Incorrect assumption about default character encoding

[GoogleCodeExporter]

There are a couple places in the OAuth client library for Java where a
String is converted to a byte[] or vis-versa without an explicit character
encoding.

On most platforms that works fine because the default character set is
mostly compatible with ASCII or UTF-8.  On some platforms, like z/OS on IBM
hardware the default character encoding is something less compatible (UBIDEC)

I've attached a patch that makes the needed changes to make the Oauth
library work on z/OS.

Original issue reported on code.google.com by matt.la...@gmail.com on 19 Apr 2010 at 5:03

Attachments:

• oauth-utf8.patch

[GoogleCodeExporter]

The character encoding on z/OS is acutally EBCDIC.  I got my characters out of 
order =).

Original comment by matt.la...@gmail.com on 19 Apr 2010 at 5:20

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Is there anything I can do to get this patch applied to the OAuth code?  

I'd like to ship an unmodified version of the OAuth libraries in our builds of 
Jazz
Foundation (from jazz.net), but this bug (not sure why it was marked an 
enhancement)
is causing problems on our z/OS platforms.

If there is something wrong with the patch, I can try to build a new one with 
any
feedback you can give.

Original comment by matt.la...@gmail.com on 14 May 2010 at 8:45

[GoogleCodeExporter]

fixed in -r1202

Original comment by jmkrist...@gmail.com on 24 May 2010 at 3:29

• Changed state: Fixed