HI Gisle, I figured out the root cause of the problem this line <META content=3.9 itemprop="rating"> in the html body will throw a START Event to perl
but this line is without END event and consequently causes TreeBuilder to start a new branch. my quick fix is to ignore the START event if tag is 'meta'. The code hparser.c just works fine as i stepped through the code. A future proof fix will require adding some logic on TreeBuilder to close any open closures if its parent is closing. Thanks On Sun, Sep 8, 2013 at 7:50 AM, Michael Song <michael.l.s...@gmail.com>wrote: > Hell All, > I've attached my html source code to demonstrate a incompatible parsing > behavior between Parser and Firefox firebug > If you open the below attached file with Firefox firebug, you will see > <div class="gd-grid-6 product-pricing"> is inside <div > class="listing-page-bucket"> enclosure > but when you parse it, the above relationship does not in the tree, > Is there anyway I can get around this problem? > > > Thanks > > > > > >
