Paul,

The issue may be that Excel expects to find only Western characters when CSV is used.
Same would apply to the text editors in Windows (unless you change the default non-Unicode
code page in the Advanced section of the Regional and Language options).

I think your best bet for Excel would be to use Excel XML with the explicitly specified UTF8 encoding.

Max


Paul Mander wrote:
Hi,

I would have thought the same, but changing 2 impacts the "degree of mess".
I have used various text editors and excel (as it is a cvs) to open the file
and they all show corruption. I'll pull together a test case to illustrate.


max.starets wrote:
  
Paul,

I would think that the only relevant encoding for the downloaded file is
the encoding you use while writing it (your item #3). What is the target 
format?
What application are you using to read the file?

Max Starets

Paul Mander wrote:
    
Another day, another encoding problem. This time I'm having problem
getting
Turkish characters to output correctly using the
fileDownloadActionListener
component.

I have 3 possible places I can specify the encoding. 

1) the contentType of on the component itself

contentType="text/csv;charset=iso-8859-9"

2) the xml page encoding of the markup

<?xml version="1.0" encoding="iso-8859-9"?>

3) The code that I use to generate the file

OutputStreamWriter w = new OutputStreamWriter(out, "iso-8859-9");

None of these or any combinations of this encoding and utf-8 results in a
file that has the correct encoding for a number of Turkish character.

In particular, the problem characters appear to be:

İı

  
      

    

  

Reply via email to