Lost support for international character sets.
Description
In 1.0.* of GeoServer users could set the charSet, which would pass it to the postgis database and set the xml mime type. It doesn't seem to be working any more in 1.1beta, needs to be investigated a bit more, but the first test on my machine did not work.
Environment
Activity
CodeHaus Comment From: cholmes - Time: Fri, 26 Dec 2003 11:44:43 -0600
---------------------
<p>Moved to 1.1.1, though might have to wait till 1.2. The problem is the current configuration loading does not make doing this easy, as you can't access the ServerConfig when loading the datastores, which is where you'd get the character set. Individual setting of the postgis charset should work fine, so we should have a note of that. It's just the convenience global one.</p>
CodeHaus Comment From: cholmes - Time: Thu, 29 Jan 2004 18:01:49 -0600
---------------------
<p>This should be tested with weird characters (oslash is the one I use to test) in the actual database. To fix it the postgis datastore factory should be called with the global GeoServer charset as its charset param. And yes, just generally test international support extensively.</p>
CodeHaus Comment From: brihaye - Time: Fri, 30 Jan 2004 03:01:02 -0600
---------------------
<p>See : <a href="http://jira.codehaus.org/secure/ViewIssue.jspa?key=GEOS-64" class="external-link" rel="nofollow">http://jira.codehaus.org/secure/ViewIssue.jspa?key=GEOS-64</a></p>
<p>Using standard ways for parsing should definitely help in handling character encodings.</p>
<p>XML -> Java Unicode Strings -> XML</p>
CodeHaus Comment From: cholmes - Time: Fri, 30 Jan 2004 12:24:07 -0600
---------------------
<p>I think most of the XML reading and writing we do is fine, the bug I'm talking about is postgis and jdbc specific, so it's not reading XML, it's reading from a database, and getting the encoding wrong occasionally.</p>
CodeHaus Comment From: cholmes - Time: Wed, 7 Apr 2004 10:06:45 -0500
---------------------
<p>I'm unable to test this on windows properly. I'm going to close it until we get real confirmation that it's broken. I think code was added to get it working again, it just hasn't been fully tested.</p>
