commons-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rahul Akolkar" <rahul.akol...@gmail.com>
Subject Re: [scxml] schema problem within <assign> node
Date Thu, 27 Apr 2006 21:21:34 GMT
On 4/27/06, Nestor Urquiza <nestoru@yahoo.com> wrote:
> So guys,
> either we change the @name by @location or we do not
> use the schema to validate ... thoughts?
>
<snip/>

There is another option since you host the schema yourself, if you'd
like. And thanks for following up on this promptly.

Anyway, from an implementation perspective, having an attribute mean
multiple things means we have to figure out beforehand if the
expression is a "degenerate path", and in one form or the other, this
will boil down to:

<pseudo>
if (means one thing) {
   do the thing one thing does;
} else (means another thing) {
  do the thing another thing does;
} else {
  throw up in the log file;
}
</pseudo>

You get none of that with the @name and @location distinction (for
"scratch space variable" and "XML data tree node" respectively). Also,
as Jim mentions, the WD is subject to change WRT the data model, so I
don't know if we should urgently move to a sub-optimal impl for
assign, especially since the @name usage does appear in the var
section of the draft.

-Rahul


> Thanks!
>
> --- Nestor Urquiza <nestoru@yahoo.com> wrote:
>
> > Thanks Jim,
> >
> > That was exactly my assumption while we were
> > discussing the commons-scxml implementation in java
> > for the scxml draft.
> >
> > Just a question about the schema. When do you think
> > it
> > will be published in a w3c url? Recently I have
> > notice
> > that urls that were working before are no longer
> > working like the one that specified the DTD for XSLT
> > "http://www.w3.org/1999/XSL/Transform"
> >
> > Thanks a lot for your help,
> > Nestor
> >
> > --- "Barnett, James" <James.Barnett@aspect.com>
> > wrote:
> >
> > >
> > > The current version of  the spec uses the
> > attribute
> > > 'location' to
> > > generalize
> > > the concept of 'name'.  The idea is that a
> > location
> > > is designated by
> > > some sort of path expression, and a simple
> > variable
> > > can be viewed
> > > as a degenerate path consisting of only one link.
> > > So to assign to a
> > > variable named 'foo', write <assign location="foo"
> > > ...>.
> > >
> > > However, we warn that the datamodel is the least
> > > stable part of the
> > > specification and may change significantly in
> > future
> > > drafts.
> > >
> > > - Jim
> > >
<snap/>

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-user-help@jakarta.apache.org


Mime
View raw message