The WebKit XSLT parser apparently ignores an XML file's DTD. This causes the following problems: 1) No custom entities are recognised. Further more, no error gets displayed stating that there were unrecognised entities found, they are instead silently ignored. 2) The XML document is not validated against the DTD. 3) Attributes of type "ID" (such as the 'id' attribute of XHTML) are not marked as such, causing the XSLT id () function to fail. An example can be found at http://astro.nickshanks.com/library/extrasolar.xml Pre-processed HTML output can be found at http://astro.nickshanks.com/library/extrasolar.en The two pages should look identical. The latter was produced by libxslt processing of the former.
Adding CC. Would also like to say that DTD does not even get parsed, but this is because Safari never downloads the DTD in the first place.
Ah Ha!, so thats whats exactly breaking the entities test., I was going to look deeper into this when I got back from my cousins- a small 2 week holiday!, but looks like you have set me onto the root of the cause, I had an idea it was something to do with the DTD!! so I am back to good ol' coding business!! Chris Jerome.
Oops. For better or worse I made a new bug: http://bugzilla.opendarwin.org/show_bug.cgi?id=7184 and have attached the fix there. I'm not sure if 7184 will address all of the issues covered here, but it probably will.
We don't respec *** This bug has been marked as a duplicate of bug 7184 ***