- From: Richard Cyganiak <richard@cyganiak.de>
- Date: Tue, 20 Dec 2011 19:12:49 +0000
- To: Bob Ferris <zazi@smiy.org>
- Cc: public-rdb2rdf-comments@w3.org
Hi Bob, This is an official response from the RDB2RDF Working Group regarding your comments [1] on the R2RML Last Call working draft: http://lists.w3.org/Archives/Public/public-rdb2rdf-comments/2011Sep/0005.html You said: > However, I guess that I'll > miss d2rq:TranslationTable [2]. I think this is a quite essential > functionality in this transformation process (e.g. you could also > utilise this functionality for data cleaning & cleansing tasks). > Unfortunately, this requires a programming language independent > representation of translation functions or an implementation that can > handle references to classes of specific programming languages (I guess, > CORBA-style or something like this). The working group considered this. As you correctly observe, the challenge is how to define an interface that is independent of programming language. Existing implementations of languages similar to R2RML (including D2RQ) only implement interfaces that rely on a specific programming language or interface. Given this lack of design experience, the WG decided not to tackle this feature in R2RML 1.0, but it is marked as one of the postponed issues that will be on the table in case a future working group is chartered for R2RML 1.1. An example for (very simple) translation using SQL functions was added to the R2RML specification: http://www.w3.org/2001/sw/rdb2rdf/r2rml/#example-translationtable For our internal Last Call housekeeping, can you please let the working group know whether you consider your comment addressed by this response? Thanks, and all the best, Richard
Received on Tuesday, 20 December 2011 19:13:20 UTC