Closed heindorf closed 3 years ago
The NT files generated by Vectograph are invalid as they are missing the final dot. For example, the output
"5800860" must be changed to "5800860" .
Consider it done.
FYI: this file is generated only for logging purposes.
It appears to me that rdflib throws a parser error due to literals not because of underscores. If an RDF triple in ntriple format contains literal even with a correct a tag, it throws ParseError. Besides, for the embedding learing part, I do not need any tag as literals are ignored. Aside from this, All is done. We have an extra default parameter for base IRI. In the below written example, base_iri=https://ai4bd.com/resource/
<Event0> <https://ai4bd.com/resource/customer_id> "5800860"^^<http://www.w3.org/2001/XMLSchema#integer> .
<Event0> <https://ai4bd.com/resource/customer_name> <customer_b> .
<Event0> <https://ai4bd.com/resource/site_id> "5317663159"^^<http://www.w3.org/2001/XMLSchema#integer> .
<Event0> <https://ai4bd.com/resource/customer_number> <S300034> .
<Event0> <https://ai4bd.com/resource/op_group_id> "5328102884.0"^^<http://www.w3.org/2001/XMLSchema#double> .
<Event0> <https://ai4bd.com/resource/logistic_type> <SMART_BIN_CLASSIC> .
<Event0> <https://ai4bd.com/resource/location_uuid> <196492C81502483B93A2628C76942705> .
According changes are performed. Invalid NT file problem appears to be solved. I kindly ask you @heindorf to validate the correctness of the changes. If problem appears to continue, please reopen this issue.
The NT files generated by Vectograph are invalid as they are missing the final dot. For example, the output
<Event_0> <customer_id> "5800860"
must be changed to<Event_0> <customer_id> "5800860" .
Moreover, the nt file should be readable by
rdflib
andrdflib
requires valid IRIs, e.g., "http://ai4bd.com/resource/Event_0"For validation, you might use https://www.w3.org/2015/03/ShExValidata/. For testing, you might create a test case and try to execute
rdflib.Graph().parse(FILE, format='nt')
.