On Jun 30, 2008, at 2:47 PM, Stefan Kuhn wrote:

Hi all,
today I had a look at the bug #1930029 and I added the test 
(org.os.cdk.io.cml.CMLCustomPropertyTest) to trunk and tried to fix the 
problems. But two problems remain:
1. In CMLCoreModule in line 608 and following there is code like
bondCustomProperty.add("true");//TODO here the value of the scalar must go in
I must admit I have no idea how to read the value of the current element here. 
Or is this considered another element? Anybody an idea?
2. The test also contains a comparision of the input to the output after the 
roundtrip. This fail (I commented them out). Differences are things like 
isotopenumber given in output, but not in input explicitly. Even if one could 
fix e. g. the isotope stuff, it is still questionable if a verbatim 
comparision makes sense for cml.
What do you think?

I plan to work on this. I got Jim Downing to put together a Jumbo release so I could get the things I needed in Jumbo into an official release that could be synched up with CDK. That has happened, but I've had some other distractions over the last 4-6 weeks. I should be able to get some significant time to work on this starting early next week.


   Daniel Zaharevitz
   Chief, Information Technology Branch
   National Cancer Institute