XML News from Tuesday, August 29, 2006

The Unicode® Consortium has released version 5.0 of the Unicode Character Database. Version 5.0 defines more than 99,000 characters. This includes the the Common Locale Data Repository 1.4 which supports 360 different locales covering 121 languages and 142 territories.

Version 5.0 adds five new scripts: Balinese, N’Ko, Phags-pa, Phoenician, and Sumero-Akkadian Cuneiform. New characters were also added for Cyrillic, Greek, Hebrew, Kannada, Latin, math, phonetic extensions, and symbols. Overall, there are 1,369 new characters in version 5.0.

New features in 5.0 include dependable caseless matching through stable case folding operations. Version 5.0 also revises and improves "property values and behavioral specifications in areas such as character, word, line, and sentence segmentation, and tightens conformance requirements on Bidi implementations (used for Arabic and Hebrew)." The Unicode Consortium has also released the Unicode Collation Algorithm 5.0 specifying default collation for all 99,000 characters.

The W3C XML Query Working Group and the XSL Working Group have released version 1.0 of the XML Query Test Suite (XQTS). According to Andrew Eisenberg,

We encourage implementors to run this test suite and request that they provide feedback to us by Sept. 29, 2006. If enough positive results are received, then we will be able to request a transition to Proposed Recommendation. During this period we will continue to respond to bug reports and will likely issue one or two point releases of the test suite.

In this release, we have added 230 test cases, including a small number of tests for fn:collection. We have changed the way that tests of fn:doc and tests that expect a context item from the host environment are expressed in our catalog (spelled out in Guidelines for Running the XQuery Test Suite).

We request that submittors re-read our guidelines if they have not done so recently and follow the guidelines for the transformation of queries and for the comparison of results as closely as possible. While we have provided a mechanism for submittors to tell us about any deviations that have been made, we hope that this will be used sparingly.

To date, we have received the results for several implementations of XQuery: Saxon-SA, xq2xsl, X-Hive/DB, xbird/open, XQuest,Qizx/open, and one anonymous implementation. A report that reflects these results is available from our web page. We will update this report as new results are received. We encourage implementors to send us results early, and then to update their results as their implementations progress.