Open beckyjackson opened 5 years ago
Unfortunately, it seems like the format data is lost after the ontology is loaded with the OWLAPI OWLOntologyManager.
@beckyjackson have you tried this method?
Yes, unfortunately it returned null after loading with the ROBOT IOHelper
🙁
Do you know a way to keep that information @balhoff ?
No, sorry, I thought that would work! It's not something I have used before though.
The root node should be rdf:rdf
- @jamesaoverton
Jena?
I think at a minimum it should parse using Jena, which gives us a guarantee that it's in some RDF format, which is a more pragmatic requirement than the stricter RDF/XML IMHO
On Thu, Aug 15, 2019 at 2:27 PM Becky Jackson notifications@github.com wrote:
The root node should be rdf:rdf - @jamesaoverton https://github.com/jamesaoverton
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/OBOFoundry.github.io/issues/1018?email_source=notifications&email_token=AAAMMOL3JOPX5M5R6XVHWKDQEXC5DA5CNFSM4IKVL6SKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4NBQ4I#issuecomment-521803889, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAMMOPFBR2CCVRUVRJKF3TQEXC5DANCNFSM4IKVL6SA .
Even though RDF/XML is not my preferred format, I like the predictability of the strict rule.
Wait, I just thought, surely we can use the owlapi with only the rdfxml parser registered? If it fails, then it's not valid rdfxml
On Mon, Aug 19, 2019 at 8:00 AM Chris Mungall cjmungall@lbl.gov wrote:
Jena?
I think at a minimum it should parse using Jena, which gives us a guarantee that it's in some RDF format, which is a more pragmatic requirement than the stricter RDF/XML IMHO
On Thu, Aug 15, 2019 at 2:27 PM Becky Jackson notifications@github.com wrote:
The root node should be rdf:rdf - @jamesaoverton https://github.com/jamesaoverton
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/OBOFoundry.github.io/issues/1018?email_source=notifications&email_token=AAAMMOL3JOPX5M5R6XVHWKDQEXC5DA5CNFSM4IKVL6SKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4NBQ4I#issuecomment-521803889, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAMMOPFBR2CCVRUVRJKF3TQEXC5DANCNFSM4IKVL6SA .
This check should be expanded: the base ontology should be rdf/xml
However, we should also mandate that any imported ontology be at least some RDF format (turtle or xml).
Some ontologies have an .obo format file in their imports and this causes problems with OWL-based toolchains like Owlready2 (reported by SciBite cc @simonjupp )
For more on this particular instance of the issue: https://github.com/HUPO-PSI/psi-ms-CV/issues/26
I am not sure how best to implement this in robot/owlapi
This check should be expanded: the base ontology should be rdf/xml
I never knew we were going that way, but it makes sense with Jena in mind. Is there any hope to propose to merge release ontologies in general? having imports makes proper versioning really difficult to manage..
Update.
@cmungall requirement seems necessary. Is there another way to validate imports in a different format.
We now have a dependency on how to verify the format of imports.
cc @bpeters42
Does the file really need to be RDF/XML? Would not other OWL RDF formats be acceptable? (Maybe this should be a separate issue.)
@ramonawalls related discussion: #360
FP 2 - Common Format
Automated checks:
Mechanism:
We can ensure that the ontology properly loads in ROBOT, but this does not confirm that the format is RDF/XML. Unfortunately, it seems like the format data is lost after the ontology is loaded with the OWLAPI
OWLOntologyManager
. We can check the first line of the file to see if it starts with<?xml version=
. I'm open to other suggestions here.