TOPCAT and STScI registry
Mark Taylor
m.b.taylor at bristol.ac.uk
Fri Nov 15 09:23:17 PST 2013
Dear Reg'n'Apps,
this is an update on the long-broken interaction between the RI1.0
registry client in TOPCAT (also STILTS and my not-much-used freestanding
RegClient library) and the NVO/VAO/STScI registry.
Topcat talks to the registry mainly to locate Cone, TAP, SIA and SSA
services. It does this using the mandatory parts of the RI1.0
standard, which should in principle allow it to work with any
compliant searchable registry. There are not so many of these:
basically AstroGrid, Euro-VO, and STScI.
For several years now, it's worked with the AstroGrid and Euro-VO
registries, but the STScI one has just returned empty responses to
all queries.
The estimable Menelaus Perdikeas, to whom my thanks, working at ESA
to upgrade their registry implementations, has now tracked down why
this was happening. It turns out there were errors in both the
topcat client and the STScI server.
The main problem was that the Search query presented by topcat contained
<Where> elements with the wrong namespace: I had them in namespace
http://www.ivoa.net/xml/ADQL/v1.0 when they should have been in
http://www.ivoa.net/wsdl/RegistrySearch/v1.0. This error was
tolerated/ignored by the AstroGrid and Euro-VO registries, which made
me suspect that the problem might lie with the STScI service.
In fact the mandated response (p.13 of RI1.0) to an incorrect query is an
ErrorResponse fault rather than just an empty list of results, so the server
was behaving incorrectly as well. Strictly speaking, all RI1.0 services
should have been giving me back these errors rather than the records I
wanted (AG/Euro-VO) or an empty list (STScI).
Since RI1.0 is presumably on its way out in favour of RegTAP in the
forseeable future this may not be of great practical significance, but
future public releases of topcat (as well as pre-release versions in
the mean time, available from
ftp://andromeda.star.bris.ac.uk/pub/star/topcat/pre/)
will be able to query the STScI registry along with the other ones
properly.
Mark
--
Mark Taylor Astronomical Programmer Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-9288776 http://www.star.bris.ac.uk/~mbt/
More information about the registry
mailing list