Re: ACTION-133, ISSUE-22

* Richard Cyganiak <richard@cyganiak.de> [2011-06-21 14:39+0100]
> Some comments:
> 
> 1. The target language is called "Core SQL" as specified in ISO/IEC 9075:2008. This is commonly known as "Core SQL:2008". This is more accurate than "SQL 2008".

+1

> 2. An unknown dialect is expressed by simply omitting the attribute, not by a blank node. This should be stated explicitly. Recall:
> 
> [[
> RESOLUTION: Re ISSUE-22 the WG decides to introduce an OPTIONAL  
> attribute. if attribute is present, we assume it's a specific dialect.  
> if attribute is absent, we do not know what dialect it is. we only  
> specify the behaviour of R2RML where the attribute says it's SQL-2008.
> ]]
> 
> I am quite strongly opposed to the blank node approach. *Obviously* the rr:sqlQuery is in some language, so a blank node value tells us nothing that we didn't already know, so we can just drop the triple and it says the same thing.

I think this fails to meet David's goal of telling the implementation not to assume that the embedded language is SQL. I'm content either way.

> 3. It should say at least a sentence on where identifiers for other languages may come from.

I had deliberately left that information out because I didn't want to predict. Precedent indicates that sometimes these are strongly associated with a registry (as with the shortnames for XPointer registration schemes), defined by later revisions of standards (XHTML namespaces), invented by external organizations (as WS-1 did with SOAP-related identifiers) or invented by interested parties in the community (like most RDF identifiers or XML Signature's *Method identifiers).

> 4. Instead of
> 
> "This document identifies the SQL 2008 language."
> 
> better:
> 
> "This document assigns the URL http://...SQL2008 as an identifier for the Core SQL:2008 language."

+1

> 5. The prose text should speak of "other SQL versions" rather than "other languages". To the best of my knowledge, we have no use cases or requirements that call for mapping other languages than SQL.

It's true that we don't have any listed, though user expectations can be much more succinctly communicated to the implementations via other table-producing languages like tuple or domain relational calculus. While it's true that implementations are free to interpret whatever language they want, I thought it would be nice to not box them in to the point of having to abuse the abuse the spec.

Here's an example which relies on algebra syntax:
_:subjMap1 rr:languageIdentifier <http://...relationalAlgebra> ;
           rr:queryString "π ID,addr,city (People) ⋈ σ state="MA" Addresses" .

Critical? no.
Useful? could be.


> 6. I'd prefer if the property was called rr:sqlVersion, as that is much more obvious and aligns nicely with the rr:sqlQuery attribute.

I'd prefer rr:queryString and rr:languageIdentifier, but I agree that if we stick with SQLQuery, then rr:SQLVersion is more appropriate.

> Best,
> Richard
> 
> 
> On 21 Jun 2011, at 13:35, Eric Prud'hommeaux wrote:
> 
> > * Michael Hausenblas <michael.hausenblas@deri.org> [2011-06-21 09:09+0100]
> >> 
> >> All,
> >> 
> >> Below the agenda for our today's meeting.
> >> 
> >> Cheers,
> >>  Michael
> >> 
> >> ---------------------------------------------------------
> >> AGENDA Teleconference
> >> W3C RDB2RDF Working Group telephone conference 2011-06-21
> >> ----------------------------------------------------------
> >> *16:00-17:00 UTC*
> >> Local time:
> >> http://timeanddate.com/worldclock/fixedtime.html?month=06&day=21&year=2011&hour=16&min=00&sec=0
> >> 
> >> Bridge US: +1-617-761-6200 (Zakim)
> >> Conference code: 7322733# (spells "RDB2RDF")
> >> IRC channel : #RDB2RDF on irc.w3.org:6665 W3C
> >> ---------------------------------------------------------
> >> 
> >> Chair: Michael
> >> Scribe: Zakim, pick victim
> >> 
> >> 
> >> 1. Admin
> >> 
> >> PROPOSAL: Accept the minutes of last meeting http://www.w3.org/2011/06/14-rdb2rdf-minutes.html
> >> 
> >> 
> >> 2. ISSUE-29
> >> 
> >> PROPOSAL: Change to OPEN
> >> 
> >> context see: http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Jun/0101.html
> >> 
> >> 3. Review of open actions/issues
> >> 
> >> http://www.w3.org/2001/sw/rdb2rdf/track/actions/open
> >> 
> >> Please prepare. The goal is to close as many actions as possible and
> >> assign actions to issues where not done yet.
> > 
> > Per ACTION-133: Write proposal for issue-22, I propose the following text to create an optional rr:languageIdentifier to be used in conjunction with rr:SQLQuery:
> > 
> > 1.3 SQL Conformance ¶2
> > http://www.w3.org/2001/sw/rdb2rdf/r2rml/#sql-conformance
> > 
> > -In places where SQL expressions are allowed, these MUST conform to Core SQL 2008.}
> > +This document specifies the behavior of R2RML applied to SQL 2008 databases. The <code>rr:languageIdentifier</code> property can be used to identify other language versions where database queries are embedded in R2RML documents.}
> > 
> > 
> > 3.3.1.3 rr:languageIdentifier
> > below http://www.w3.org/2001/sw/rdb2rdf/r2rml/#LogicalTable_SQLQuery_Property
> > 
> > +    <h5><a name="LogicalTable_languageIdentifier_Property" id="LogicalTable_languageIdentifier_Property"></a>3.3.1.2 rr:languageIdentifier</h5>
> > +    <pre class="vocab">rr:languageIdentifier rdfs:range xsd:string .
> > +</pre>
> > +
> > +    <p>
> > +      While R2RML is defined only with respect to SQL 2008[<cite><a href="#SQL1">SQL1</a></cite>][<cite><a style="" href="#SQL2">SQL2</a></cite>], the it can be extended to apply to other languages.
> > +      The <code>rr:languageIdentifier</code> property is <em class="rfc2119">MAY</em> be used to identify the language embedded in the object of the <a href="#LogicalTable_SQLQuery_Property">rr:SQLQuery</a> property.
> > +      The identifier &lt;<a href="http://www.w3.org/2001/sw/rdb2rdf/r2rml/SQL2008">http://www.w3.org/2001/sw/rdb2rdf/r2rml/SQL2008</a>&gt; corresponds to SQL 2008.
> > +    </p>
> > +
> > +    <pre class="ex-mapping">
> > +[] rr:languageIdentifier <http://www.w3.org/TR/r2rml/SQL2008>
> > +   rr:SQLQuery """...""" .
> > +</pre>
> > +
> > +    <p>If the SQLQuery language is not SQL 2008 and there is no convenient language identifier for the SQLQuery language, a blank node may be used to indicate that the SQLQuery language is unknown.</p>
> > +
> > +    <pre class="ex-mapping">
> > +[] rr:languageIdentifier [] .
> > +</pre>
> > 
> > 
> > new document under in r2rml specification directory:
> > http://www.w3.org/2001/sw/rdb2rdf/r2rml/SQL2008.html
> > +<?xml version="1.0" encoding="utf-8"?>
> > +<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
> > +    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
> > +<html xmlns="http://www.w3.org/1999/xhtml" lang="EN" xml:lang="EN">
> > +<head>
> > +  <title>R2RML Identifier for the SQL 2008 Language</title>
> > +  <meta name="RCS-Id" content=
> > +  "$Id$" />
> > +  <link rel="stylesheet" type="text/css" href=
> > +  "http://www.w3.org/StyleSheets/TR/W3C-ED.css" />
> > +  <body>
> > +    <h1>R2RML Identifier for the SQL 2008 Language</h1>
> > +    <p>
> > +      This document identifies the SQL 2008 language.
> > +      Please see the ISO specifications [<cite><a href="../#SQL1">SQL1</a></cite>][<cite><a style="" href="../#SQL2">SQL2</a></cite>] for the language definition. and <cite><a style="" href="../">R2RML: RDB to RDF Mapping Language</a></cite> for the use of the identifier <a href="http://www.w3.org/2001/sw/rdb2rdf/r2rml/SQL2008">http://www.w3.org/2001/sw/rdb2rdf/r2rml/SQL2008</a> in R2RML.
> > +    </p>
> > +  </body>
> > +</html>
> > 
> > 
> >> 4. AOB
> >> 
> >> --
> >> Dr. Michael Hausenblas, Research Fellow
> >> LiDRC - Linked Data Research Centre
> >> DERI - Digital Enterprise Research Institute
> >> NUIG - National University of Ireland, Galway
> >> Ireland, Europe
> >> Tel. +353 91 495730
> >> http://linkeddata.deri.ie/
> >> http://sw-app.org/about.html
> >> 
> >> 
> > 
> > -- 
> > -ericP
> > 
> 

-- 
-ericP

Received on Tuesday, 21 June 2011 14:44:38 UTC