I'm going to be speaking at the Yet Another Perl Conference later this
month about HTML::Element and HTML::TreeBuilder.  I've done my
write-up for the procedings, and Kevin Lenzo says he won't need it in
final form for another few days (maybe until Jun 3?).  In the
meantime, all are welcome to read the addled mess I've put together as
notes for my talk and email me any corrections, questions, or
suggestions.

The write-up is at:
  http://www.speech.cs.cmu.edu/~sburke/y2c.html

The "look_up" and "look_down" methods mentioned are, incidentally, new
(as in about 12 hours old as I write this), and not yet in CPAN.
They, well, behave as described in that document.  look_down is
basically an extension of find_by_attribute.  look_up operates the
same but recurses up from the currentnode instead of traversing down
from that node.

(The emphasis of the talk shifted as I did the write-up: it started
out with me spending a lot of time bitching about some of the more
difficult cases of weird/bad HTML code that HTML::TreeBuilder has to
handle.  But then I thought it's more relevant to people if I talk more
about extracting information from trees.  And anyhow, hopefully
TreeBuilder is at a stage now where everyone can just treat it like a
black box and never have to worry about the Lovecraftian terror that
its actual parsing logic consists of.)

-- 
Sean M. Burke    [EMAIL PROTECTED]    http://www.spinn.net/~sburke/

Reply via email to