Lists Home |
Date Index |
On 4/15/05, Ronald Bourret <firstname.lastname@example.org> wrote:
> Ken North wrote:
> > Developers wanted more, of course, such as being able to describe document
> > structure and shred documents (map document content to columns). That's also
> > been available for years with SQL platforms such as Microsoft SQL Server,
> > Oracle, Sybase and DB2 (XML Extender).
> Shredding is very useful for storing certain classes of data-centric
> documents, especially when non-XML documents need to access that data.
> In fact, I'd bet that shredding accounts for the vast majority of XML
> use with relational databases.
> However, shredding shouldn't be viewed as a substitute for native XML
> data storage as it doesn't preserve sibling order, comments, processing
> instructions, etc. and usually doesn't support mixed content. It also
> requires design-time mapping of the XML schema to the relational schema
> -- something that won't work with schemaless documents. So it shouldn't
> be viewed as a substitute for native XML storage but as a complementary
I think you're over generalizing more than a bit. Over the years I've
done a couple shredding implementations. In particular, I've
previously built a generalized approach on top of SQL server that
supported mixed content, sibling order and did not require a schema.
Adding comments and PI's to that would have been pretty much trivial.
There's nothing magic here; generalized hierarchical to relational
mapping algorithms have been around for a long time...