Search Mailing List Archives

Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-owl] Variations in owl output of actual identical ontologies

Christian Knuepfer tral at
Thu Aug 9 06:26:22 PDT 2007

Hi all,

I found a interim solution for "normalising" owl files in order to get 
better results from CVS diff.

I use cwm

for this purpose. A single run of
   python --rdf <protege-output.owl> > <canonical-version.owl>

produces a normalised owl file, i.e. the entries are sorted in always 
the same way and have the same inner structure.

Then I put this normalised files under CVS...


Nick Drummond wrote:
>> There is an option in Protege 3.2.1:
>>    OWL/Preferences/"RDF/XML Writer Settings"
>> where you can choose between the Jena writer and an "Experimental  
>> native
>> writer".
> This was an attempt to "regularise" the ordering etc somewhat and may  
> be enough to help you with your problems.
> To deal with versioning properly, we need structural (or semantic?)  
> diff tools that ignore the syntax.
>> In Protege 4 I didn't found this option. Which writer will
>> Protege 4 use? Will it be the (completed) "native" one? There is no
>> statement about this in the Rlease Notes
>> release_notes_alpha.html
> All of the serialisation is done by the OWLAPI [1].
> Nick
> [1]
> _______________________________________________
> protege-owl mailing list
> protege-owl at
> Instructions for unsubscribing: 

Christian Knuepfer
Friedrich-Schiller-Universitaet Jena
Institut fuer Informatik
Ernst-Abbe-Platz 1-4
D-07743 Jena, Germany

currently at:
EMBL-EBI, Wellcome-Trust Genome Campus
Hinxton, Cambridge
CB10 1SD, UK

Mobile: Intl.+44-(0)7942-333513

e-mail: tral at

PGP Key-ID: 0xC3329342
PGP Public Key:
Fingerprint 5656 7F57 8E58 FC7B AC58 6714 3DA4 3181 C332 9342

More information about the protege-owl mailing list