[Owl-devel] Null Pointer Exception in readCrystalData

[Jose M. Duarte]

Hi Henning

(posting to owl-devel directly)

Thanks for that, I didn't even know this was happening. Just checked with
2l02 and indeed I can reproduce the problem, I think there might be
something strange in that entry, the parsing tests do pass (which means that
the tests are not good enough!). Have you seen the problem with other cif
files too?

Re the casp file problems, do you have an example of one of those casp
files? I think both problems should be easily solvable without having to go
for optional parsing. If you can send me one example I'll look into it.

Cheers

Jose



On 09/01/2010 12:37 PM, Henning Stehr wrote:

Continuing with the next problem right away...

I'm getting such errors when loading cif files (e.g. 2l02):

Exception in thread "main" java.lang.NullPointerException
        at owl.core.structure.CiffilePdb.readCrystalData(CiffilePdb.java:374)
        at owl.core.structure.CiffilePdb.parseCifFile(CiffilePdb.java:296)
        at owl.core.structure.CiffilePdb.load(CiffilePdb.java:173)
        at owl.core.structure.Pdb.load(Pdb.java:259)
        at owl.casp.tools.MapStru2Seq.main(MapStru2Seq.java:69)

I had similar problems earlier with the new b-factor parsing and when
I checked I found that Casp models often break the correct format for
b-factors (which we had previously ignored). As a workaround I fixed
the format in the files but I probably can't do that for cyrstal data.
Can we somehow make parsing all these things optional? I'm not sure
how to best implement this, but mabe some kind of legacy or lenient
parse mode?

Thanks,
Henning
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.bioinformatics.org/pipermail/owl-devel/attachments/20100901/a755b173/attachment.html>