Re: [xsl] XPath 2.0 Best Practice: wrap the first node of every path expression within schema-element?

[Florent Georges <lists@xxxxxxxxxxxx>]

"Costello, Roger L." wrote:

  Hi

> I would like to request a small change to the XPath 2.0
> specification.

> I would like to request that XPath implementations be required to
> generate an error or warning message for expressions like this:

>     schema-element(Book)/Foo

> where Foo is not a valid child of Book, according to the in-scope
> schema.

  While I think it would be *very* useful to have some kind of static
type analysis/checking, I feel that would be a mistake to request
processors to generate errors in such cases.

  For the most part of us, I gess, we are used to developing a
stylesheet and applying it to an XML file on the disk (by building the
tree in memory.)  But there are a lot of possible other applications:
think for instance to databases or streaming.  I am not a processor
internals expert, but I guess that can have implications on some
features those specific areas could need.

  Most of the projects I worked on had the same developing model: some
kind of editorial phase, where as much checks as possible are required,
then the production phase where time-consuming checking is not welcome.
 For the editorial phase, I always had problems with typos that were
not caught by the processor, and I think there can still be a lot of
improvements to help the XSLT developer.

  But I think those are really two different things: all the help a
processor can give the developer, thanks to the creativity of its
authors, and all the work its is required to do (that also means
"everything it can not do") in order to be conformant, because of the
RECs.

  Regards,

--drkm
























      _____________________________________________________________________________ 
Envoyez avec Yahoo! Mail. Capaciti de stockage illimitie pour vos emails. http://mail.yahoo.fr