- From: Jan Wielemaker <J.Wielemaker@vu.nl>
- Date: Sat, 1 Jul 2017 13:49:23 +0200
- To: Wouter Beek <wouter@triply.cc>, Gregg Kellogg <gregg@greggkellogg.net>
- CC: Dan Brickley <danbri@google.com>, Ivan Herman <ivan@w3.org>, public-rdf-comments Comments <public-rdf-comments@w3.org>, "Peter F. Patel-Schneider" <pfpschneider@gmail.com>
On 07/01/2017 11:52 AM, Wouter Beek wrote:
> Hi Gregg,
>
> I’m not aware of processors that don’t parse these tests.
>
>
> Two out of five processors I've tested did not pass the minimal
> whitespace test. These are rdflib 4.2.2 and SWI-Prolog's Semweb library
> 7.5.10 (the ones who did pass are Serd, Raptor, and Jena). Even though
> my sample size is small, this 2-out-of-5 score indicates that there may
> be other processors that do not pass this test either.
At least the SWI-Prolog parser evolved from the pre-1.1 n-triples specs.
It has always been in my mind that there should be white space between
the three elements of a triple. I really dislike the (I fear) fact that
this has been abandoned. It complicates managing triple files using
generic text tools significantly. With the demand to have comments on
their own line and white space between the three elements you can easily
process n-triples files with very simple tools and you can give more
reliable syntax errors. If we are allowed to write
_:a<http://example.com/p>"42" ^^ xsd:integer .
n-triples files get IMHO less readable, harder to parse and it gets
harder to give sensible error messages while I see no benefits.
Regards --- Jan
>
> I’m also not aware of any previously-ok-now-illegal tests based on
> our discussion.
>
>
> I agree, Peter's tests all seem to be of the
> previously-unclear-or-illegal-now-legal kind.
>
> ---
> Best regards,
> Wouter Beek.
>
> Email: wouter@triply.cc <mailto:wouter@triply.cc>
> WWW: http://triply.cc
> Tel: +31647674624 <tel:%2B31647674624>
Received on Saturday, 1 July 2017 11:50:01 UTC