Re: [xml] xmlParseFile and xmlns

[Michael Ludwig <mlu as-guides com>]

Juhani Connolly schrieb:
> When using xmllib to parse an rss feed that includes an unquallified
> xml namespace the document will include nothing.

Hi Juhani,

I guess you mean what's called the default namespace (no prefix).

> Are we meant to be doing something special before calling xmlParseFile
> (or another similar function)?

I'd say so, if we want to work with namespaces.

> <feed xmlns='http://www.w3.org/2005/Atom'
> xmlns:openSearch='http://a9.com/-/spec/opensearch/1.1/' >
>
> Feeding any similar feed through a barebones test program with an
> xmlParseFile and xpath operation for something like //item
> there are no hits.

The expression //item looks for element nodes named "item" *in no
namespace*. Your feed has "item" nodes in the Atom namespace.

> Upon deleting [xmlns='http://www.w3.org/2005/Atom'] it works fine, all
> the items are picked up.

Yes, because now the default namespace (no prefix) is mapped to the
empty namespace, not to the Atom namespace.

> If anyone knows what causes this issue and how to get around it
> correctly I would really appreciate any solution/explanation that
> could be given.

I don't know the C API, but you have to register namespaces with
prefixes in the XPath context and then search for //atom:item. Maybe
this pointer to the API doc will lead you on the right track:

http://xmlsoft.org/html/libxml-xpath.html#xmlXPathContext

Note the reference to "xmlNsPtr * namespaces : Array of namespaces".
I think this is what you need to fill, but not knowing the C API, I'm
not 100 % sure.

Michael Ludwig