Bruce Eckel doesn't like XML. But alas - it's everywhere and he has to deal with it. So as you can expect, he goes and creates "general purpose XML manipulation library called xmlnode." for Python. That should be easy, right? Just one class, no need for more. Alas, it doesn't support namespaces, mixed content, CDATA sections, comments, processing instructions, DTD, Doctype, doesn't check well-formedness rules such as element and attribute names or allowed in XML characters etc. Well, that must be version 0.0...
Bruce Eckel's general purpose XML manipulation libraryRelated Blog Posts
No TrackBacksTrackBack URL: http://www.tkachenko.com/cgi-bin/mt-tb.cgi/593 2 CommentsLeave a comment |
Maybe XML is just too darn complex.. causes slow programs... and doesn't really solve any problems?
Yes, it looks like the code will break badly when fed mixed content, or even just contiguous text nodes. However it is built over an existing DOM-like library (minidom) which presumably handles things like CDATA and name validity constraints.
But no namespaces ? Bleh...
And he missed the single most useful feature in an XML library, that MSXML got right 6+ years ago : XPath navigation !
Cheers,
--Jonathan