Can't retrieve urlencoded umlauts. [solution: use UTF8]

Question

Hello, I'm programming some websites with JSP-Scripting and I encounterd a strange problem with urlencoded web-formular parameters. The site itself is encoded in iso-8859-1.

I have a simple webform with a field called description.

If I enter german Umlauts or specialchars like "ü" or "ß", these get automatically urlencoded. But if I want to read this parameter, I always get null.

String description = request.getParameter("description");

If I enter some chinese words, like 專業人士, they are urlencoded, too. However, I can read them without retrieving null.

This behaviour doesn't change whether I use "post" or "get" as method. I tried to "pimp" my webform with the text below - but that didn't help either.

accept-charset="ISO-8859-1" enctype="application/x-www-form-urlencoded"

My question is: Why can't I retrieve urlencoded german Umlauts but some chinese words?

Answer 1

I suspect your page or servlet encoding is UTF-8. The Latin-1 encoded Umlaut is an invalid UTF-8 sequence so you get NULL.

When you enter Chinese, the browser knows it can't encode this in Latin-1 so it changes into UTF-8 automatically. That's why Chinese works.

If you can post some HTTP trace, we can confirm this.

Answer 2

Chinese with ISO-8859-1 wont't work, but german should. If chineese is encoded anyway, maybe your form is encoded in UTF-8? Which browser so you use? What is the encopding of the JSP? You should always use UFT-8, not Latin 1. Nowadays every browser and server should support UTF-8 on every OS.

My experience is, that GET ist often not working with encoding, but POST should do. The form parameters accept-charset and enctype are correct. What server do you use?

The JSP content type is typically set by:

<%@ page language="java" contentType="text/html; charset=ISO-8859-1" 
                         pageEncoding="ISO-8859-1" %>