cr pingback

What is first

• It's useful to report Finding Linksets among Linked Data Bubbles back to datahub.io.
• cr pingback requires [CSV2RDF4LOD_PUBLISH_DATAHUB_METADATA_OUR_BUBBLE_ID](CSV2RDF4LOD environment variables) to know what datahub.io dataset to update.
• cr-pingback gets hung up by https://github.com/timrdf/DataFAQs/issues/125, it's currently a manual one-time fix.

What we will cover

This page describes how cr-pingback.sh dereferences a Prizms node's /void and submits it to its datahub.io listing using DataFAQs' add-metadata.py.

Let's get to it!

add-metadata.py only processes instances of datafaqs:CKANDataset. /void is typed to datafaqs:CKANDataset as part of the processing in the following two places. So, one of them must occur before datahub.io will be updated.

• bin/secondary/cr-linksets.sh is enabled as a secondary derived dataset and asserts:
  • owl:sameAs datahub.io; a datafaqs:CKANDataset; void:subset linkset
  • void:vocabulary
  • cr-dataset-uri.sh argument --abstract
• bin/util/cr-full-dump.sh asserts:
  • a void:Dataset, dcat:Dataset; void:rootResource; void:dataDump; dcat:distribution...
  • a datafaqs:CKANDataset; owl:sameAs datahub
  • a doap:Project with doap:repository
  • void:exampleResource
  • void:triples and dcterms:date

add-metadata.py only processes instances of datafaqs:CKANDataset, which isn't asserted until cr-full-dump.sh has a pass.

cr-dataset-uri.sh asserts <$CSV2RDF4LOD_BASE_URI/void> a void:Dataset;, so anything that publishes its response into the triple store will get /void to return some RDF description.

What is next