[xquery-talk] BumbleBee XQuery Test Harness

Patrick Lehti lehti at ipsi.fhg.de
Fri Oct 17 11:29:09 PDT 2003


> 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. 

So far as I see, ERROR is not one of two legal results, but the only
legal result for all numeric functions using decimal, integer,
nonPositiveInteger, nonNegativeInteger, positiveInteger and
negativeInteger. This results in "passed" tests for all non- or
minimal-conforming query engines and "failed" tests for conforming
engines! This seems not to be fair and makes the (NIST)results unable to
compare!
If you need a tool to calculate the correct results for these tests, I
suggest using IPSI-XQ, which supports decimal und integer values with
arbitrary length since its very first release! ;-)

Patrick



More information about the talk mailing list