Hi John,

I too wrote an RSS feed for someone's blog a couple of weeks ago. (I tried to tell you on the 6th and 7th, but I couldn't access your blog from work -- my work's connection? Anyway ...) I used pxmlutils to read the HTML, but surmised I had to write my own "accessors" to get elements out of the DOM tree (effectively), which was not hard but I thought pxmltuils (or some other package) would already have such routines. If you know about any I should have used, please let me know.

The DOM element accessor works by giving it a path of an element in the DOM tree to retrieve (I'm sure you've seen this before). For instance I learned where the blog title was located by reading in the blog via 'parse-html' as LHTML into the emacs scratch buffer and cruising it using the emacs sexp navigation functions. This was *much* easier than even reading prettified HTML to locate blog content components. So, I defined the following path to the blog title based on my scratch buffer cruising.

(defparameter *blog-title-path* '(:html :head (:body :p (:div (:center (:table (:tbody (:tr :td :td (:td :p :p 'string :p)))))))))

Then to get at that element I said

(path->element *blog-title-path*)

There are other details to talk about, such as text mastication routines. If you want, can we compare notes? I'm a relative newbie, so beware that you may not profit as much as I from such an exchange. :-)