N-Triples vs. RDF Test Cases compatibility
--- a/rdf-turtle/index.html Wed Mar 21 14:42:43 2012 -0700
+++ b/rdf-turtle/index.html Wed Mar 21 15:15:17 2012 -0700
@@ -1029,17 +1029,23 @@
<section id="n-triple-changes" class="informative">
<h3>Changes from RDF Test Cases format</h3>
<ul>
- <li>Default encoding is UTF-8 rather than US-ASCII only
+ <li>Encoding is UTF-8 rather than US-ASCII
<li>Uses IRIs rather than RDF URI References
<li>Defines a unique media type <code>application/ntriples</code>
<li>Subset of Turtle rather than Notation 3
<li>Comments may occur after a triple production
+ <li>Allows <code>\b</code> and <code>\f</code> for backspace and form feed
+ <li>More than one way to represent a single character
</ul>
</section>
<section id="n-triples-compatibility" class="informative">
<h3>Compatibility with previous RDF Test Cases N-Triples</h3>
+ <p class="issue">This section is very prescriptive and contains no information not contained above.</p>
<p>
- Full backwards compatibility with RDF Test Cases N-Triples is supported. An RDF Test Cases document written using only absolute IRIs is a valid N-Triples document generating the same triples. Both use <code>\u</code> escape sequences for characters outside US-ASCII and processing will have turned these into the original character. A N-Triples document that is serialized into ASCII and uses <code>\u</code> escape sequences for any character outside US-ASCII is equivalent to a RDF Test Cases document. Any comments may also be removed to avoid changes in the locations that comments are allowed.
+ Full backwards compatibility with RDF Test Cases N-Triples is supported. An RDF Test Cases document written using only absolute IRIs is a valid N-Triples document generating the same triples. Both use <code>\u</code> escape sequences for characters outside US-ASCII and processing will have turned these into the original character.
+ </p>
+
+ <p>N-Triples is also forwards compatible with existing RDF Test Cases format parsers. A N-Triples document that is serialized into US-ASCII and uses <code>\u</code> escape sequences for any character outside US-ASCII is equivalent to a RDF Test Cases document. In addition the backspace (<code>U+0008</code>) and form feed (<code>U+000C</code>) characters SHOULD be escaped using the <code>\u</code> sequence rather than using <code>\b</code> or <code>\f</code>. Any comments SHOULD also be removed to avoid changes in the locations that comments are allowed. Such a document provided as <code>text/plain</code> is indistinguishable from an RDF Test Cases format document.
</p>
</section>
<section id="n-triple-grammar">