如何从 String 中过滤掉 Unique 元素并将它们存储到 ArrayList

Question

Given a filtered string. I want to add its substrings to an ArrayList only if it doesn't contain it already.

My code:

Pattern reg = Pattern.compile(",");
ArrayList<String> SANR = reg.splitAsStream(joiner.toString())    
        .filter(role -> role.contains("ABB"))
        .map(str -> str.replaceAll("ABB", ""))
        .collect(Collectors.toCollection(ArrayList::new));

As input, it gets all the values with ABB appended such as [123,123,244,244,255] , so I want to make sure the output is just:

[123,244,255]

Answer 1

You need to apply distinct() operation to ensure that the result doesn't contain duplicates. To achieve it and preserve the initial order of elements in the stream, it uses LinkedHashSet under the hood.

Don't use ArrayList as a type, it makes your code rigid. See What does it mean to "program to an interface"?

By the way, Collectors.toList() as well as Java 16 toList() operation will actually give you an ArrayList , but your code should not depend on it. If the in future there would be designed a better general purpose implementation of the List interface which will be capable of everything that ArrayList does, you'll get it for free if you would not demand specifically an ArrayList . Write your code against interfaces.

Pattern reg = Pattern.compile(",");
List<String> sanr = reg.splitAsStream(joiner.toString())
    .filter(role -> role.contains("ABB"))
    .map(str -> str.replaceAll("ABB" , ""))
    .distinct()
    .collect(Collectors.toList());

Sidenote: by convention uppercase-names should be used for constant fields (ie marked as final ), for local values use camel-case starting with a lower-case letter. And names by itself should be meaningful, if sanr is a some kind abbreviation it should be documented somewhere.

Answer 2

If it fits your use case, I would suggest using a Set abstract data type instead of an ArrayList for this purpose. It allows you to not store any duplicates in the list of Strings.

Check out: Java Sets

If you'd like to keep it as an ArrayList you can do:

myArrayList = new ArrayList<String>(new LinkedHashSet<String>(myArrayList));

to delete any duplicates from "myArrayList"