javaregexjava.util.scannerpunctuationaccent-insensitive

Java remove punctuation on a String (also ’ “ ” and all of these) maintaining accents characters


I need to remove punctuation reading on a file, maintaining accents character I tried this code but don't work how I would.

Expectation: input=> ’'qwe..,rty ‘èeéò’“ ”o" "à     output=> qwertyèeéòoà

Effective result: input=> ’'qwe..,rty ‘èeéò’“ ”o" "à   output=>’qwerty ‘èeéò’“ ”o" "à

I can't remove ’“” symbols and other of these

Note: Eclipse and filetext.txt are set to UTF-8.

Thank you

import java.io.*;
import java.util.Scanner;

public class DataCounterMain {
    public static void main (String[] args) throws FileNotFoundException {

    File file = new File("filetext.txt");

    try {
        Scanner filescanner = new Scanner(file);
        while (filescanner.hasNextLine()) {

            String line = filescanner.nextLine();
            line=line.replaceAll ("\\p{Punct}", "");

            System.out.println(line);
        }
    }
    catch(FileNotFoundException e) {
        System.err.println(file +" FileNotFound");
    }
    }
}

Solution

  • The regex \p{Punct} only matches US-ASCII punctuation by default, unless you enable Unicode character classes. That means that your code, as written, would only remove these characters:

    !"#$%&'()*+,-./:;<=>?@[\]^_`{|}~
    

    If you want to match everything the Unicode Consortium classified as punctuation, try \p{IsPunctuation} instead, which always checks Unicode character properties and matches all the punctiuation in your example (and more!).

    To replace whitespace as well as punctuation, like in your example, you would use:

                 
            line = line.replaceAll("\\p{IsPunctuation}|\\p{IsWhite_Space}", "");