That was actually my original impl -- but it broke a number of weird test cases (I didn't dive into why), and in general I wanted to be as conservative as possible with this change. Non-lossy conversions work fine already; it's just the lossy ones that break, and AFAIK nobody's actually relying on the lossy conversion anywhere.
On Wed, Mar 17, 2010 at 3:39 PM, <[email protected]> wrote: > > http://codereview.appspot.com/620041/diff/3001/4002 > File > > java/common/src/main/java/org/apache/shindig/common/util/JsonConversionUtil.java > (right): > > http://codereview.appspot.com/620041/diff/3001/4002#newcode201 > > java/common/src/main/java/org/apache/shindig/common/util/JsonConversionUtil.java:201: > return value; > Can you skip the check and just always return value for length == 1? > > > http://codereview.appspot.com/620041/show >
