Java change system new-line character - java

On Windows, using System.out.println() prints out \n\r while on a Unix system you would get \n.
Is there any way to tell java what new-line characters you want to use?

As already stated by others, the system property line.separator contains the actual line separator. Strangely, the other answers missed the simple conclusion: you can override that separator by changing that system property at startup time.
E.g. if you run your program with the option -Dline.separator=X at the command line you will get the funny behavior of System.out.println(…); ending the line with an X.
The tricky part is how to specify characters like \n or \r at the command line. But that’s system/environment specific and not a Java question anymore.

Yes, there is a way and I've just tried it.
There is a system property line.separator. You can set it using System.setProperty("line.separator", whatever)
To be sure that it indeed causes JVM to use other separator I implemented the following exercise:
PrintWriter writer = new PrintWriter(new FileWriter("c:/temp/mytest.txt"));
writer.println("hello");
writer.println("world");
writer.close();
I am running on windows now, so the result was 14 bytes long file:
03/27/2014 10:13 AM 14 mytest.txt
1 File(s) 14 bytes
0 Dir(s) 409,157,980,160 bytes free
However when I added the following line to the beginning of my code:
System.setProperty("line.separator", "\n");
I got 14 bytes long file:
03/27/2014 10:13 AM 14 mytest.txt
1 File(s) 14 bytes
0 Dir(s) 409,157,980,160 bytes free
I opened this file with notepad that does not recognize single \n as a new line and saw one-line text helloworld instead of 2 separate lines. So, this works.

Because the accepted answer simply does not work, as others pointed out before me, and the JDK only initialises the value once and then never reads the property anymore, only an internal static field, it became clear that the clean way to change the property is to set it on the command line when starting the JVM. So far, so good.
The reason I am writing yet another answer is that I want to present a reflective way to change the field, which really works with streams and writers relying on System.lineSeparator(). It does not hurt to update the system property, too, but the field is more important.
I know that reflection is ugly, as of Java 16+ needs an extra JVM command line parameter in order to allow it, and only works as long as the internals of System do not change in OpenJDK. But FWIW, here is my solution - don't do this at home, kids:
import java.io.*;
import java.lang.reflect.Field;
import java.nio.file.Files;
/**
* 'assert' requires VM parameter '-ea' (enable assert)
* 'Field.setAccessible' on System requires '--add-opens java.base/java.lang=ALL-UNNAMED' on Java 16+
*/
public class ChangeLineSeparator {
public static void main(String[] args) throws IOException, NoSuchFieldException, IllegalAccessException {
assert System.lineSeparator().equals("\r\n") : "default Windows line separator should be CRLF";
Field lineSeparator = System.class.getDeclaredField("lineSeparator");
lineSeparator.setAccessible(true);
lineSeparator.set(null, "\n");
assert System.lineSeparator().equals("\n") : "modified separator should be LF";
File tempFile = Files.createTempFile(null, null).toFile();
tempFile.deleteOnExit();
try (PrintWriter out = new PrintWriter(new FileWriter(tempFile))) {
out.println("foo");
out.println("bar");
}
assert tempFile.length() == "foo\nbar\n".length() : "unexpected file size";
}
}

You may try with:
String str = "\n\r";
System.out.print("yourString"+str);
but you can instead use this:-
System.getProperty("line.separator");
to get the line seperator
Returns the system-dependent line separator string. It always returns
the same value - the initial value of the system property
line.separator.
On UNIX systems, it returns "\n"; on Microsoft Windows systems it
returns "\r\n".

As stated in the Java SE tutorial:
To modify the existing set of system properties, use
System.setProperties. This method takes a Properties object that has
been initialized to contain the properties to be set. This method
replaces the entire set of system properties with the new set
represented by the Properties object.
Warning: Changing system
properties is potentially dangerous and should be done with
discretion. Many system properties are not reread after start-up and
are there for informational purposes. Changing some properties may
have unexpected side-effects.
In the case of System.out.println(), the line separator that existed on system startup will be used. This is probably because System.lineSeparator() is used to terminate the line. From the documentation:
Returns the system-dependent line separator string. It always returns
the same value - the initial value of the system property
line.separator.
On UNIX systems, it returns "\n"; on Microsoft Windows systems it
returns "\r\n".
As Holger pointed out, you need to overwrite this property at startup of the JVM.

Windows cmd: Credit jeb at https://superuser.com/a/1519790 for a technique to specify a line-feed character in a parameter using a cmd variable. This technique can be used to specify the java line.separator.
Here's a sample javalf.cmd file
#echo off
REM define variable %\n% to be the linefeed character
(set \n=^^^
^
)
REM Start java using the value of %\n% as the line.separator System property
java -Dline.separator=%\n% %1 %2 %3 %4 %5 %6 %7 %8 %9
Here's a short test progam.
public class ReadLineSeparator {
public static void main(String... ignore) {
System.out.println(System.lineSeparator().chars()
.mapToObj(c -> "{"+Character.getName(c)+"}")
.collect(java.util.stream.Collectors.joining()));
}
}
On Windows,
java ReadLineSeparator produces
{CARRIAGE RETURN (CR)}{LINE FEED (LF)}
.\javalf.cmd ReadLineSeparator produces
{LINE FEED (LF)}

The method System.lineSeparator() returns the line separator used by the system. From the documentation it specifies that it uses the system property line.separator.

Checked Java 11 implementation:
private static void initPhase1() {
lineSeparator = props.getProperty("line.separator");
}
public static String lineSeparator() {
return lineSeparator;
}
So altering system property at runtime doesn't change System.lineSeparator().
For this reason some projects re-read system property directly, see my answer: How to avoid CRLF (Carriage Return and Line Feed) in Logback - CWE 117
The only viable option is to set system property during app startup.
For Bash it is as simple as: java -Dline.separator=$'\n' -jar my.jar.
For POSIX shell it is better to save that character in some variable first:
LF='
'
java -Dline.separator="$LF" -jar my.jar
If you are not sure debug it:
printf %s "$LF" | wc -c
1
printf %s "$LF" | od -x
0000000 000a
For Gradle I use:
tasks.withType(Test).configureEach {
systemProperty 'line.separator', '\n'
}
bootRun {
systemProperty 'line.separator', '\n'
}

Related

How to get charset of Process in Java?

Charset.defaultCharset() and file.encoding is JVM's charset, not charset of OS, console and terminal.
Now I create a Process to run a program, and use process.getInputSteam() to read the output, how to find the correct charset for the process (sun.jnu.encoding is right but it seems not general)?
From Java 17 onwards, there is a method in Process class named inputStream.
The source code taken from Process.java:
public final BufferedReader inputReader() {
return inputReader(CharsetHolder.nativeCharset());
}
The nativeCharset is obtained from System property "native.encoding".
This property was introduced in Java 17. This is probably what you want.
Reference: https://bugs.openjdk.java.net/browse/JDK-8266075

System.lineSeparator() returns nothing

What should I see when I use the following?
System.out.println("LineSeperator1: "+System.getProperty("line.separator"));
System.out.println("LineSeperator2: "+System.lineSeparator());
I get the following back:
LineSeperator1:
LineSeperator2:
Is it empty? invisible? shouldn there be something like \r or \n?
I use windows 7, eclipse and jdk 1.8.
As you expect, the line separator contains special characters such as '\r' or '\n' that denote the end of a line. However, although they would be written in Java source code with those backslash escape sequences, they do not appear that way when printed. The purpose of a line separator is to separate lines, so, for example, the output of
System.out.println("Hello"+System.lineSeparator()+"World");
is
Hello
World
rather than, say
Hello\nWorld
You can even see this in the output of your code: the output of
System.out.println("LineSeperator1: "+System.getProperty("line.separator"));
had an extra blank line before the output of the next statement, because there was a line separator from System.getProperty("line.separator") and another from the use of println.
If you really want to see what the escaped versions of the line separators look like, you can use escapeJava from Apache Commons. For example:
import org.apache.commons.lang3.StringEscapeUtils;
public class LineSeparators {
public static void main(String[] args) {
String ls1 = StringEscapeUtils.escapeJava(System.getProperty("line.separator"));
System.out.println("LineSeperator1: "+ls1);
String ls2 = StringEscapeUtils.escapeJava(System.lineSeparator());
System.out.println("LineSeperator2: "+ls2);
}
}
On my system, this outputs
LineSeparator1: \n
LineSeparator2: \n
Note that I had to run it in the same folder as the .jar file from the Apache download, compiling and running with these commands
javac -cp commons-lang3-3.4.jar LineSeparators.java
java -cp commons-lang3-3.4.jar:. LineSeparators
Printing something afterwards will show the effect:
System.out.println("a" + System.lineSeparator() + "b");
Gives
a
b

Java compilaton issue with ß character

I have come across a strange issue. In the below piece of code, I am searching for the presence of ß.
public static void main(String[] args) {
char [] chArray = {'ß'};
String str = "Testß";
for(int i=0; i<chArray.length; i++){
if(str.indexOf(chArray[i])>-1){
System.out.println("ß is present");
break;
}
}
}
I have a web application running on JBOSS in linux, Java 6. The above code doesn't detect the presence of ß when include the code in the above specified application.
Surprisingly, if I compile the same file in my eclipse workspace and then apply the patch in the application it runs as expected!
Points to note:
The application build environment is a black-box to me, hence no idea if there is any -encoding option is present for the javac command or something like that
My eclipse's JRE is java8, but the compiler version set for the project is Java6
I changed the value from ß to unicode equivalent to \u00DF in the array declaration, but still the behavior is same.
char [] chArray = {'\u00DF'};
When I decompiled the class file generated the character array declared value was shown as 65533, which is \uFFFD, nothing but replacement character which is used for unidentified symbol. I used JD-GUI as decompiler, which I dont see trustworthy!
Need your help folks! I am sure it is not same as: case sensitive issue of beta Java's equalsIgnoreCase fails with ß ("Sharp S" used in German alphabet)
Thanks in advance
I think your problem is the encoding of ß. You have two options to solve your error:
First convert your java source code into ascii chars, and then compile it:
native2ascii "your_class_file.java"
javac "your_class_file.java"
Compile your java file with your encoding, utf-8 on linux and iso-8859-15 on windows:
javac -encoding "encoding" "your_class_file.java"
As far as I can judge it, it should have worked with replacing "ß" with "\u00df". If the solutions above don't work, print every char and its unicode value to System.out and look which char is 'ß'.
Another error might be that you read the text in an encoding that doesn't support ß; try reading your String by reading the bytes and call:
String input = new String(input_bytes, StandartCharsets.UTF_8); // on linux
String input = new String(input_bytes, StandartCharsets.ISO_8859_1); // on windows
For more information on charsets, see StandartCharsets class reference.
Thanks for your time and responses!
The actual problem was the class file was not generated in the build, hence the change was not reflecting. Using ß's unicode value \u00DF in the java source file should work fine.

Portable newline transformation in Java

Suppose I have a string, which consists of a few lines:
aaa\nbbb\nccc\n (in Linux) or aaa\r\nbbb\r\nccc (in Windows)
I need to add character # to every line in the string as follows:
#aaa\n#bbb\n#ccc (in Linux) or #aaa\r\n#bbb\r\n#ccc (in Windows)
What is the easiest and portable (between Linux and Windows) way to do it Java ?
Use the line.separator system property
String separator = System.getProperty("line.separator") + "#"; // concatenate the character you want
String myPortableString = "#aaa" + separator + "ccc";
These properties are described in more detail here.
If you open the source code for PrintWriter, you'll notice the following constructor:
public PrintWriter(Writer out,
boolean autoFlush) {
super(out);
this.out = out;
this.autoFlush = autoFlush;
lineSeparator = java.security.AccessController.doPrivileged(
new sun.security.action.GetPropertyAction("line.separator"));
}
It's getting (and using) the system specific separator to write to an OutputStream.
You can always set it at the property level
System.out.println("ahaha: " + System.getProperty("line.separator"));
System.setProperty("line.separator", System.getProperty("line.separator") + "#"); // change it
System.out.println("ahahahah:" + System.getProperty("line.separator"));
prints
ahaha:
ahahahah:
#
All classes that request that property will now get {line.separator}#
I don't know what exactly you are using, but PrintWriter's printf methods lets you write formatted strings. Withing the string you can use the %n format specifier which will output the platform specific line separator.
System.out.printf("first line%nsecond line");
The output:
first line
second line
(System.out is a PrintStream which also supports this).
As of Java 7 (also noted here) you can also use the method:
System.lineSeparator()
which is the same as:
System.getProperty("line.separator")
to get the system-dependent line separator string.
The description of the method from the official JavaDocs is quoted below:
Returns the system-dependent line separator string. It always returns
the same value - the initial value of the system property
line.separator.
On UNIX systems, it returns "\n"; on Microsoft Windows systems it
returns "\r\n".

How to compile a java source file which is encoded as "UTF-8"?

I saved my Java source file specifying it's encoding type as UTF-8 (using Notepad, by default Notepad's encoding type is ANSI) and then I tried to compile it using:
javac -encoding "UTF-8" One.java
but it gave an error message"
One.java:1: illegal character: \65279
?public class One {
^
1 error
Is there any other way, I can compile this?
Here is the source:
public class One {
public static void main( String[] args ){
System.out.println("HI");
}
}
Your file is being read as UTF-8, otherwise a character with value "65279" could never appear. javac expects your source code to be in the platform default encoding, according to the javac documentation:
If -encoding is not specified, the platform default converter is used.
Decimal 65279 is hex FEFF, which is the Unicode Byte Order Mark (BOM). It's unnecessary in UTF-8, because UTF-8 is always encoded as an octet stream and doesn't have endianness issues.
Notepad likes to stick in BOMs even when they're not necessary, but some programs don't like finding them. As others have pointed out, Notepad is not a very good text editor. Switching to a different text editor will almost certainly solve your problem.
Open the file in Notepad++ and select Encoding -> Convert to UTF-8 without BOM.
This isn't a problem with your text editor, it's a problem with javac !
The Unicode spec says BOM is optionnal in UTF-8, it doesn't say it's forbidden !
If a BOM can be there, then javac HAS to handle it, but it doesn't. Actually, using the BOM in UTF-8 files IS useful to distinguish an ANSI-coded file from an Unicode-coded file.
The proposed solution of removing the BOM is only a workaround and not the proper solution.
This bug report indicates that this "problem" will never be fixed : https://web.archive.org/web/20160506002035/http://bugs.java.com/view_bug.do?bug_id=4508058
Since this thread is in the top 2 google results for the "javac BOM" search, I'm leaving this here for future readers.
Try javac -encoding UTF8 One.java
Without the quotes and it's UTF8, no dash.
See this forum thread for more links
See Below
For example we can discuss with an Program (Telugu words)
Program (UnicodeEx.java)
class UnicodeEx {
public static void main(String[] args) {
double ఎత్తు = 10;
double వెడల్పు = 25;
double దీర్ఘ_చతురస్ర_వైశాల్యం;
System.out.println("The Value of Height = "+ఎత్తు+" and Width = "+వెడల్పు+"\n");
దీర్ఘ_చతురస్ర_వైశాల్యం = ఎత్తు * వెడల్పు;
System.out.println("Area of Rectangle = "+దీర్ఘ_చతురస్ర_వైశాల్యం);
}
}
This is the Program while saving as "UnicodeEx.java" and change Encoding to "unicode"
**How to Compile**
javac -encoding "unicode" UnicodeEx.java
How to Execute
java UnicodeEx
The Value of Height = 10.0 and Width = 25.0
Area of Rectangle = 250.0
I know this is a very old thread, but I was experiencing a similar problem with PHP instead of Java and Google took me here. I was writing PHP on Notepad++ (not plain Notepad) and noticed that an extra white line appeared every time I called an include file. Firebug showed that there was a 65279 character in those extra lines.
Actually both the main PHP file and the included files were encoded in UTF-8. However, Notepad++ has also an option to encode as "UTF-8 without BOM". This solved my problem.
Bottom line: UTF-8 encoding inserts here and there this extra BOM character unless you instruct your editor to use UTF8 without BOM.
Works fine here, even edited in Notepad. Moral of the story is, don't use Notepad. There's likely a unprintable character in there that Notepad is either inserting or happily hiding from you.
I had the same problem. To solve it opened the file in a hex editor and found three "invisible" bytes at the beginning of the file. I removed them, and compilation worked.
Open your file with WordPad or any other editor except Notepad.
Select Save As type as Text Document - MS-DOS Format
Reopen the Project
To extend the existing answers with a solution for Linux users:
To remove the BOM on all .java files at once, go to your source directory and execute
find -iregex '.*\.java' -type f -print0 | xargs -0 dos2unix
Requires find, xargs and dos2unix to be installed, which should be included in most distributions. The first statement finds all .java files in the current directory recursively, the second one converts each of them with the dos2unix tool, which is intended to convert line endings but also removes the BOM.
The line endings conversion should have no effect as it should already be in Linux \n format on Linux if you configure your version control correctly but be warned that it does that as well in case you have one of those rare cases where that is not intended.
In the Intellij Idea(Settings>Editor>File Encodings), the project encoding was "windows-1256". So I used the following code to convert static strings to utf8
protected String persianString(String persianStirng) throws UnsupportedEncodingException {
return new String(persianStirng.getBytes("windows-1256"), "UTF-8");
}
Now It is OK!
Depending on the file encoding you should change "windows-1256" to a proper one

Categories