Deleting Duplicate Data in a List

Question

The following is what my List looks like, it is a web service call that pulls Itinerary data from an external source. There are two passengers for this specific call and I don't need both of their information because its the same flight, how can I remove this extra data?

Test Flight(

ItineraryFlights:[accountCompanions=(), active=true, airline=Delta, arrival=2013-10-12T15:38:00, confirmationNumber=1234565, departure=2013-10-12T12:05:00 , flightNumber=15161, fromAirport=Atlanta, itinerary=null, itineraryDay=2013-10-12T12:05:00, 
itineraryFlightID=3492, status=Active, success=true, toAirport=San Juan], 

ItineraryFlights:[accountCompanions=(), active=true, airline=Delta, arrival=2013-10-12T15:38:00, confirmationNumber=1234565, departure=2013-10-12T12:05:00 , flightNumber=15160, 
fromAirport=Atlanta, itinerary=null, itineraryDay=2013-10-12T12:05:00, itineraryFlightID=3493, status=Active, success=true, toAirport=San Juan], 

ItineraryFlights:[accountCompanions=(), active=true, airline=Delta, arrival=2013-10-12T15:38:00, 
confirmationNumber=1234565, departure=2013-10-12T12:05:00 , flightNumber=15161, fromAirport=Atlanta, itinerary=null, itineraryDay=2013-10-12T12:05:00, itineraryFlightID=3494, status=Active, success=true, toAirport=San Juan], 

ItineraryFlights:[accountCompanions=(), 
active=true, airline=Delta, arrival=2013-10-12T15:38:00, confirmationNumber=1234565, departure=2013-10-12T12:05:00 , flightNumber=15160, fromAirport=Atlanta, itinerary=null, itineraryDay=2013-10-12T12:05:00, itineraryFlightID=3495, status=Active, success=true, 
toAirport=San Juan])

By concatenation of the airline and flight number you can find the dupes and assume all the data underneath it is similar and can be removed. This is what I was thinking for my logic, but I'm unsure now. Am I making this too complicated?

private static List<ServiceBusServices.ItineraryFlights> getFlightInfo(String dryFlyID)
{   
    intineraryFlights = TripProfileProxy.retreiveFlightInformation(dryFlyID);
    Diagnostics.debug('intineraryFlights' + intineraryFlights);
    Integer count = -1;
    String concat;
    String airlines;
    String flightNumber;
    for(ServiceBusServices.ItineraryFlights intinF : intineraryFlights)
    {

        if(count <  0)
        {
            airlines = intinF.airline;
            flightNumber = intinF.flightNumber;
            concat = airlines + ' ' + flightNumber;

        }
        if(intinF.airline + ' ' + intinF.flightNumber == concat)
        {
            intineraryFlights.remove(not sure what do.);
            count + 1;
        }
    }
    return intineraryFlights;
}

Answer 1

Simply making a map would suffice. I'd make a custom class to actually hold the key, because simple concatenation might be bad. Here's an implementation.

First, this class should be saved somewhere. I use it pretty much everywhere when I can help it, so I have it as a top-level class, along with NameValuePair, a class that I use for URL parsing/generation.

class KeyPair {
    String key1, key2;
    public KeyPair(String k1, String k2) {
        key1 = k1;
        key2 = k2;
    }
    public Boolean equals(Object other) {
        Keypair kp = (KeyPair)other;
        return kp.key1 == key1 && kp.key2 == key2;
    }
    public Integer hashCode() {
        return System.hashCode(key1+key2);
    }
}

For the actual function, you pretty much just need to use the map against your callout results:

private static List<ServiceBusServices.ItineraryFlights> getFlightInfo(String dryFlyID)
{   
    intineraryFlights = TripProfileProxy.retreiveFlightInformation(dryFlyID);
    Map<KeyPair, ServiceBusServices.ItineraryFlights> flights = new Map<KeyPair, ServiceBusServices.ItineraryFlights>();
    for(ServiceBusServices.ItineraryFlights flight: itineraryFlights) {
        KeyPair kp = new KeyPair(flight.airline, flight.flightNumber);
        if(!flights.containsKey(kp)) {
            flights.put(kp, flight);
        }
    }
    return (intineraryFlights = flights.values());
}

Answer 2

I think you need to do something more like the following:

A list can't contain more than one identical value when performing DML. If you try to add a 2nd value that's identical to it, you could throw a list exception at some point in your code later. So, each time before you add a new value to the list, you run a test to see whether it contains the new value. If it does, you've found a duplicate and can either "break" or add it to a list of dupes so you can continue looking for additional duplicate values.

Another way that would work if you have a unique identifier that you can use as a map key, would be to put the strings into a map and test for whether the map.values() contain the new string value before you add each additional key and string to the map.

If necessary, you could use something as simple as a count integer for the key as that would always be unique for that run through your class. Again, if a dupe was found, you could put it to a different map or add it to a different list, assuming you need to keep track of them (I'm guessing you probably don't).