Re: Editorial bugs?

HI Souri,

On Nov 1, 2010, at 15:39 , Souri Das wrote:

[snip]
>> 
>> Aha. That makes it a little bit more understandable. However... isn't it true that, somehow, extra triples generated by this foreign key should also appear in the A.2.2 example? I am not 100% what they should be. Is it something like
>> 
>> <xyz.com/emp/7369> emp:c_ref_deptno _:Department10 .
>> 
>>   
>> 
> That is correct. We will add it in the new version.
>> Plus all the triples generated by the example in A.2.1.?
>>   
>> 
> No additional triples (other than the triple you mentioned above) will be generated that will have emp:c_ref_deptno as predicate. 
> (I may be misunderstanding your comment here. Please elaborate if necessary.)

No, I think we are in the same line, it is purely editorial. If I look at the A.2.2 example, what it means is that it refers back to the triples generated in A.2.1. Up to that point all the examples were pretty much independent from one another but this is the point that binds them. It is just that making this clear to the reader might be helpful.

>> This is very unclear (for my taste) in the description of the foreign key stuff, which only speaks about the RDF property but does not speak about the object of the generated triple. The triple I have here is just, well, something that makes sense to me...
>>   
>> 
> The object for the predicate corr to a rr:foreignKeyMap is expected to be retrieved by the following query: 
> (it could return multiple objects if the foreign key constraint is not really enforcing a many-to-one child-to-parent relationship resulting in multiple triples with one child instance as subject)
> 
> SELECT parent.<rr:column in rr:subjectMap of rr:parentTriplesMap> 
>   FROM <rr:logicalTable in rr:parentTriplesMap> parent, 
>               <rr:logicalTable> child 
> WHERE <rr:joinCondition>

Ok. I think I understand that now, but I think this pseudo code should be in the text.  It also answers one of the questions I just had for myself, namely whether the 'parent' is the logical table or the original.

> 
> Would it help if we include a polished version of the above text in the rr:foreignKeyMap or rr:ForeignKeyClass section?
> 

Absolutely!

> --------------
> So, for A.2.1, the expected query will be:

You mean A.2.2, right?

> 
> SELECT parent.deptId
>   FROM 
>                (
>                 Select ('_:Department' || deptno) AS deptId
>                           , deptno
>                           , dname
>                           , loc
>                  from dept
>                 ) parent,
>                 (
>                 Select ('<xyz.com/emp/' || empno||'>') AS empURI
>                          , empno
>                          , ename
>                          , ('<xyz.com/emp/job/'|| job||'>') AS jobTypeURI
>                          , job
>                          , deptno
>                          , ('<xyz.com/emp/etype/'|| etype||'>') AS empTypeURI
>                          , etype
>                          , ('<xyz.com/graph/'|| job || '/' || etype||'>') AS graphURI
>                  from emp
>                 ) child
>                 WHERE (
>                      child.deptno = parent.deptno
>                  );
> 
> 

Thanks a lot Souri!

Ivan

----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf