On Nov 3, 2007, at 10:03 PM, Charlie Groves wrote: > Hi all, > > I've been doing some work that involves parsing Hessian object > definitions, and have found a small difference between the example at > http://hessian.caucho.com/doc/hessian-serialization.html#anchor29 and > the format used by Hessian2Output and Hessian2Input in > hessian-3.1.3.jar. The spec says an object definition starts with > 'O' which is followed by a type definition starting with 't'. > However, both Hessian2Output and Hessian2Input move on to the two > bytes for String length following the 'O' and skip the 't'. I can > see the 't' isn't necessary as in the map and list case since the > type definition is mandatory for an object, but it'd be nice if the > spec were updated to reflect that. I'd actually prefer it if object > definitions were updated to be just an 'O' followed by a regular > Hessian string. That would shave off a byte in the most common case > where a type name is less than 32 characters and keep type > definitions confined to a single special case.
I've added it as http://bugs.caucho.com/view.php?id=2142. It's a good point. -- Scott > > Thanks, > Charlie > > > _______________________________________________ > hessian-interest mailing list > [email protected] > http://maillist.caucho.com/mailman/listinfo/hessian-interest _______________________________________________ hessian-interest mailing list [email protected] http://maillist.caucho.com/mailman/listinfo/hessian-interest
