[xquery-talk] BumbleBee XQuery Test Harness

[Jason Hunter]

Sander Bos wrote:
>>>I think we produce the result in the right way, so surrounded by a
>>>results element? But BumbleBee does not like it.
>>
>>Yes, you are correct.  That also is fixed in 1.0.1.
> 
> 
> 'fraid not, it just went from 596 such errors to 546, e.g.

Alas.

> Because of these errors, is it reasonable to assume that none of the
> products you wrote adapters for currently have a close to 100% score on
> all the tests?

That would be a good guess, but actually the issue with these NIST tests 
is that they use numeric values exceeding the bounds that engines are 
required to support.  That's why you see ERROR listed as one of two 
legal results.  I don't recall finding an engine until now that both 
supported the arbitrarily sized numeric values and had a good 
understanding of the surrounding calls used by the NIST tests. 
Awkwardly the NIST tests themselves are just the queries, no results, so 
I had to write some of them in by hand.  Your engine is better at 
calculating results than my hands.  :-)

I'm hoping Carmelo updates the NIST tests to have numeric values within 
the mandatory-to-support range.  That'll make the NIST set of tests more 
interesting.

> Your other changes helped quite a bit. We now have an 83.3% (*) score on
> the usecases tests (I may as well say it as I attach an Adapter for
> X-Hive/DB). We have failures in Strong.bee because we do not implement
> import schema yet, and errors in R.bee that we have to look into
> further.

OK, great!

> I attached my adapter implementation for X-Hive/DB as indicated that I
> would yesterday, maybe it can be useful to others.

Absolutely.

-jh-