如何使用 Stream 从 Java 中的字符串中提取三元组

Question

Let's say I have a String = "abcabc". trigrams are all unique consecutive groups of three letters, in this case

abc
bca
cab

I would like to extract them and put them inside a Map, ideally counting them. So I'll get

abc, 2
bca, 1
cab, 1

I would like to do that with Streams, not with loops.

My idea is

1. convert the string to a HashMap: (index, 3-characters String)
2. collect, group, and count over the stream of the Map values

But I can't put it in code...

I came up with something like this:

        String testString = "abcabc";
        //I can't get point 1...this doesn't compile
        Map trigrams = IntStream.range(0, testString.length()-2).collect(Collectors.toMap(i->i, testString.substring(i,i+3));

        //point 2 seems to work
        //Map<Integer,String> trigrams =  Map.of(0,"abc", 1, "bca",2,"cab",3,"abc");
        
        List<String> trigramsList = trigrams.values().stream().collect(Collectors.toList());
        Map<String, Long> result = trigramsList.stream()
                .collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));

Also I'm not convinced about IntStream, maybe there is another way.

Answer 1

This can be simplified somewhat.

import java.util.Map;
import java.util.function.Function;
import java.util.stream.IntStream;

import static java.util.stream.Collectors.*;

public class Main {

    public static void main(String[] args) {
        String s = "a";

        Map<String, Long> map = IntStream.range(0, s.length() - 2)
                .mapToObj(i -> s.substring(i, i + 3))
                .collect(groupingBy(Function.identity(), counting()));

        System.out.println(map);
    }
}

But, keep in mind that substring() method will throw an exception if the second index is greater than the length of the string or the second index is less than the first. This only works because IntStream.range() generates numbers in ascending order. So, IntStream.range(0, -2) returns an empty stream.

It would be better if you explicitly check that there are at least 3 characters in the string.

Answer 2

One possibility would be to use Guava Mutliset :

Mutliset trigrams = IntStream.range(0, testString.length()-2)
    .mapToObj(i->testString.substring(i,i+3))
    .collect(Collectors.toCollection(Mutltiset::new));

If you don't want to use Google library, it could be done like so:

Map trigrams = IntStream.range(0, testString.length()-2)
    .mapToObj(i->testString.substring(i,i+3))
    .collect(Collectors.groupingBy(x->x, Collectors.counting()));

Answer 3

It's very well possible with streams:

    String testString = "abcabc";
    Map<String, Long> trigramCnt = 
            // range through index 0-3
            IntStream.rangeClosed(0, testString.length() - 3)
            // create substrings abc (idx 0-2), bca (idx 1-3), cab (idx 2-4), abc (idx 3-5)    
            .mapToObj(i -> testString.substring(i, i + 3))
            // count them
            .collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));
    System.out.println(trigramCnt);
    //-> {bca=1, abc=2, cab=1}