can struts read hebrew from property files?

E

Elhanan

hi..

my jsp and property files are encoded in UTF-8:

if i use this:

<%@ taglib uri="/tags/struts-bean" prefix="bean" %>
<html>
<head>

</head>

<body>נסיון
<h3><bean:message key="welcome.heading"/></h3>
</body>
</html>

i get my hebrew in gibbirsh, but changing the encoding setting in
explorer corrects this: (using meta didn't help).
however if i add
<%@ page contentType="text/html; charset=UTF-8" pageEncoding="UTF-8"
%>

to the jsp , i get gibbrish value from the bean tag, no matter what i
do.
 
T

Thomas Fritsch

Elhanan said:
my jsp and property files are encoded in UTF-8:
if i use this:
<%@ taglib uri="/tags/struts-bean" prefix="bean" %>
<html>
<head>

</head>

<body>?????
<h3><bean:message key="welcome.heading"/></h3>
</body>
</html>
i get my hebrew in gibbirsh, but changing the encoding setting in
explorer corrects this: (using meta didn't help).
however if i add
<%@ page contentType="text/html; charset=UTF-8" pageEncoding="UTF-8"
%>
to the jsp , i get gibbrish value from the bean tag, no matter what i
do.


I would recommend to avoid the problematic non-ASCII characters altogether
(both in HTML and in properties files) and instead use the HTML or Java
escape syntax to write the hebrew characters as hexadecimal codes.
In HTML files like this:
<body>ןויםנ
In properties files (and in Java source code) like this:
key=value\u05df\u05d5\u05d9\u05dd\u05e0
This strategy makes your HTML and properties totally independent of
encoding, thus taking you to the safe side.

(You can get the hebrew unicode chart from <http://www.unicode.org/charts/>)
 
E

Elhanan

actually i've just read about a utility in java called native2ascii
which converts files to such charachters, i tried, and did a simple
java application of a property class reading such a file, and altough
the get property returned a readible string, i still got gibbrish in
jsp version.

remember that if i try my page with out the page directive it defaults
to western encoding, but i can still turn it manually, if i add the
page directive, it confuses it.
 
R

Roedy Green

actually i've just read about a utility in java called native2ascii

Let's look at your *.properties file first.

quoting from http://mindprod.com/jgloss/properties.html

In contrast to the natively generated system properties, ordinary user
Properties are usually loaded into RAM in their entirety and indexed
for rapid Hashtable access and are later saved to flat files on disk
with the *.properties extension. Properties files on disk are similar
to the old Windows 3.1 INI files, except they have no […] sections.
They are text files encoded with ISO 8859-1 keyword=value pairs. You
can get additional characters with \uxxxx escapes or with \t, \r, \n,
or \f. Comments begin with #. Lines may be continued by ending them in
a trailing \.

NOTE: you must use 8859-1 encoding for your *.properties file. I
don't know if you encode your strings left to right or right to left.

Dump your strings out on the console in hex so you can tell precisely
what you have in each slot.

/** convert a String to a hex representation of the String,
* with 4 hex chars per char of the original String.
* e.g. "1abc \uabcd" gives "0031_0061_0062_0063_0020_abcd"
* @param s String to convert to hex equivalent
*/
public static String displayHexString ( String s )
{
StringBuilder sb = new StringBuilder( s.length() * 5 - 1 );
for ( int i=0; i<s.length(); i++ )
{
char c = s.charAt(i);
if ( i != 0 )
{
sb.append( '_' );
}
// encode 16 bits as four nibbles

sb.append( hexChar [ c >>> 12 & 0xf ] );
sb.append( hexChar [ c >>> 8 & 0xf ] );
sb.append( hexChar [ c >>> 4 & 0xf ] );
sb.append( hexChar [ c & 0xf ] );
}
return sb.toString();
}

// table to convert a nibble to a hex char.
static final char[] hexChar = {
'0' , '1' , '2' , '3' ,
'4' , '5' , '6' , '7' ,
'8' , '9' , 'a' , 'b' ,
'c' , 'd' , 'e' , 'f'};
 
E

Elhanan

i think this program does what native2ascii program does, would like me
to send you the result from native2ascii?
 
R

Roedy Green

i think this program does what native2ascii program does, would like me
to send you the result from native2ascii?

Post your properties file. If all is well it will be expressed in
7-bit ASCII so your newsreader encoding won't matter.
 
E

Elhanan

here it is with the \u and all

# -- validator --
bean.title=\u05f3\u00a0\u05f3\u00a1\u05f3\u2122\u05f3\u2022\u05f3\ufffd

bean.body= \u05f3\u2019\u05f3\u2022\u05f3\u00a3
errors.required={0} \u05f3\u201c\u05f3\u00a8\u05f3\u2022\u05f3\u00a9
# -- welcome --
welcome.title=Struts Blank Application
welcome.heading=\u05e9\u05dc\u05d5\u05dd

the one i'm aiming for is 'welcome.heading'
 
R

Roedy Green

here it is with the \u and all

# -- validator --
bean.title=\u05f3\u00a0\u05f3\u00a1\u05f3\u2122\u05f3\u2022\u05f3\ufffd

bean.body= \u05f3\u2019\u05f3\u2022\u05f3\u00a3
errors.required={0} \u05f3\u201c\u05f3\u00a8\u05f3\u2022\u05f3\u00a9
# -- welcome --
welcome.title=Struts Blank Application
welcome.heading=\u05e9\u05dc\u05d5\u05dd

the one i'm aiming for is 'welcome.heading'

I wrote this little SSCCE to read your file. It seems to be reading it
correctly under Win2K with JDK 1.5.0_06.

import java.io.FileInputStream;
import java.io.IOException;
import java.util.Properties;

public class Hebrew
{
/**
* test harness
*
* @param args not used
*/
public static void main ( String[] args )
{

// get an ordinary property
Properties props = new Properties();
try
{
FileInputStream fis = new FileInputStream(
"hebrew.properties" );
props.load( fis );
fis.close();
}
catch ( IOException e )
{
return;
}
String desc = props.getProperty( "welcome.heading" );
System.out.println( desc );
System.out.println( displayHexString( desc ) );

}

/** convert a String to a hex representation of the String,
* with 4 hex chars per char of the original String.
* e.g. "1abc \uabcd" gives "0031_0061_0062_0063_0020_abcd"
* @param s String to convert to hex equivalent
*/
public static String displayHexString ( String s )
{
StringBuilder sb = new StringBuilder( s.length() * 5 - 1 );
for ( int i=0; i<s.length(); i++ )
{
char c = s.charAt(i);
if ( i != 0 )
{
sb.append( '_' );
}
// encode 16 bits as four nibbles

sb.append( hexChar [ c >>> 12 & 0xf ] );
sb.append( hexChar [ c >>> 8 & 0xf ] );
sb.append( hexChar [ c >>> 4 & 0xf ] );
sb.append( hexChar [ c & 0xf ] );
}
return sb.toString();
}

// table to convert a nibble to a hex char.
static final char[] hexChar = {
'0' , '1' , '2' , '3' ,
'4' , '5' , '6' , '7' ,
'8' , '9' , 'a' , 'b' ,
'c' , 'd' , 'e' , 'f'};
}
 
E

Elhanan

i know it reads it correctly i also did a small application only i
simply displayed the results from the properties class on the eclipse
console, and it was correct, in the struts it doesn't work.
 
R

Richard Wheeldon

Elhanan said:
my jsp and property files are encoded in UTF-8:

Properties files cannot be in UTF8. However, there is
a quick and dirty hack to allow this: Create a wrapper
which reads the String, gets the byte array of the
string, then re-parses the byte array as a string of
UTF8 characters,

Regards,

Richard
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,982
Messages
2,570,185
Members
46,736
Latest member
AdolphBig6

Latest Threads

Top