[closed] Re: Another look at validate-with-xml-schema

"Vasil Rangelov" <boen.robot@gmail.com> writes:

> OK, after reading the previous thread, I re-read the definition of
> p:validate-with-xml-schema, and I have some thoughts (not related to the
> previous thread):
>
> 1. As far as I know, "lax" mode refers to validating the XML document
> against the schema declared inside of it if available, and assume valid if
> no such schema is declared. "strict" refers to validating the XML document
> against the schema declared inside of it and assume invalid if no such
> schema is declared. Shouldn't those two modes be clarified? If this is said
> elsewhere (say, the XML Schema spec), there should be a reference to the
> relevant section.

We've attempted to clarify that strict and lax.

> 2. Is the "schema" port required? If not, how does the processor behave when
> it's not present? Against the mode I suppose? This should be said
> explicitly.

We've added options try-namespaces and use-location-hints to give the
pipeline author more control.

> 3. What exactly is the point of the assert-valid option? If assert-valid is
> "false" and the document turns out to be invalid, what should happen?
> Nothing?!? Why so?

In the case of XSD validation, you might get a PSVI.

If these changes do not satisfy your concerns, please let us know.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | Always do one thing less than you think
http://nwalsh.com/            | you can do.--Bernard Baruch