Hi Hoss,
Comments inline:
>
> : Potentially -- my problem is, why is XMLWriter a sacred cow? There's not
>
> I'm not trying to suggest that it should be ... my point wasn't to
> suggest that we shouldn't modify XMLWriter because it needs to be
> preserved as is -- my point was that given XMLWriter's origins, trying to
> shoehorn general purpose functionality into it is prbably going to be an
> uphill battle, and it might be easier and more beneficial in the long term
> to approach the problem from a "what kind of utilities/helper code would
> make it easier for people to write custom Response Writers that generate
> structures of their own choosing in XML?"
+1, gotcha.
>
> : even a SOLR namespace or DTD to go along with it? Why should users be able
> : to inject their own custom XML into it -- it would seem that FieldTypes have
> : the ability to do this anyways (e.g., I can declare any new type of field
> : and its XML output [without namespaces sure, but still])...
>
> i'm not understanding what you mean ... i'm not fond of the way we have
> FieldType.write(XMLWriter xmlWriter, String name, Fieldable f) and
> FieldType.write(TextResponseWriter writer, String name, Fieldable f) ...
> but those methods don't allow custom fieldtypes to inject arbitrary XML,
> they're still confined to the existing XMLWriter and TextResponseWriter
> APIs -- those hooks just give the FieldType the ability to call the
> approraite write method (writeStr, writeInt, etc...) based on the
> inherient type of the data.
Sure, but my point was, I am still able to call xmlWriter.writePrim which
lets me write XML tags with unbound names in an arbitrary fashion. In other
words, I can make my field (for XML) instead of spitting out:
<arr name="booya">
<doc>
<int name="booya2">2</int>
...
</doc>
</arr>
Spit out:
...
<mycustomtag name="somename">someval</mycustomtag>
...
Using the writePrim method as it stands. So, I'm looking at a near term
solution versus a longer term solution. Your proposal is the correct
long-term solution, for sure. I would imagine then the XmlWriter that exists
would become nothing more than a thin wrapper around our robust
XMLReponseWriter framework that we'll write in order to keep backwards
compat with XmlWriter peeps, no?
Cheers,
Chris
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Email: chris.mattm...@jpl.nasa.gov
WWW: http://sunset.usc.edu/~mattmann/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adjunct Assistant Professor, Computer Science Department
University of Southern California, Los Angeles, CA 90089 USA
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
