I have a few lines of text containing street address information from an HTML textarea input. In my application I can't break this up into five separate form input fields (MailingStreet, MailingCity, MailingState, MailingPostalCode and MailingCountry).
Is there a parser that is available in Apex to break this down for me in the five address components? Salesforce must be doing this when they parse the web-to-lead address textarea to populate the Lead address field.
I have provided a similar solution in this (SFSE) forum link. I have used googleapis as parser to parse the text area into different text fields. It is a complete example to start with. You can update further as per your requirement. Hope this helps.
You can parse it by writing your own code in trigger if address is separated by any delimiter like ',' and the address is in fixed order like house no., city, state, postal code, country.
You can use following code in your before insert/update trigger:
List<String> splittedAddress = TextAreaField__c.split(',');
String street = splittedAddress.get(0);
String city = splittedAddress.get(1);
String state = splittedAddress.get(2);
String postalCode = splittedAddress.get(3);
String country = splittedAddress.get(4);
Hope it helps. Thanks
I know this is pretty late, but maybe someone else will see this and find it useful.
I wrote my own class to do just this. It assumes that the addresses are primarily US addresses (as is the case for the org I wrote it in), and then made adjustments for Canadian and UK zip codes. I know that it won't work for Somalia or Montserrat because of their zip codes.
It's a surprisingly difficult thing to do because of the varied ways that addresses are written. But this has worked correctly for every address in my org.
public class addressParser {
public static Map<string, string> parseAddress(String address){
address.replace('\r\n', '\n').replace('\r', '\n');
Map<String, String> addressMap = new Map<String, String>();
String tempAddress = address.trim();
Boolean International = false;
String Country = '';
if (checkForCountryString(address) != ''){
country = checkForCountryString(address).trim();
tempAddress = address.removeEnd(country).trim();
}
String zipAddress = tempAddress;//zip should be after a comma or after a line break if there is none
if (tempaddress.contains(','))
zipAddress = tempAddress.remove(tempAddress.split(',')[0]);
else if (tempaddress.contains('\n'))
zipAddress = tempAddress.remove(tempAddress.split('\n')[0]);
Pattern zipPattern = Pattern.compile('\\d{5}(?:[-\\s]\\d{4})?'); //US Zip code pattern, accounts for 12345, 12345-6789 or 12345 6789 patterns at the end of the address string
Matcher zipMatcher = zipPattern.matcher(zipAddress);
String zip = '';
while (zipMatcher.find()){
if (zipMatcher.group() != null && zipMatcher.group() != '')
zip = zipMatcher.group(); //zip code that matches us pattern
}
if (zip == null || zip == ''){
International = true;
addressMap = getInternationalZip(zipAddress);
zip = addressMap.get('Postal Code');
if (addressMap.containsKey('Country'))//&& tempAddress.contains(addressMap.get('Country'))
tempAddress = address.removeEnd(addressMap.get('Country')).trim();
}
else
addressMap.put('Postal Code', zip);
if (country != ''){
addressMap.put('Country', country);
tempAddress = tempAddress.remove(Country).trim();
}
tempAddress = tempAddress.trim().remove(zip).trim();
if (tempAddress.substringAfterLast(',').trim() == '' )//If comma is between state & zip or zip and country, remove it
tempAddress = tempAddress.removeEnd(',');
String state = tempAddress.substringAfterLast(',').trim(); // state is block after comma that isn't zip
if ((zip != null && zip != '') && (state == null || state == '') && International == false){
state = tempAddress.substringAfterLast(' ').trim(); //if user didn't add comma, only space
String temp2 = tempAddress.removeEnd(state).trim();
String stringBeforeState = temp2.substringAfterLast(' ').trim();
if (stringBeforeState.contains('New') || stringBeforeState.contains('North') || stringBeforeState.contains('South') || stringBeforeState.contains('East') || stringBeforeState.contains('West'))
state = stringBeforeState + ' '+ state;
}
addressMap.put('State', state);
tempAddress = tempAddress.removeEnd(state);
tempAddress = tempAddress.trim().removeEnd(',');
String city = tempAddress.trim().substringAfterLast('\n').trim();//in between line break and comma
if (city == null || city == '')
city = tempAddress.substringAfterLast(','); //If comma was used in place of a line break
addressMap.put('City', city);
tempAddress = tempAddress.remove(city);
String street = tempAddress;
addressMap.put('Street', street);
if (addressMap.get('Country') != '' && !International)
addressMap.put('Country', 'USA');
return addressMap;
}
public static Map<String, String> getInternationalZip(String Address){
Map<String, String> addressMap = new Map<String, String>();
String zip = '';
Map<string, string> regexMap = new Map<String, String>();
regexMap.put('Canada', '[ABCEGHJKLMNPRSTVXY]{1}\\d{1}[A-Z]{1} *-* *\\d{1}[A-Z]{1}\\d{1}');
regexMap.put('UK', '(GIR 0AA)|((([A-Z-[QVX]][0-9][0-9]?)|(([A-Z-[QVX]][A-Z-[IJZ]][0-9][0-9]?)|(([A-Z-[QVX]][0-9][A-HJKPSTUW])|([A-Z-[QVX]][A-Z-[IJZ]][0-9][ABEHMNPRVWXY])))) [0-9][A-Z-[CIKMOV]]{2})');// Official UK Regex
regexMap.put('Netherlands', '[0-9]{4}\\s*[A-Za-z]{2}');
for (String s: regexMap.keyset()){
Pattern zipMatcher = Pattern.compile(regexMap.get(s));
Matcher zipMatch = zipMatcher.matcher(Address);
while (zipMatch.find())
zip = zipMatch.group();
if (zip != '' && zip != null){
addressMap.put('Postal Code', zip);
addressMap.put('Country', s);
return addressMap;
}
}
if (zip == '' || zip == null)
addressMap = getZip(address);
//Not included -- Somalia AA NNNNN Montserrat MSR 1110-1350
return addressMap;
}
public static Map<String, String> getZip(String address){ //If all else fails, get the string containing numbers from the lowest line available
Map<String, String> addressMap = new Map<String, String>();
Boolean foundZip = false;
Pattern containsNumbers = Pattern.compile('.*[0-9].*');
String zip = '';
List<String> lines = address.split('\r\n');
for(Integer i = (lines.size()-1); i >= 0; i--){
String s = lines[i].trim();
Matcher mixedMatcher = containsNumbers.matcher(s);
if (mixedMatcher.matches()){
List<String> tokens = s.split(' ');
for (string token: tokens){
Matcher zipMatcher = containsNumbers.matcher(token);
if (zipmatcher.matches())
zip += token;
}
break;
}
}
addressMap.put('Postal Code', zip);
return addressMap;
}
public static string checkForCountryString(String address){ //We're going to assume that it's a country string if the last line is all characters and there are at least 2 preceding lines
String country = '';
Pattern onlyLetters = Pattern.compile('^[a-zA-Z\\s]*$');
List<String> lines = address.split('\r\n');
if (address.contains('<br/>'))
lines = address.split('<br/>');
if (lines.size() > 2){
String lastLine = lines[(lines.size()-1)];
Matcher lettersMatcher = onlyLetters.matcher(lastLine.trim());
if (lettersMatcher.matches())
country = lastLine;
}
return country;
}
}
@brezotom, that is fantastic. Will need a little tweaking, but a great start!
For anybody reading this, I'm starting to put this into a trigger, and calling it like this:
trigger beforeAccountActions on Account (before insert, before update) {
for(Account a:Trigger.New) {
if(String.isBlank(a.BillingStreet) && !String.isBlank(a.BillingAddressParse__c)) {
Map<String,String> bAdd = addressParser.parseAddress(a.BillingAddressParse__c);
if(!bAdd.isEmpty()) {
if(bAdd.containsKey('Street')) a.BillingStreet = bAdd.get('Street');
if(bAdd.containsKey('City')) a.BillingCity = bAdd.get('City');
if(bAdd.containsKey('State')) a.BillingState = bAdd.get('State');
if(bAdd.containsKey('Postal Code')) a.BillingPostalCode = bAdd.get('Postal Code');
if(bAdd.containsKey('Country')) a.BillingCountry = bAdd.get('Country');
}
}
}
}
