- ConfigProfile_p.html now transmits usable encoding for other than 7-bit ASCII charset, see TODO in httpd.parseArg(String)

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@3174 6c8d7289-2bf4-0310-a012-ef5d649a1542

htroot/ConfigProfile_p.html

@@ -14,7 +14,7 @@
     <p>
       You do not need to provide any personal data here, but if you want to distribute your contact information, you can do that here.
     </p>
-    <form action="ConfigProfile_p.html">
+    <form action="ConfigProfile_p.html" accept-charset="ascii">
       <fieldset>
         <dl>
           <dt><label for="name">Name</label></dt>

source/de/anomic/http/httpd.java

@@ -777,6 +777,19 @@ public final class httpd implements serverHandler {
             		baos.write(Integer.parseInt(s.substring(pos + 2, pos + 6), 16));
             		pos += 6;
             	} else {									// normal escape
+                    // currently one escape is treated as one char, which only works if
+                    // formulars accept-charset=ascii are used. This is wrong if
+                    // formulars are set to accept-charset=UTF-8, then many escaped bytes
+                    // are sent, but they belong together to one char
+                    
+                    // TODO: UTF-8 escapes: i.e. "%C3%A4" should be 'ä', but is now "ä"
+                    // - parse out subsequent escapes
+                    // - check which of these 'belong together', see http://de.wikipedia.org/wiki/UTF-8
+                    // - write the UTF-8 escapes together as one char
+                    
+                    // XXX: implement an own UTF-8 converter as described above or
+                    // revert this method to use bytes and let Java do the UTF-8-parsing?
+                    // i.e. new String(result[], "UTF-8");
                 	baos.write(Integer.parseInt(s.substring(pos + 1, pos + 3), 16));
             		pos += 3;
             	}