- From: Andy Seaborne <andy@apache.org>
- Date: Thu, 22 May 2014 11:48:15 +0100
- To: public-csv-wg@w3.org
On 22/05/14 04:29, Yakov Shafranovich wrote: > During the call earlier today I mentioned a distinction that may be > useful for the mailing list. There are two ways to look the conversion > templates we are working on. One way is that they will be the > canonical way to convert between CSV and the specified formats (XML, > RDF, JSON, etc.). A second way would be to develop generic conversion > guidelines and use these formats as a way to illustrate those with > concrete implementations. I am tending to lean towards #2. > > Thanks, > Yakov > Hi Yakov, #2 is, I think, the ideal we're working towards, with a more definite definition for RDF/JSON/XML. At the moment, there is some general text in http://w3c.github.io/csvw/metadata/#converting-tables that gives the conversion framework. There is a charter requirement: [[ 4. Standard mapping mechanisms transforming CSV to other formats (e.g., RDF, JSON, or XML). ]] and deliverable: [[ Mapping mechanism to transforming CSV into various Formats (e.g., RDF, JSON, or XML) (Recommendation) ]] If the graph-template style is adopted as the general underlying definitional framework (with code callouts, with a way to use metadata without templates and also without any metadata), we will have a way forward. That said, each syntax also adds it's own idiosyncrasies. We are a bit resource-constrained as well. Andy
Received on Thursday, 22 May 2014 10:48:45 UTC