W3C home > Mailing lists > Public > www-rdf-calendar@w3.org > September 2003

Updates to tools

From: George Huo <ghuo@w3.org>
Date: Tue, 2 Sep 2003 13:45:12 -0400
To: www-rdf-calendar@w3.org
Message-ID: <20030902174512.GA20674@w3.org>

	
	I tried to send this message originally over a week ago,
but ran into problems with the W3C List Manager registering my 
approval. Hopefully this makes it through! 

----- Forwarded message from ghuo -----

To: www-rdf-calendar@w3.org
Subject: Updates to tools


Hello,

	I have been working on 2000/10/swap/pim/toIcal.py and
2002/12/cal/ical2rdf.pl for the past few weeks. I had the original goal
of round-trip testing, and I soon discovered that I had to add a good
deal of functionality to pass any test cases. Currently, all 20 test
cases in $CAL/test round-trip successfully. I adopted Tim's regression
test harness pieces from cwm, which now live at $CAL/test/cal-retest.py
and $CAL/test/cal-regression.n3. To run the tests:

$ cd 2002/12/cal/
$ make test 

Note that I've been testing using an old copy of swap with cwm 1.132;
something related to URI.py seems to break the tests in the current swap.
	toIcal.py is more modular now. It should be easier to add
enumerated and text fields; see exportGeneral().`X-' fields work, though
they currently depend on the namespace being called "x:". 
	In ical2rdf.pl, I fixed parameters that were enumerated types
(and should be symbols) that were being ignored. The parameter parsing
no longer trips up on values that contain colons enclosed in quotes (see
cal02). It also correctly handles fields that can be both enumerated
values or free text. 
	While coding and reading the spec, there were a few issues that
were unclear. First, toIcal.py was originally stripping spaces from
every field exported as text. This broke round-tripping in many of the
examples that contained "TZNAME=BST ". I removed the strip; is this how
we want to handle TEXT types? (see http://www.ietf.org/rfc/rfc2445.txt
4.8.3.2, 4.3.11)
	The ATTACH field is supposed to be of type `uri,' so it is
represented using rdf:resource. However, in a couple cases (see
20030115mtg), the usage appears:
	ATTACH;VALUE=URI:Ping
which results in:
	<attach rdf:resource='Ping'/>
Is it okay to directly convert something that's not a real URI?

Cheers,
-George

----- End forwarded message -----
Received on Tuesday, 2 September 2003 13:45:48 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:14:11 UTC