W3C home > Mailing lists > Public > public-rdf-wg@w3.org > June 2012

ACTION-158 Turtle review

From: Pierre-Antoine Champin <pierre-antoine.champin@liris.cnrs.fr>
Date: Wed, 13 Jun 2012 12:38:07 +0200
Message-ID: <CA+OuRR-hpbUWU8Zh4jySb1k9b98L1RkmREZNhwOgJwQZkZgPLw@mail.gmail.com>
To: public-rdf-wg@w3.org

here comes (finally) my review of the turtle document;

* Sec1, Par2 : "a RDF graph" -> "an RDF graph"

* Sec1, example: a figure would be nice here, although not strictly

* Sec2, Sec3 and Sec4: are those sections really normative? It seems to me
that they are reader-friendly introductions to Sec5 and Sec6; which contain
the core of the specification.

* Sec2.3, Par1: "following a subject and predicate" -> well, not
necessarily directly. If a predicate list is used, the subject could be
much higher in the document. So I would rather write "following a
predicate", then later on replace "that subject" by "the corresponding

* Sec3.1, exemple: the example comes either too soon or too late: some ways
to write IRIs have already been presented (<>, a) but other have not been
yet (prefixed name). I would rather have the example at the end of Sec3.1
(may be as subsection 3.1.3 ?)

* Sec3.2.3 "the URL from which a particular SPARQL query was retrieved";
you mean "Turtle document" instead of "SPARQL query", right??

* Sec3.2, last paragraph before 3.2.1: the order of the 3 last sentences is
weird; I would rather put them in order 2,1,3, resulting in:

If there is no language tag, there may be a datatype IRI, preceeded by ^^.
The datatype IRI in Turtle may be written using either an absolute IRI, a
relative IRI or prefixed name. If there is no datatype IRI and no language
tag, the datatype is xsd:string.

* Sec3.2.3, example: prefix my: is not declared

* Sec3.3: "liberalized to allow:" -> "liberalized as follows" (gramatically
more consistent with the following list, containing whole sentences)

* Sec5.1, Par2: it is confusing to mention IRIREF here, as IRIREF is
already addressed by the previous paragraphn being in capital letters (I
smell that it comes from the renaming of production rules...). I would
replace that sentence by "White space is also significant in the production

* Sec5.3, table: I would introduce the table by a short sentence, e.g. "The
table below specifies in which context each kind of escape sequence can be

* Sec6.1, Par1: "four items" -> "five items"

* Sec6.1, bullet 3: "from string to blank node label" -> "from string to
blank node"

* Sec6.2, procedure for PNAME_LN: remove the last sentence (about relative
IRI resolution). If I got it right, namespace prefixes are already
absolute, so this is not required.

* Sec6.2, procedure for RDFLiteral: replace <code>IRIref</code> by

* Sec6.2, procedure for RDFLiteral: add the sentence "if neither a language
tag nor a datatype IRI is provided, the literal has a datatype of

* Sec6.3, Par3: "Each object <code>O</code>" -> "Each <code>object</code>"

* Annex A, Encoding consideration: "[0-9A-F]" -> "[0-9A-Fa-f]" (Turtle
accepts lower-case hex digits)

* everywhere: there are a number of references to production rule IRI_REF
(including links), which should be replaced by IRIREF

Received on Wednesday, 13 June 2012 10:38:40 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:49 GMT