XOM API for Processing XML with Java

[Elliotte Rusty Harold <elharo AT metalab.unc.edu>]

Wolfgang Hoschek wrote:
> The bug occurs because Crimson is reporting rubbish as the internal 
> subset, breaking roundtrip ability on subsequent setInternalDTDSubset().
> I know you worked around similar bugs before, so I'm wondering if that 
> could be worked around as well. It might well be unfixable. Here is a 
> test (running on the latest stable JDK 1.4.2 on MacOSX):

OK. I get it. The problem is not that Crimson is somehow misparsing the 
internal DTD subset like I initially thought, but that it's reporting 
bad data for that subset. Thus the problem is in the Builder, not the 
setinternalDTDSubset method. Maybe I can reproduce this now.

I'm not sure how to work around it though. Actually, maybe I can now 
that I realize what's going on but it's going to be a nasty hack. I'll 
have to recognize if I'm using Crimson, check the content model for 
#PCDATA, and then append the asterisk manually. Yuck. I'll explore this 
further.

Have you tested if you can reproduce this with a much simpler document? 
e.g.:

<?xml version="1.0"?>
<!DOCTYPE doc [
  <!ELEMENT doc (#PCDATA|a)*>
]>
<a />


--
Elliotte Rusty Harold  elharo AT metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim