On Fri, Jan 22, 2010 at 1:03 PM, <lind...@inuus.com> wrote: > > http://codereview.appspot.com/186278/diff/1/2 > File > > java/gadgets/src/main/java/org/apache/shindig/gadgets/servlet/MakeRequestHandler.java > (right): > > http://codereview.appspot.com/186278/diff/1/2#newcode119 > > java/gadgets/src/main/java/org/apache/shindig/gadgets/servlet/MakeRequestHandler.java:119: > encoding = "UTF-8"; > On 2010/01/22 18:16:57, beaton wrote: > >> Why would we ever receive a request that was not UTF-8 encoded? >> > > From the HttpServletRequest javadoc: > > Returns the name of the character encoding used in the body of this > request. This method returns null if the request does not specify a > character encoding > > > http://codereview.appspot.com/186278/diff/1/2#newcode129 > > java/gadgets/src/main/java/org/apache/shindig/gadgets/servlet/MakeRequestHandler.java:129: > req.setPostBody(getParameter(request, POST_DATA_PARAM, > "").getBytes(encoding)); > On 2010/01/22 18:16:57, beaton wrote: > >> paul lindner ran into nasty performance problems with code like this: >> > > > > http://paul.vox.com/library/post/the-mysteries-of-java-character-set-performance.html > > If the other character set is only used sporadically, then yes, you'll > be okay. Since most sites will tend to stick with one or the other > character set this should be okay. Id recommend passing the charset > value through a toUpperCase to reduce variance though...
Good suggestion, and "other" charsets *should *only be used sporadically. Still, they might not, which leads me to ask the inevitable: what are the hurdles in Shindig upgrading to Java 6? And I quote Mr Lindner: "If you're using Java 6 you can use pre-calculated static Charset objects which gets you more concurrency. This works for getBytes() and some others. Of course this doesn't help with external class libraries... I added this to our local copy of the postgresql jdbc driver and it helped out a bit." ...among likely other benefits. > > > http://codereview.appspot.com/186278/show >
