From: chap at anastigmatix.net (Chapman Flack) Date: Sat, 19 Sep 2015 23:41:04 -0400 Subject: [Pljava-dev] PL/java kills unicode chars? In-Reply-To: References: Message-ID: <55FE2AD0.1040305@anastigmatix.net> Srivatsan Ramanujam wrote: > I believe PL/java is killing unicode characters (it is probably converting > text to a byte stream and reading them as single byte characters - perhaps > Latin-1 and not as UTF-8). ... Well, this one has been open for a while. It's also in the github tracker, https://github.com/tada/pljava/issues/21 which I've just updated with confirming test code. I think vatsan's comment (on the github issue) about http://bugs.sun.com/view_bug.do?bug_id=5030776 is probably spot on. We seem to handle the whole basic multilingual plane correctly (nearly 64k codepoints), it's just all the planes above that getting messed up. Should be a straightforward fix, once I've found how many places those not-really-UTF JNI functions are really used in the code. -Chap