Isidre, The client/server version of the BaseX Driver will do exactly what you want Isidre, in exactly the same way the Sedna, eXist and MarkLogic drivers do. It will not be long until this driver will be available. Regards, Charles

Hi Isidre,

as our XQJ implementation will soon be replaced by a new version of Charles Foster (who also did the versions for eXist, Sedna and MarkLogic), we won't put any more effort into the existing code. The new version will also be based on the client/server architecture of BaseX.

Hope this helps, Christian ___________________________

2012/4/3 Isidre Guixà <iguixa@xtec.cat>:

...

Hi

I am experiencing the BaseX API XQJ and I compare to other XQJ NXD and I find the following problem.

Suppose we have the next XML file in database "test":

<?xml version="1.0" encoding="ISO-8859-1"?>   <bookstore param="hao">     <book>       <title>Book01</title>     </book>     <book>       <title>Book02</title>     </book> </bookstore>

If I execute:

    XQDataSource xqs = ...     XQConnection conn = xqs.getConnection();     XQExpression xqe = conn.createExpression();     String cad= "collection('test')//title";

    XQResultSequence xqrs = xqe.executeQuery(cad);     XMLStreamReader result = xqrs.getSequenceAsStream();     System.out.println("Iteration on XMLStreamReader...");     for(; result.hasNext(); result.next())       System.out.println(getEventString(result));

    xqrs = xqe.executeQuery(cad);     System.out.println("Using a Transformer...");     Source resultXML = new StAXSource(xqrs.getSequenceAsStream());     StreamResult resultSR = new StreamResult(System.out);     TransformerFactory factory = TransformerFactory.newInstance();     Transformer tr = factory.newTransformer();     tr.setOutputProperty(OutputKeys.INDENT, "yes");     tr.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "no");     tr.setOutputProperty("{http://xml.apache.org/xslt}indent-amount","2");     tr.transform(resultXML, resultSR);

where   static String getEventString(final XMLStreamReader reader) {     switch(reader.getEventType()) {       case XMLStreamConstants.START_ELEMENT:         return "START_ELEMENT:\t\"" + reader.getLocalName() + "\"";       case XMLStreamConstants.END_ELEMENT:         return "END_ELEMENT:\t\"" + reader.getLocalName() + "\"";       case XMLStreamConstants.START_DOCUMENT:         return "START_DOCUMENT";       case XMLStreamConstants.END_DOCUMENT:         return "END_DOCUMENT";       case XMLStreamConstants.CHARACTERS:         return "CHARACTERS:\t\"" + reader.getText() + "\"";     }     return "";   }

I get:

Iteration on XMLStreamReader... START_ELEMENT: "title" CHARACTERS: "Book01" END_ELEMENT: "title" END_DOCUMENT START_ELEMENT: "title" CHARACTERS: "Book02" END_ELEMENT: "title" END_DOCUMENT Using a Transformer... <?xml version="1.0" encoding="UTF-8"?> <title>Book01</title>

It's not good!

When I use a Transformer I should get something similar to: <?xml version="1.0" encoding="UTF-8"?> <result> <title>Book01</title> <title>Book02</title> <result>

That is, if result sequence length is greater than 1, getSequenceAsStream should wrap the sequence in a containing result element, so to not break JAXP Transformers and XMLStreamReader events.

Is there any solution in BaseX ?

Thanks a lot!

Isidre Guixà