Dear Hans-Juergen, it's true, JSON names and values may contain characters that have no valid equivalent in XML 1.0 (such as the backspace character [1]) or cannot be represented in QNames. As you already wrote, in the W3C there is an ongoing discussion on how to best represent JSON in XQuery. I haven't followed the last discussions, but as far as I know, there is no decision if JSON names should be represented as element names; instead, they could as well be stored in a (yet to be standardized) map structure, similar to Michael Kay's proposal which we implemented in BaseX, in which case they would have to be mapped differently than QNames. I hope this is a helpful hint? Christian [1] http://www.json.org/ [2] http://docs.basex.org/wiki/Map_Module ___________________________ On Fri, Apr 20, 2012 at 6:36 PM, Hans-Juergen Rennau <hrennau@yahoo.de> wrote:

Thank you, Christian! And one more question.

You use a bidirectional mapping between JSON names and XML names. So does MarkLogic (MLJSON), and I think others too. Is there any W3C work in progress which aims at a standardization of this JSON/XML name mapping? I think it might deserve a miniature spec.

Kind regards, Hans-Juergen

Von: Christian Grün <christian.gruen@gmail.com> An: Hans-Juergen Rennau <hrennau@yahoo.de> CC: "basex-talk@mailman.uni-konstanz.de" <basex-talk@mailman.uni-konstanz.de> Gesendet: 0:07 Donnerstag, 19.April 2012 Betreff: Re: [basex-talk] JSON/XML mapping

Dear Hans-Juergen,

...

1) The text says: "As arrays have no names, <value/> is used as element name." But I suppose it should be "array members", rather than "arrays", as an array may of course be associated with a name as any other JSON item. Correct?

thanks for the hint; I have updated our documentation.

...

2) So are there all in all 6 reserved attributes names (type, numbers, booleans, nulls, objects, arrays) and 2 reserved element names (json, value)? Or have I forgotten some names?

Exactly, these should be all of them.

Feel free to ask for more, Christian