Re: [xml] Best Process for InnerXML from Xpath Query

[Armin Bauer <armin bauer desscon com>]

Alex Neblett wrote:
> Hello again:)
>
> I must have caught a nerve or two judging by the response...not really sure
> why...
>
> Maybe I should try rewording the question...
>
> I would like to be able to read the innerxml as a string result of an Xpath
> query.
>
> So, for example, if I have some xml such as:
>
> ...
> <topnode>
> <anode>
> <bnode>sjdlkdsf</bnode>
> <cnode>lkjlkdsf</cnode>
> </anode>
> <tnode>
> <unode>sjsddlkdsf</unode>
> <vnode>lkjlkfddsf</vnode>
> </tnode>
> </topnode>
> ...
>
> and queried //anode the resulting string would be...
>
> <bnode>sjdlkdsf</bnode>
> <cnode>lkjlkdsf</cnode>
>
>
> I wrote a C# wrapper which calls libxml2 to do this, but I am not entirely
> happy with my approach as was looking for guidance as to a better way to do
> it...

I had a similar problem. But i dont know if my approach is better :)

I used a xmlwriter with a buffer for creating the output. So im looping
over the nodes i get and use the writer to create a new doc. In the end
i just return the content of the buffer.

> My example function is below:
>
> [STAThread]
> public static unsafe string singlenodeinnerxml(bool SourceIsString,string
> sourcexml, string XPathExpression, System.Text.Encoding xmlencoding)
> {
>         xmlInitParser();
>         _xmlDoc *doc;
>         if (SourceIsString == true)
>         {
>                 doc = xmlParseMemory(new
> System.Text.StringBuilder(sourcexml),sourcexml.Length);
>         }
>         else
>         {
>                 doc = xmlParseFile(sourcexml);
>         }
>         if (doc == null)
>         {
>                 return "";
>         }
>         _xmlXPathContext *xpathCtx = xmlXPathNewContext(doc);
>         if(xpathCtx == null)
>         {
>                 xmlFreeDoc(doc);
>                 return "";
>         }
>         _xmlXPathObject *xpathObj = xmlXPathEval(XPathExpression ,xpathCtx);
>         if(xpathObj == null || xpathObj->nodesetval == null)
>         {
>                 xmlXPathFreeContext(xpathCtx);
>                 xmlFreeDoc(doc);
>                 return "";
>         }
>         Console.WriteLine("\n[" + xpathObj->type1.ToString() + "]\n");
>         _xmlBuffer *mybuffer = xmlBufferCreate();
>         Console.WriteLine("\n[" + xmlNodeDump(mybuffer, doc,
> xpathObj->nodesetval->nodeTab[0], 0, 0).ToString() + "] bytes written\n");
>         byte *mybyte = xmlBufferContent(mybuffer);
>         int xpathlen = xmlStrlen(mybyte);
>         byte[] bytesArray = new byte[xpathlen];
>         for (int i = 0; i < xpathlen; i ++)
>         {
>                 bytesArray[i] = *(mybyte + i);
>         }
>         xmlBufferFree(mybuffer);
>         xmlXPathFreeObject(xpathObj);
>         xmlXPathFreeContext(xpathCtx);
>         xmlFreeDoc(doc);
>         string mystring = xmlencoding.GetString(bytesArray);
>         // mystring contains .outerXml
>         // the below string manipulations convert .outerXml into .innerXml,
> but the design is poor...
>         mystring = mystring.Substring(0,mystring.LastIndexOf("<")); //
> remove closing outerelement
>         mystring = mystring.Substring(mystring.IndexOf(">") + 1); // remove
> starting outerelement
>         System.GC.Collect();
>         System.Windows.Forms.Application.DoEvents();
>         return mystring;
> }
>
>
> Best Regards,
>
> Alex
>
>
> -----Original Message-----
> From: Daniel Veillard [mailto:veillard redhat com]
> Sent: Wednesday, July 13, 2005 7:25 AM
> To: Alex Neblett
> Cc: xml gnome org
> Subject: Re: [xml] Best Process for InnerXML from Xpath Query
>
> On Wed, Jul 13, 2005 at 07:21:25AM -0500, Alex Neblett wrote:
>
> > Hello,
> >
> > What is the best process to get the innerxml from an Xpath query using
> > the latest libxml2?
>
>
>   Do NOT use Reply on an existing mail thread to ask a completely unrelated
> question :-(
>   And I don't understand your question ... at all !
>
> Daniel

Attachment: signature.asc
Description: OpenPGP digital signature