R help - Jul 2008 - Removing or overwriting an XML node

Hi,

I have an existing XML document on disk, which I'd like 
to use as a template, and exchange a subnode with my own
newly created subtree:

	<?xml version="1.0"?>
	<Duncan>
	  <name a="1" b="xyz">
	    <first>Duncan</first>
	    <last>Temple Lang</last>
	  </name>
	</Duncan>

created by e.g.

?	library(XML)
	tr <- xmlTree("Duncan")
	  tr$addTag("name", attrs=c(a=1,b="xyz"), close=F)
	    tr$addTag("first", "Duncan")
	    tr$addTag("last", "Temple Lang")
	tr$closeTag()
	cat(saveXML(tr))

And now imagine I want to change <first>Duncan</first>
into e.g.? <initials>D.</initials>. How to do that ?

I am able to add my node:

	library(XML)
	x <-  xmlTreeParse("duncan.xml", useInternalNodes = TRUE)

        # find parent, add as last child:
	name <- getNodeSet(x, "//name")[[1]]
	newXMLNode("initials", "D.", parent=name)

        first <- getNodeSet(x, "//first")[[1]]
?	 # wanted: 
	# deleteXMLNode(name)
	# or
?	 # replaceXMLNode("initials", "D.", replace=first)

	cat(saveXML(x))
	free(x)

As seen above: how do I get rid of (no offense meant ;-)
<first>Duncan</first> ?
Alternatively, how do I overwrite <first>...</first> with my own
stuff ?

I checked the provided examples, and grep'ed the manpages, to no avail.

Thanks in advance, 
yours,
Steffen

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL:
<https://stat.ethz.ch/pipermail/r-help/attachments/20080702/9d71da4c/attachment.bin>

?Hi Martin,

thanks for your reply, I'll give it a try. My current problem is 
to map my problem to your suggestion of modifying the tree 
*at parsing time*. 

(Of course) my problem is more complex than my first example,
sorry for any confusion caused.

The XML looks more like this:

<root>
  <complex>stuff</complex>
  <list>
    <element>
      <param name="x" value="a"/>
?      <param name="y" value="b"/>
??      <more params/>
    </element>
?    <element>
?      <param name="x" value="i"/>
?      <param name="y" value="b"/>
?????      <more params/>
    </element>
  <list>
?  <more complex>stuff</more complex>
</root>

and I have my own set of values for "x",
so I want to use the first <element> as a template 
for my own <element>s, setting the value="" of
name="x"
to my new values. ?Finally I want to replace the whole <list> 
with my own <list> of my new <element>s.

My number of elements will not correspond 
to the number of <element>s in the template file.

So the handlers would need to read ahead to find the first <element>,
create copies of that for each of my values="" and return 
the new <list> instead of the one in the file.

I might have to read the file twice, first to obtain the <element>
as a template, create my <list> of these new <element>s, and then 
use the handler approach to insert my replacement <list>.

Because there is a lot of <complex stuff/> in there, 
I can't write the XML tree from scratch.
If anyone has suggestions how to parse the file, 
find the template <element>, create a <list> of them,
and finally replace the original <list>, I am all ears ;-)

Yours,
Steffen


On Wed, 2008-07-02 at 21:00 -0700, Martin Morgan wrote:
> Hi Steffen...
> 
> Steffen Neumann <sneumann at ipb-halle.de> writes:
...
> I'm not sure how to do this with internal nodes, but
> 
> handlers = list(
>   first=function(node, ...) {
>       init <- sub("^([[:upper:]]).*$", "\\1.",
xmlValue(node))
>       xmlNode("initial", init)
>   })
> xml <- xmlTreeParse("~/tmp/tr.xml", handlers=handlers,
asTree=TRUE)
> 
> gives
> > xmlRoot(xml)
> <Duncan>
>  <name a="1" b="xyz">
>   <initial>D.</initial>
>   <last>Temple Lang</last>
>  </name>
> </Duncan>
> 
> The documentation of xmlTreeParse is written to almost imply that this
> would also work with internal nodes, but I was not able to work
> through this.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL:
<https://stat.ethz.ch/pipermail/r-help/attachments/20080703/cf0dd2fe/attachment.bin>