Hi,
The parsers/serializers have been updated to the latest encoding of the
specification.
Some additional optional values have been included for the encoding of
the Document information item:
- additional data
- character encoding scheme
(still need to add decoding functionality for these).
Changes were made to support built-in entries for the XML namespace
prefix and XML namespace name in the prefix and namespace name tables. I
also put in error checking when parsing for literal values of the XML
namespace prefix and XML namespace name and for occurences of the XMLNS
prefix and namespace name. I have tried to do this efficiently as
possible taking advantage of the known lengths for specific states in
switch statements and comparing characters from end to start for XML and
XMLNS namespace names.
This should be the last changes we make to the encoding, unless there
are bugs in the code. The specification will not change now.
We can now add a set of XML documents and a set of fast infoset
documents that are created from the XML documents using the Fast Infoset
SAX serializer. We can then test for consistent encoding over releases
in addition to the existing per-release round tripping tests.
Paul.
--
| ? + ? = To question
----------------\
Paul Sandoz
x38109
+33-4-76188109