ISSUE-58 (turtle-introduction): R2RML doesn't introduce Turtle well enough [R2RML]

ISSUE-58 (turtle-introduction): R2RML doesn't introduce Turtle well enough [R2RML]

http://www.w3.org/2001/sw/rdb2rdf/track/issues/58

Raised by: Richard Cyganiak
On product: R2RML

(Originally raised by Sören)

The R2RML spec assumes familiarity with Turtle, as stated in the introduction.

Nevertheless, one can *mostly* understand the R2RML spec while knowing very little Turtle. To read most of the spec, and to make simple changes to existing mapping documents, or even to create a complete mapping file by copy-pasting stuff, one really only needs to understand the concept of objects that are nested and have named keys.

There are a few places that break this understanding by diving deeper into Turtle details without sufficient explanation or examples. Smoothing them out would make the spec more accessible.

The following two concrete occurrences were pointed out:

1. The following commentary in 2.3:
http://www.w3.org/2001/sw/rdb2rdf/r2rml/#example-mapping

[[
To complete the mapping document, the DEPT table needs to be mapped in a triples map called <#TriplesMap2>, which is already referenced above.
…
The definition of <#TriplesMap2> follows.
]]

2. The following note in 4.2:
http://www.w3.org/2001/sw/rdb2rdf/r2rml/#syntax

[[
It is common to use document-local IRIs in mapping documents by defining the default prefix in the beginning of the document, and using it for creating IRIs for mapping components such as triples maps:

@prefix : <#>
…
:EmpQuery rr:sqlQuery """SELECT * FROM EMP WHERE …""".
…
:EmpTriples rr:logicalTable :EmpQuery.
]]

Received on Monday, 25 July 2011 13:39:13 UTC