Hi, I don't know why but resolve_entities=False works for me only when DOCTYPE declaration is present. For your case will work: tree = etree.parse(StringIO('<!DOCTYPE aRoot SYSTEM "some_dtd_link"><aRoot>&aReference;</aRoot>'), parser) Just be careful because unparsed entities are not just parts of element's text, they are this element's child objects. This causes problems with attributes (I don't know if this is a bug or feature) because attributes are strings and can't contain objects inside. By this I mean the following. If you have xml like: <aRoot>bla &aReference; foo</aRoot> You'll get:

...

...

print tree.getroot().text bla print tree.getroot()[0] &aReference; print tree.getroot()[0].tail foo

But when you have xml like: <aRoot><aChild attr="bla &aReference; foo" /></aRoot> You'll get two child elements inside aRoot:

...

...

print tree.getroot()[0] &aReference; print tree.getroot()[1] <Element aChild at c35e10> print tree.getroot()[1].get('attr') bla foo

So, parser prepends entity object before child node. That doesn't look nice... -- Marat On Tue, Nov 10, 2009 at 2:08 AM, Emanuele D'Arrigo <manu3d@gmail.com> wrote:

Hi everybody,

today I tried this:

parser = etree.XMLParser(resolve_entities=False) tree = etree.parse(StringIO("<aRoot>&aReference;</aRoot>"), parser)

I was expecting it to simply treat the entity reference as text but it raises an error. What is the parser's parameter resolve_entities for then?

Manu

_______________________________________________ lxml-dev mailing list lxml-dev@codespeak.net http://codespeak.net/mailman/listinfo/lxml-dev