W3C home > Mailing lists > Public > public-grddl-wg@w3.org > November 2006

Re: Hello world test case

From: Dan Connolly <connolly@w3.org>
Date: Wed, 15 Nov 2006 23:24:46 -0600
To: Ian Davis <ian.davis@talis.com>
Cc: public-grddl-wg <public-grddl-wg@w3.org>
Message-Id: <1163654686.30415.4.camel@dirk>

On Wed, 2006-11-15 at 22:27 +0000, Ian Davis wrote: 
> I've reviewed the first testcase[1] and in my view to make this test 
> case as clear and unambiguous as possible ...

Calling it the "hello world" test case might be misleading.

It's not my goal to make this test "as clear and ambiguous as possible".
My goal is just to decide that it's correct.

This is actually my Nth attempt at a hello-world example.

The 1st one was the DC.Subject example that was in the introduction
of the GRDDL spec for a few years; it actually goes waaay
back to 2000...

I wrote the guts of dc-extract.xsl on my palm pilot, over drinks
with Eric Miller and Dan Brickley in Amsterdam after WWW9 in an
effort to show them how easy it is to use XSLT to extract RDF
from real-world data.

  -- http://www.w3.org/2000/06/dc-extract/form

Then, after I we moved the tutorial introduction to GRDDL to
the primer and I re-organized it to explain the XML case first,
I couldn't find a test case for the XML case, so I made
one about extracting RDF from a spreadsheet.
  http://www.w3.org/2001/sw/grddl-wg/td/testlist1#projectsSpreadsheet

I then re-discovered the test case of this form that Dom had
already made... the P3P example...
  http://www.w3.org/2001/sw/grddl-wg/td/testlist1#xmlWithGrddlAttribute

Meanwhile, section 2 on GRDDL with XML had no example before
the formal rules. When I started thinking about slides to present
the rules, it became clear that I was going to need an example,
and I might as well put it in the spec. So I put the
spreadsheet example in there.

Then I realized that there was no good reason for the intro
to talk about Stephen King and The Stand and then switch to
spreadsheets. So I went back to Steven King and The Stand.
An example with 2 transformations and a relative URI is
not the simplest case, but it works reasonably well when
woven into the formal rules. I could successively elaborate
it in 3 steps, but I think that would be uncomfortably
verbose for this spec; it would start to look like
the primer.

The testlist1#xmlWithGrddlAttribute has been in the GRDDL
test suite since Feb 2005. I'm pretty sure every GRDDL implementation
ever built passes it. We are pretty much compelled to approve
it or else radically re-charter the group. Perhaps that
should be considered the hello-world test case.

I think testlist1#projectsSpreadsheet is only different in
ways that have nothing to do with GRDDL, so I recommend
the chair put one question to approve both of those tests.

If people aren't comforable approving testlist1#title_author until
issue-base-param is resolved, then perhaps we can do these
other 2 first.


> [1] http://www.w3.org/2001/sw/grddl-wg/td/testlist1.html#title_author

Where did you come up with that address, by the way? Are
there some links to it? I'd like to fix them; they're
probably not right.

For test suite purposes, I'm using the address without the .html
in it, since there are both .html and .rdf representations.
The name of the test is
  http://www.w3.org/2001/sw/grddl-wg/td/testlist1#title_author

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
Received on Thursday, 16 November 2006 05:25:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:11:46 GMT