Reported by jteh on 2013-09-23 00:52
XHTML documents break very badly with the MSHTML engine, including IE. The problem is that nodeName returns lower case names in XHTML (as per the spec), but we expect upper case, which is what is returned for normal HTML.
The simplest way to fix this is probably to upper case the node name before we use it in both the Python and C++ MSHTML code. (We could lower case everything, but that would mean a much larger diff.) For searching, I guess we'll need to search for both upper and lower case versions.
Note that in the MSHTML vbuf backend, we currently test for the math tag in lower case, as it always seems to appear like this, at least with MathPlayer installed. This will of course need to be changed if the node name is converted to upper case.
Reported by jteh on 2013-09-23 00:52
XHTML documents break very badly with the MSHTML engine, including IE. The problem is that nodeName returns lower case names in XHTML (as per the spec), but we expect upper case, which is what is returned for normal HTML.
The simplest way to fix this is probably to upper case the node name before we use it in both the Python and C++ MSHTML code. (We could lower case everything, but that would mean a much larger diff.) For searching, I guess we'll need to search for both upper and lower case versions.
Note that in the MSHTML vbuf backend, we currently test for the math tag in lower case, as it always seems to appear like this, at least with MathPlayer installed. This will of course need to be changed if the node name is converted to upper case.