DataHub Issue: Load into DataStore error 409 with rdf/turtle ontology

Open Knowledge Labs DataHub
1

title: Load into DataStore error 409 with rdf/turtle ontology
slug: rdf-turtle-datastore-error-409
username: jziemer
dataset: https://datahub.io/dataset/fro-test

Hi,
Hope someone can help …
The purpose is to make US laws & regulations available for SPARQL querying on semantic.ckan.net/isparql/.

The test data set has 3 resources, rdf/turtle OWL ontology files. Two resources (US_LegalReference.ttl, LegalReference.ttl) load into the DataStore successfully. Resource (Code_Federal_Regulations.ttl) fails with error 409.

Protégé can open all 3 files open at the original source. And all have same header starting with "#baseURI: " - no extra keys in row one.

Many thanks
Jurgen

Error: CKAN DataStore bad response. Status code: 409 Conflict. At: https://datahub.io/api/3/action/datastore_create.
HTTP status code: 409
Response: {“help”: “https://datahub.io/api/3/action/help_show?name=datastore_create”, “success”: false, “error”: {“records”: ["row "1" has extra keys "# baseURI: http://finregont.com/fro/cfr/Code_Federal_Reg…
Requested URL: https://datahub.io/api/3/action/datastore_create

2

Hi @jziemer,

The DataStore is designed around tabular data, so it only accepts files like CSVs, Excel, etc

Having said that, I’m not familiar with http://semantic.ckan.net or how do you get data added to it. Looking at the domain name it looks like it’s an alias for http://linkeddata.openlinksw.com/, which would suggest that OpenLink are the maintainers. I’d suggest reaching out to them ( http://www.openlinksw.com) to get support on this front.

Hope this helps,

Adrià

3

Hi Adrià
Thanks for your reply. Following your advice, I submitted a ticket (# WEB 21803) to OpenLink. I will update the post with their response.

Still weird, if DataStore isn’t designed for non-tabular dataset resources. As per my ticket, 2 of the RDF/turtle files uploaded successfully. And there are many more RDF datasets from other users.

Regards
Jurgen

4

Hi Adrià and Admins
Please see the OpenLink response below:

  1. Open Link is no longer administrator for the DataStore (who is?).
  2. CKAN has been hosting RDF datasets for years
  3. The error is raised by the 'CKAN proprietary API ', data hub.io API

Can an administrator please try to upload the Code_Federal_Regulations.ttl. Do you get the same error?
It looks like the API is returning a wrong error code. There definitely are no extra keys in row 1. Can you see the original DataStore endpoint error code?

Thanks & regards
Jurgen

OpenLink Response
Tue, 05 Jul 2016 22:01:57 GMT

HI Jurgen,

Note this online support site if for OpenLink commercial customers to which priority support is given …

The http://semantic.ckan.net/sparql/ endpoint is something we hosted for the OKFN as part of of the LOD2 EU project (http://lod2.eu) we were partners with them in and host the metadata describing datasets hosted in there CKAN repository via the data hub.io API they provided for extracting it, and hasn’t really been maintained since 2014.

It is not clear to me what the 409 error you are receiving has to do with the SPARQL endpoint though, as you seem to be seeking to input data into the CKAN repository, which is not done via SPARQL as it uses the CKAN proprietary API as far as I am aware i.e.

Error: CKAN DataStore bad response. Status code: 409 Conflict. At: https://datahub.io/api/3/action/datastore_create.

You indicate the “US_LegalReference.ttl, LegalReference.ttl” RDF file loaded successfully but the “Code_Federal_Regulations.ttl” failed . But the CKAN guy says there datastore only accepts files in tabular form, which seems odd has CKAN has been hosting RDF datasets for years to my knowledge …

But they certainly are not uploaded via http://semantic.ckan.net/sparql/ which is just a read only sparql endpoint of the Metadata hosted in the datahub.io CKAN repository …

Best Regards
Hugh Williams
OpenLink Software

Your Update
Wed, 06 Jul 2016 13:04:40 GMT

Hello Hugh,
Thanks a lot for your prompt response. I will revert back to DataHub support.

Indeed,the files were uploaded via the DataHub. So the issue must be with the hub.io API.I only use iSPARQL to query the data.

Regards
Jurgen

5

Hi @jziemer,

The DataStore definitely only works with tabular data. The two resources that got imported were purely incidental, just because the parser identified the rows starting with a ‘#’ on the ttl files as records.

If OpenLink is not pulling data somehow from datahub.io then I’m afraid there is no active connection with http://semantic.ckan.net anymore. You are still more than welcome to host the files in datahub.io, but as I said, the DataStore only works with tabular data.

Sorry about that,

Adrià