This file records difference sbetween the RDF Tests CG test suite and Apache Jena.

== RDF/XML

/rdf/rdf11/rdf-xml/xml-canon/

These two tests include XML Literals.

The RDF test suite includes the enclosing XML namespaces in the literal.

"<br xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" xmlns:eg=\"http://example.org/\"></br>"^^rdf:XMLLiteral

but this is not necessary.

Jena (both ARP-based and RRX-based parsers only includes XML namespaces tyhat are used.

Hence: the result "<br></br>"^^rdf:XMLLiteral.

The manifgest processor requires the tests to file in RIOT/Lang/rdf-xml/xml-canon.

== Semantics and Model Theory tests are not run.

/rdf/rdf11/rdf-mt/
/rdf/rdf12/rdf-semantics/

ManifestProcessor has a list of manifests to ignore

== SPARQL

-- Codepoint escapes

/sparql/sparql12/codepoint-escapes#codepoint-esc-07
/sparql/sparql12/codepoint-escapes#codepoint-esc-bad-03

Use of \U unicode escpae sequences outside strings and URIs is not supported.
\u prociessing is done by javacc natively.

-- /sparql/sparql11/aggregates/manifest#agg-groupconcat-04

GROUP_CONCAT of "1"@en and "2"@en

ARQ does not add the language tag.

-- /sparql/sparql11/property-path/manifest#values_and_path

Returns one result ?v = 1, the test requires zero.
This is due to the optimizer.
--optimize=false is correct.
