view release on metacpan or  search on metacpan

redland/raptor/RELEASE.html  view on Meta::CPAN

</p>

<p>The RSS tag soup parser was updated to work better when there is
no base URI given.  It also now supports reading the 
<a href="http://inamidst.com/rss1.1/">RSS 1.1</a> format and
turning it into RSS 1.0 model triples.</p>

<p>Deprecated <code>raptor_ntriples_string_as_utf8_string</code> as
rather too internal to be useful, since it only works with a parser.
</p>

<p>More fixes to work around the broken libxml2 on Apple OSX 10.3.x
with inconsistent shared libraries / headers.</p>

<p>Experimental and incomplete Notation 3 parser - updated to match
changes to Turtle.  CVS changes only, not enabled in standard builds.
</p>



<h2 id="rel1_4_4"><a name="rel1_4_4">Raptor 1.4.4 Changes</a></h2>

<p>Make the RSS tag soup parser handle RSS 0.9 namespace elements by
turning them into RSS 1.0.</p>

<p>Fix a couple of crashes in the RSS 1.0 serialiser when
no base URI is used.</p>

<p>Make <code>raptor_uri_to_relative_counted_uri_string</code>
work when the base or reference URI have no paths such as like
<em>http://example.org</em>
</p>

<p>Added portability fixes for Win32 to get Raptor 1.4.3 building
with MS Visual Studio using expat and libcurl.  The RAPTOR_INTERNAL
define was moved to the build configuration and defines added for
integral type sizes.  Patch from Dave Viner (dviner at apache dot
org).</p>



<h2 id="rel1_4_3"><a name="rel1_4_3">Raptor 1.4.3 Changes</a></h2>

<p>A release with the major new feature of an XML writer API.
This is now used along with a new supporting XML element class to
improve the existing RDF/XML serializer and to provide a new
<a href="http://purl.org/rss/1.0/spec">RSS 1.0</a> serializer.
</p>

<p>This API it is also used by the next release of
<a href="http://librdf.org/rasqal/">Rasqal</a> to provide serializing
of query results to XML.
</p>

<p>The new <code>raptor_xml_writer</code> class functions added are:
<code>raptor_new_xml_writer</code> (constructor),
<code>raptor_free_xml_writer</code> (destructor),
<code>raptor_xml_writer_empty_element</code>,
<code>raptor_xml_writer_start_element</code>,
<code>raptor_xml_writer_end_element</code>,
<code>raptor_xml_writer_cdata</code>,
<code>raptor_xml_writer_cdata_counted</code>,
<code>raptor_xml_writer_raw</code>,
<code>raptor_xml_writer_raw_counted</code>,
<code>raptor_xml_writer_comment</code> and
<code>raptor_xml_writer_comment_counted</code>.
</p>

<p>The new <code>raptor_xml_element</code> class functions added are:
<code>raptor_new_xml_element</code> (constructor),
<code>raptor_free_xml_element</code> (destructor),
<code>raptor_xml_element_get_name</code>,
<code>raptor_xml_element_set_attributes</code>,
<code>raptor_xml_element_declare_namespace</code> and
<code>raptor_iostream_write_xml_element</code>.
</p>


<h3>Parser Changes</h3>

<p>RSS tag soup parser now works with older libxml2s (2.5.10+),
including the one shipped with some Apple OSX versions that has an
inconsistent header file and library.</p>

<p>RSS tag soup parser recognises/scores more common XML RSS file names.</p>

<p>RSS tag soup parser turns XML RSS
<code>&lt;guid isPermaLink="true"&gt;val&lt;/guid&gt;</code>
into RDF/XML form <code>&lt;guid rdf:resource="val"/&gt;</code>, leaving
the non isPermaLink form to be a literal value.</p>

<p>A bug was found in libxml2 that causes double expanding of XML
entities in RDF/XML.  This has been reported but cannot be worked
around from raptor.  The expat XML parser can be used as an
alternative, as it does not have this problem.  A test was added for
this bug but it will not cause the test suite ('make check') to
fail.</p>

<p>Added additional 
<a href="http://www.dajobe.org/2004/01/turtle/">Turtle</a>
parser tests that cover Notation 3 syntax that is not part of the
Turtle language.</p>

<p>Added
<code>raptor_parser_set_feature_string</code> and
<code>raptor_parser_get_feature_string</code>
methods to set/get string feature values.</p>


<h3>Serializer Changes</h3>

<p>Added feature <code>relative_uris</code> for serializers.  This
is used by the RDF/XML serializer and enabled by default.</p>

<p>Added feature <code>start_uri</code> for serializers with a string
value to set the start URI for serializing.  Not used at present.</p>

<p>Added new methods <code>raptor_serializer_features_enumerate</code>
to list serializer features and functions to set/get serializer
feature integer or strings values:
<code>raptor_serializer_set_feature</code>,
<code>raptor_serializer_get_feature</code>,