Thread: [Rdfapi-php-interest] N3Parser -> TurtleParser?

So, I really like the N3Parser used by RAP. I'm wondering, though: since
the parser handles just about the same subset of Notation3 that's
implemented in Turtle (http://www.dajobe.org/2004/01/turtle/ ), I wonder
if it wouldn't make sense to revise the parser to parse Turtle
correctly.

Turtle was designed to be Notation3 with all the bits that aren't really
RDF left out; since this seems to be the point of the N3 parser in RAP,
it should be a good fit. Using a "real" spec (as much as Turtle is a
real spec) would be fairer to users, since they'd have a good idea what
will and won't work with the library.

In short: I'd like to hack on the N3Parser and N3Serializer such that
they parse and produce correct Turtle. In particular, I'd like to make
sure that the parser can handle the conformance test cases for Turtle
(http://www.dajobe.org/2004/01/turtle/#sec-conformance  ). I'd also
rename them to TurtleParser and TurtleSerializer and make the N3 classes
deprecated compatibility classes. And I'd add the conformance tests as
unit tests.

Any feelings on the issue? If I make a patch and it's righteous, can it
go in?

~ESP

-- 
Evan Prodromou <eva...@us...>

Hi Evan,

what you are saying makes total sence. There were no clearly defined =
subsets of N3 when we implemented the first version of the =
parser/serializer. As there is Turtle now and the parser (hopefully) =
does Turtle, we should rename it.=20

It would be great if you could make the parser pass the Turtle =
conformance tests. Contributions there and also better error handling =
are highly welcome.

Cheers from ISWC in Ireland,

Chris
=20
  ----- Original Message -----=20
  From: Evan Prodromou=20
  To: RAP Interest=20
  Sent: Tuesday, November 08, 2005 8:11 PM
  Subject: [Rdfapi-php-interest] N3Parser -> TurtleParser?


  So, I really like the N3Parser used by RAP. I'm wondering, though: =
since the parser handles just about the same subset of Notation3 that's =
implemented in Turtle (http://www.dajobe.org/2004/01/turtle/ ), I wonder =
if it wouldn't make sense to revise the parser to parse Turtle =
correctly.

  Turtle was designed to be Notation3 with all the bits that aren't =
really RDF left out; since this seems to be the point of the N3 parser =
in RAP, it should be a good fit. Using a "real" spec (as much as Turtle =
is a real spec) would be fairer to users, since they'd have a good idea =
what will and won't work with the library.

  In short: I'd like to hack on the N3Parser and N3Serializer such that =
they parse and produce correct Turtle. In particular, I'd like to make =
sure that the parser can handle the conformance test cases for Turtle =
(http://www.dajobe.org/2004/01/turtle/#sec-conformance  ). I'd also =
rename them to TurtleParser and TurtleSerializer and make the N3 classes =
deprecated compatibility classes. And I'd add the conformance tests as =
unit tests.

  Any feelings on the issue? If I make a patch and it's righteous, can =
it go in?

  ~ESP

        --=20
        Evan Prodromou <eva...@us...> =20