Hi The characters should be converted to UTF-8 encoding for representation in the stream. If you create a Reader on-top of the stream, you would have to declare using UTF-8 do properly decode the bytes into characters.
Hope this helps. Regards Felix Am 01.04.2013 um 15:44 schrieb Dimitar Dimitrov: > Hello everyone. > Is there anyone who has faced the problem with the Cyrillic support in > Apache Sling. > When I put some string as a property value by the JCR Explorer, and then > get this node back in SystemView or DocumentView format I received > unreadable characters. > By debugging I saw that when I use > > nextNode.getSession().exportSystemView(nextNode.getPath(), > nodeByteArrayOutputStream, false, noRecursive); > > nodeByteArrayOutputStream which is implementation of > ByteArrayOutputStream, has filled with one byte characters for the Latin > symbols and four bytes characters for the Cyrillic. > At first I had not noticed that, so I tried some native methods to > convert the whole stream into the correct format, but unsuccessfully. > Then I start to look more precisely into the ByteArray and catch that > the cyrillic strings has four bytes. > The same can be seen in the JCR Explorer also if you try to enter the > cyrillic characters as a property value. > > Thanks in advance. > Dimitar > > -- Felix Meschberger | Principal Scientist | Adobe
