prototype TAP-1.1 service

[Mark Taylor]

On Tue, 2 Jun 2015, Patrick Dowler wrote:

> 
> We did recently switch to tomcat7 and it could be less tolerant about repeated
> /. I don't see any downside so will modify our service registration.

thanks.

> Pat
> 
> PS-But, there is a general issue with appending stuff to URLs that usually
> requires parsing them first to see exactly what state they are in (how they
> end)... e.g. do you start with a ? or a & to append query params? I'm always
> writing code to be careful of leading and trailing / so it seems like one of
> those fiddly issues that have to be dealt with by everyone :-(

I'm quite happy to write defensive code that strips off trailing
slashes where present if that's considered the right thing to do.
But I can (just about) imagine a service which requires a double
slash since that's what TAP seems to say.  ? and & is a bit 
different, I don't think base_url??a=b or base_url?a=b&&c=d is
ever going to be correct, so it's more obvious what's the
right/paranoid thing to do there.

> 
> On 02/06/15 06:49 AM, Mark Taylor wrote:
> > Pat,
> > 
> > On Wed, 27 May 2015, Patrick Dowler wrote:
> > 
> > > I have updated the CADC TAP service with some TAP-1.1 goodies:
> > 
> > I think there's an issue with the service registration and/or
> > implementation of the TAP service ivo://cadc.nrc.ca/tap.
> > The service URL is listed in the registry as
> > 
> >     http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/tap/
> > 
> > If I submit jobs to:
> > 
> >     http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/tap/async
> > 
> > it works.  But if I append /async to the registered service URL I get
> > 
> >     http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/tap//async
> > 
> > and attempting to POST there gets me a 404.  My reading of TAP 1.0
> > is that appending /async is the right thing to do, and that's what
> > topcat's TAP client does.  Can you either fix it or argue it's
> > somebody else's fault (possibly mine)?  Of course I could work
> > around it easily enough.
> > 
> > I don't *think* I used to see this; I suspect that the implementation
> > has become less tolerant of the repeated solidus, but possibly the
> > registration has changed.
> > 
> > Thanks
> > 
> > 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/
> > .
> > 
> 
> -- 
> 
> Patrick Dowler
> Canadian Astronomy Data Centre
> National Research Council Canada
> 5071 West Saanich Road
> Victoria, BC V9E 2E7
> 
> 250-363-0044 (office) 250-363-0045 (fax)
> 

--
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/