Re: form encoding UTF-8 / ISO-8859-1

[Murad Jura]

Artur Bialecki wrote:

I looked at the javadoc for java.lang package (jdk1.3.1)
and java.nio.charset.Charset (jdk1.4.1) and it
looks like the UTF-8 is the correct name.

I have a feeling setting the container encoding to ISO-8859-1
is what fixed your problem.

Another word of caution is the java.net.URLEncoder class that
doesn't play nice with UTF-8 in pre 1.4 jdk version.

Artur...

-----Original Message-----
From: Murad Jura [mailto:[EMAIL PROTECTED]] Sent: January 16, 2003 6:55 AM
To: [EMAIL PROTECTED]
Subject: Re: form encoding UTF-8 / ISO-8859-1


Martin Koeppe wrote:

Hello,

I'm using a recent developer version of cocoon 2.1 (Jan 9 2003),
tomcat 4.1.12 on SuSE Linux 8.0 Kernel 2.4.20,
and I found a problem with parsing the request parameters from a GET
request:

The browser (both Mozilla and IE) sends them as UTF-8, but cocoon
interprets them as ISO-8859-1. The problem can be seen e.g.

with form

validator: if you enter a non-ASCII7 character, and there

is something not

correct for the form validator within ANOTHER field, you

get the first

field back changed.

I have enabled "container-encoding" and "form-encoding" in

web.xml, and

set both to "utf-8", but with no success (no change in behaviour).

1) Can someone help?


I looked at the source, too:
org.apache.cocoon.environment.http.HttpRequest

There are getCharacterEncoding() and setCharacterEncoding().
Whereas get...() is mapped to the servlet engine function,
set...() seems to be implemented within cocoon.

In my experiments a call to set...("UTF-8") directly before

reading the

parameters within a flow script didn't work, either. (no change)
get...() always returned null.

2) Can someone explain the idea of the set...() function,

in combination

with decode(), i.e. why that should work? Or: what should

be done to get

it work?

3) Why doesn't get...() return the value that was set before with
set...(), i.e. why is the servlet value returned, whereas the cocoon
internal value seems to be used?


Thanks in advance

Martin Koeppe

---------------------------------------------------------------------

Please check that your question has not already been

answered in the

FAQ before posting.

<http://xml.apache.org/cocoon/faq/in> dex.html>

To

unsubscribe, e-mail: <[EMAIL PROTECTED]>

For additional commands, e-mail:

<[EMAIL PROTECTED]>

Hello Martin,

I want to inform you that I solved such problem by setting the web.xml's "container-encoding" parameter to ISO-8859-1 and the "form-encoding" parameter to "UTF8" (not "UTF-8" because "UTF8" is the internal Java's name of this encoding type, see Java Documentation).


---------------------------------------------------------------------
Please check that your question has not already been answered in the
FAQ before posting. <http://xml.apache.org/cocoon/faq/index.html>

To unsubscribe, e-mail: <[EMAIL PROTECTED]>
For additional commands, e-mail: <[EMAIL PROTECTED]>

---------------------------------------------------------------------
Please check that your question  has not already been answered in the
FAQ before posting.     <http://xml.apache.org/cocoon/faq/index.html>

To unsubscribe, e-mail:     <[EMAIL PROTECTED]>
For additional commands, e-mail:   <[EMAIL PROTECTED]>

Thank you Artur!

I have tested both UTF-8 and UTF8 and in all cases I have obtained successfull results. But I have to set "form-encoding" parameter to "UTF-8/UTF8" to get my russian data in the correct form.

Murad Jura.


---------------------------------------------------------------------
Please check that your question has not already been answered in the
FAQ before posting. <http://xml.apache.org/cocoon/faq/index.html>

To unsubscribe, e-mail: <[EMAIL PROTECTED]>
For additional commands, e-mail: <[EMAIL PROTECTED]>