Java: Multibyte string length

Question

I have a method which prints "header text" for command line programs, much like the syntax of Markdown:

1. =======================
2. This is a header string
3. =======================

This method takes a char c for lines 1 and 3 and repeats it n times based on the length of s.

String.length() works fine with the English alphabet, but how can I find the length (the visual length, that is) of a string containing foreign multibyte characters like "Å" and "Ç"?

Perhaps you're not reading the string in the correct encoding — ReyCharles
– ReyCharles, Commented Oct 3, 2012 at 16:21

Ian Roberts · Accepted Answer · 2012-10-03 16:12:02Z

11

String.length will be fine for those sorts of characters, as Java strings work in UTF-16, which is sufficient to represent the vast majority of characters in common use (Latin, Greek, Arabic, Hebrew, Chinese, Thai, Devanagari, ...).

If you might need to deal with characters above U+FFFF then you need to use codePointCount instead of length to cope with surrogate pairs.

edited Oct 3, 2012 at 16:12

answered Oct 3, 2012 at 15:56

Ian Roberts

123k17 gold badges177 silver badges191 bronze badges

Sign up to request clarification or add additional context in comments.

7 Comments

josocblaugrana Over a year ago

s.length() gives me a value of 2 if s is "Ø". I want this to be 1. And I believe you might get a different value depending on the file encoding you compile the program with.

ReyCharles Over a year ago

@josocblaugrana: I cannot confirm that. I tried running the following program and it printed 1: public class A { public static void main(String[] args) { System.out.println("Ø".length()); } }

josocblaugrana Over a year ago

@ReyCharles: On my end, running System.out.println("Ø".length()) (and nothing else) returns a value of 2… For the record, I am using a Mac.

Ian Roberts Over a year ago

@josocblaugrana clearly the encoding you tell the compiler to use must match the encoding of the Java source files, in the same way that you have to pass the right encoding to an InputStreamReader when you're reading a text file. If your .java file is in UTF-8 but your compiler reads it as if it were ISO-8859-1 or MacRoman then it will see that string literal as "\u00c3\u0098" instead of "\u00D8".

Ian Roberts Over a year ago

If you can't be sure what encoding your compiler will be set to then you'll have to use Unicode escapes in the source code (i.e. say "\u00D8" instead of "Ø"). You can automate this using the native2ascii tool.

|

Community · Accepted Answer · 2020-06-20 09:12:55Z

3

String.length() is fine for most Unicode characters including Å and Ç.

A Java string is utf-16 encoded where each Character takes up 2 or 4 bytes.

Supplementary characters denotes the characters taking 4 bytes and is implemented by pairing two characters, in which case the codePointCount operation must be used instead of length.

Characters though most certainly exist in the standard unicode specification.

edited Jun 20, 2020 at 9:12

CommunityBot

11 silver badge

answered Oct 3, 2012 at 15:55

Johan Sjöberg

49.4k22 gold badges135 silver badges150 bronze badges

2 Comments

Rok Kralj Over a year ago

Lol no. In utf16 each character takes 2 or 4 bytes.

Johan Sjöberg Over a year ago

@RokKralj, which is exactly what I was trying to say. Good that you spotted that it wasn't clear enough

Collectives™ on Stack Overflow

Java: Multibyte string length

2 Answers 2

7 Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

7 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related