non-robust import: recursive imports failing on non-matching URIs

[GoogleCodeExporter]

What steps will reproduce the problem?
1. import http://usefulinc.com/ns/doap# by using http://usefulinc.com/ns/doap 
(no #) as an uri or
1a. import http://ai-nlp.info.uniroma2.it/ontologies/heritage by using the 
address http://art.uniroma2.it/ontologies/heritage (art is alias for 
ai-nlp.info)

What is the expected output? What do you see instead?

doap should import foaf
heritage should import geographic 
(http://ai-nlp.info.uniroma2.it/ontologies/geographic)

recursive imports are not resolved

this happens because the import on a URI which is not the exact URI of the 
ontology but is a valid URL for accessing it succeds, but then the imported 
data is about the real URI of the ontology and not on the one used for the 
import.

Actions:
1. a good URL but no real URI should be replaced with the correct URI if an 
ontology declaration is present.
1a. use a warning reply for the user when non-matching domain/url are being 
prompted, don't put any warning for differences like trailing / or # and just 
replace them

Original issue reported on code.google.com by stell...@info.uniroma2.it on 24 Nov 2011 at 9:59

[GoogleCodeExporter]

Original comment by stell...@info.uniroma2.it on 24 Nov 2011 at 10:06

[GoogleCodeExporter]

Original comment by stell...@info.uniroma2.it on 24 Nov 2011 at 10:07

[GoogleCodeExporter]

Original comment by stell...@info.uniroma2.it on 26 Nov 2011 at 12:56

• Changed state: Fixed