Re: [Rsslibj-developer] patch of a sort for rss100.java

[Joseph O. <jo...@en...>]

Yeah, I saw that - and I'm applying the fix now. (I was coding it as your
email came in, oddly enough.)

Are there any other extant issues? I'd like to do a call for votes on a
1.0 release soon soon soon, like today today today if everyone agrees.

On Tue, 25 Feb 2003, Joseph Shelby wrote:

> Joseph Ottinger wrote:
> >
> > Argh, the problem lives on. I'll examine it more closely. Joe, can you
> > show me some code that duplicates the issue? I'm hoping to have a new
> > release today and I'd like it to be correct.
>
> Well, part of it was my issue -- I hadn't set "about" in my generation
> of rss, relying just on the link tags to do what I needed to do.  I guess
> when writing rss100, either 1) an exception be thrown, or 2) the link value
> be used, if dealing with an Item that's missing its about property.
>
> However, it is very definitely "li" and not "item" that goes inside
> the rdf:Seq tag, so my fix is partially correct.
>
> > On Tue, 25 Feb 2003, Joseph Shelby wrote:
> >
> > > On Tue, 25 Feb 2003, Joseph Ottinger wrote:
> > >
> > > actually, 2 things are fixed:
> > > what i was getting was <item/>
> > > what i should get is <li rdf:resource="uri"/>
> > >
> > > getAbout() seems to be empty.  even when i used it first with the
> > > li tag, i got <li rdf:resource=""/> or <li/> (when EXML nicely ignored
> > > an attribute with a null value).
> > >
> > > again, if the about property isn't set by something, its empty and that's
> > > causing problems.  My reading of the O'Reilly article is that the uri needs
> > > to be the same as the one in the Item tag, so why not use getLink in both?.
> > >
> > > The reason is that in RDF, the order of elements technically doesn't matter,
> > > unless the elements are wrapped in an rdf:Seq container.  Even if RSS 1.0
> > > were to specify that that the Item tags are in the right order, to be RDF
> > > compliant (which I need to work with existing RDF tools here @ my company),
> > > the Seq is all that matters in the order of things.
> > >
> > > Joe
> > >
> > > > Joe, the only issue with applying this patch is that I'm not sure what
> > > > it's solving. You're adding the item to the sequence, and the *main*
> > > > difference I see is that you're using the item LINK instead of the item
> > > > ABOUT data... and those should be the same reference. Am I missing
> > > > something?
> > > >
> > > > On Mon, 24 Feb 2003, Joseph Shelby wrote:
> > > >
> > > > >     private void addItemsRef(Element channel, List is) {
> > > > >         if(is!=null && is.size()>0) {
> > > > >             Element items=channel.addElement("items");
> > > > >             Element seq=items.addElement("rdf", "Seq");
> > > > >             Iterator iter=is.iterator();
> > > > >             while(iter.hasNext()) {
> > > > >                 Item item=(Item)iter.next();
> > > > >                 // addItem(seq, item, "resource");
> > > > >           Element li = seq.addElement("li");
> > > > >           li.setAttribute("rdf", "resource", item.getLink());
> > > > >             }
> > > > >         }
> > > > >     }
> > > > >
> > > > > that now follows the rdf:Seq->li syntax described by O'Reilly's
> > > > > page
> > > > >
> > > > > http://www.oreillynet.com/pub/a/network/2000/08/25/magazine/rss_tut.html?page=3
>
>

---------------------------------------------------------
Joseph B. Ottinger                 jo...@en...
http://enigmastation.com                    IT Consultant