I liked the way this book was structured - it was a quick read over a thanksgiving family visit, and it gave a good overview of what XML is and what XML isn't, and what tools are available already in Perl to use it. I particularly liked the middle of the book, and how it dealt with trees and event streams. First there was in short intro chapter on why you'd want to parse XML as an event stream, and some simple modules that give back event streams. Then the next chapter would be a longer one on SAX - the definitive way to do event streams. Same thing with trees and DOM. It ends with some examples of real-life XML processing, such as consuming and producing RSS feeds. The book also has the best description of Unicode I've ever read, and did it with just a few pages.
What ticked me off about this book were the egregious errors in the sampe code. The very first piece of code they show in chapter 3 is a 100-line XML parser that doesn't need any support modules. The problem is that it doesn't recognize any XML because the regular expressions are wrong, which was pretty confusing for me (I'm relatively new to Perl, so I figured they were just "another way of doing it" that I didn't understand). I downloaded the examples frm the O'Reily website, and they're wrong there too - so it's not just a printing error. Worse, the example XML file I tried to test the parser on was also from the tarfile I downloaded - but it was invalid XML! (example 3.4). So I was trying to learn XML with a sample parser that didn't work, on invalid XML! This is not the quality I am expecting from O'Reilly!
(In fairness, both of these errors were in the online errata, but I'm not sure if they were corrected in the 7/04 reprint)
There are other errors in the code too - so be sure and check back with the errata page if you're going to seriously use the code. If they'd run their sample code before printing, I'd probably give this book a better rating.