" method="squareSearch" radius="25"
zip="08826" returnvariable="results"></cfinvoke>


<cfquery name="get_resellers" datasource="TEST">
select *
from PartnerView
where zipcode IN (#ListQualify(ValueList(results.zip),"'")#)
and Country = 'USA'
order by CompanyName
</cfquery>

Here is the full zipfinder.cfc code.  Zip is set as a string in the first line (from what I can tell).  I understand what you are saying and that definitely seems to be the problem, but where does Zip stop being a string? I appreciate the help!

<cfcomponent>   
     <cffunction name="zipLocation" access="public" hint="Given a zip code, it will look up the city and state">
         <cfargument name="zip" type="string" required="true">
        
         <cfquery name="qGetZip" datasource="sc">
             select city,state
             from zipcode
             where zipcode = <cfqueryparam value="#zip#" cfsqltype="CF_SQL_CHAR">
         </cfquery>
        
         <cfreturn qGetZip>
     </cffunction>

     <cffunction name="zipToLL" access="public">
         <!---
             This is a helper function.  Given a zip code, it will look up the
             relevant lats and longs (and their corresponding values in radians)
             and then pass back a structure containing this information --->
            
         <cfargument name="zip" type="string" required="true">
        
         <!--- gets a new coordinate pair --->
         <cfset cp = getNewCoordinate()>   
        
         <cfquery name="qGetLL" datasource="sc">
         select latitude, longitude
         from zipcode
         where zipcode = <cfqueryparam value="#zip#" cfsqltype="CF_SQL_CHAR">
         and latitude IS NOT NULL  <!--- tpullis fix --->
         </cfquery>
        
         <cfif qGetLL.recordcount gt 0>
             <cfset cp.latitude        = qGetLL.latitude>
             <cfset cp.longitude        = qGetLL.longitude>
             <cfset cp.rlatitude     = 0> <!--- qGetLL.rlatitude> --->
             <cfset cp.rlongitude     = 0> <!--- qGetLL.rlongitude> --->
         </cfif>
        
         <cfreturn cp>
     </cffunction>

     <!--- +++++++++++++++++++++++++++++++++++++++++++++ --->       
    
     <cffunction name="getNewCoordinate" access="public">
         <!---
             Basically, this is a constructor that gives us a blank coordinate pair.
          --->
         
         <cfset retVal                 = structNew()>
         <cfset retVal.latitude        = 0>
         <cfset retVal.longitude        = 0>
         <cfset retVal.rlatitude        = 0>
         <cfset retVal.rlongitude    = 0>
         <cfreturn retVal>       
     </cffunction>
    
    
     <!--- +++++++++++++++++++++++++++++++++++++++++++++ --->       
    
     <cffunction name="squareSearch" access="public">   
         <!---
             This function performs a proximity search by building out a rectangle
             from a given set of coordinates, and then returning matching items that
             fall within that area.  It is not the most accurate way to search, but
             for smaller distances, it is okay.  It is also very fast.
         --->
         <cfargument name="radius"     type="numeric"     required="true">
         <cfargument name="zip"         type="string"     required="true">
        
         <cfset radius         = arguments.radius>
         <cfset zip            = arguments.zip>       
         <cfset z1            = zipToLL(zip)>
        
         <cfset lat_miles    = 69.1>    <!--- You can change this if you need more precision --->
         <cfset lon_miles    = abs(lat_miles * cos(z1.latitude * 0.0174))>
         <cfset lat_degrees    = radius / lat_miles>
         <cfset lon_degrees    = radius / lon_miles>
        
         <!--- This is where we calculate the bounds of the search rectangle --->
         <cfset lat1            = z1.latitude - lat_degrees>
         <cfset lat2            = z1.latitude + lat_degrees>
         <cfset lon1            = z1.longitude - lon_degrees>
         <cfset lon2            = z1.longitude + lon_degrees>
        
        
         <!---
             To perform the search, we're going to use trigonometry.  Remember the equation, x^2 + y^2 = z^2,
             aka the Pythazizzle Thizzle? If you look closely, you can see that we are using that in order
             to calculate the distance (dist) in the query below.
            
             This is good, because it is a fast calculation.  But, it is bad because it is calculating the
             distance as if it were a line.  If the world were flat, this would be perfect.  But, since it isn't,
             this will start to show errors the larger the radius gets.
            
             Still, for your applications, the errors might be small enough to justify the BLAZING SPEED.
          --->
         <cfquery name="qSquareSearch" datasource="sc">
             select     zipcode as zip, state, city,
                 SQRT(SQUARE(#lat_miles# * (latitude - (#z1.latitude#)))
                     +
                     square(#lon_miles# * (longitude - (#z1.longitude#)))
                     ) as dist
             from zipcode
             where latitude between #lat1# AND #lat2#
             AND longitude between #lon1# AND #lon2#
             order by dist asc
         </cfquery>
        
         <!---
             This is just a quick filter query that will remove some of the zips that get erroneously
             included in the result set.  This helps to offset the errors that this method introduces,
             but only just a little.
          --->
         <cfquery name="qRefine" dbtype="query">
         select * from qSquareSearch where dist < <cfqueryparam value="#radius#" cfsqltype="CF_SQL_INTEGER">
         </cfquery>
        
         <cfreturn qRefine>   
     </cffunction>
    
     </cffunction>
</cfcomponent>