Converting a sentence string to a string array of words in Java

Question

I need my Java program to take a string like:

\"This is a sample sentence.\"

and turn it into a string array like:

{\"this\         


        

Answer 1

String.split() will do most of what you want. You may then need to loop over the words to pull out any punctuation.

For example:

String s = "This is a sample sentence.";
String[] words = s.split("\\s+");
for (int i = 0; i < words.length; i++) {
    // You may want to check for a non-word character before blindly
    // performing a replacement
    // It may also be necessary to adjust the character class
    words[i] = words[i].replaceAll("[^\\w]", "");
}

Answer 2

Most of the answers here convert String to String Array as the question asked. But Generally we use List , so more useful will be -

String dummy = "This is a sample sentence.";
List<String> wordList= Arrays.asList(dummy.split(" "));

Answer 3

Try this:

String[] stringArray = Pattern.compile("ian").split(
"This is a sample sentence"
.replaceAll("[^\\p{Alnum}]+", "") //this will remove all non alpha numeric chars
);

for (int j=0; i<stringArray .length; j++) {
  System.out.println(i + " \"" + stringArray [j] + "\"");
}

Answer 4

Following is a code snippet which splits a sentense to word and give its count too.

 import java.util.HashMap;
 import java.util.Iterator;
 import java.util.Map;

 public class StringToword {
public static void main(String[] args) {
    String s="a a a A A";
    String[] splitedString=s.split(" ");
    Map m=new HashMap();
    int count=1;
    for(String s1 :splitedString){
         count=m.containsKey(s1)?count+1:1;
          m.put(s1, count);
        }
    Iterator<StringToword> itr=m.entrySet().iterator();
    while(itr.hasNext()){
        System.out.println(itr.next());         
    }
    }

}

Answer 5

Now, this can be accomplished just with split as it takes regex:

String s = "This is a sample sentence with []s.";
String[] words = s.split("\\W+");

this will give words as: {"this","is","a","sample","sentence", "s"}

The \\W+ will match all non-alphabetic characters occurring one or more times. So there is no need to replace. You can check other patterns also.

Answer 6

string.replaceAll() doesn't correctly work with locale different from predefined. At least in jdk7u10.

This example creates a word dictionary from textfile with windows cyrillic charset CP1251

    public static void main (String[] args) {
    String fileName = "Tolstoy_VoinaMir.txt";
    try {
        List<String> lines = Files.readAllLines(Paths.get(fileName),
                                                Charset.forName("CP1251"));
        Set<String> words = new TreeSet<>();
        for (String s: lines ) {
            for (String w : s.split("\\s+")) {
                w = w.replaceAll("\\p{Punct}","");
                words.add(w);
            }
        }
        for (String w: words) {
            System.out.println(w);
        }
    } catch (Exception e) {
        e.printStackTrace();
    }