On Wed, Sep 29, 2010 at 12:30 PM, Scott Kurz <scottk...@gmail.com> wrote: > Simon, like you mentioned, the original thinking must simply be that > in most cases, the right thing to do is transform the contents of one > array into the contents of the resulting array. > > I just wanted to chime in and ask if it might be worth trying to use > the transformer weight mechanism to make a direct array->JSON > conversion a higher-priority transform than a transformer chain which > includes Array2Array. > I admit I'm a little fuzzy as to precisely how to do that. Also, I > think this would have to make sense for array->JSON conversions, in > general, not just from this particular binding... not sure if this is > the case or not. > > Still, I wonder if this avenue is worth a try? > > Scott >
Certainly worth thinking about Scott. I am a little concerned though that I'm missed something obvious in the array->JSON conversion. I'm just trying to avoid having JSON conversion code in the binding if I can avoid it and it works OK for JSONP. So I can certainly try to see if it works for others like JSONRPC. What this does point out is that we don't have a good cross binding test for all the various types that that can be transferred as this issue was only just pointed out to me. We need to look at extending the reach of itest/databinding in 2.x. Simon -- Apache Tuscany committer: tuscany.apache.org Co-author of a book about Tuscany and SCA: tuscanyinaction.com
