This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
These two tests (in fn/matches.re.xml) assess characters that match the regex constructs \d and \D. These map to the Unicode category Nd. In Unicode 4.0.1 the Ethiopic digits were moved from category Nd to No, for reasons explained at http://unicode.org/review/resolved-pri.html issue 26. XPath processors are free to implement versions of Unicode before or after this change. Rather than make the test results conditional on which version of Unicode is in use, I propose simply to remove these characters from the tests.
The same two tests are affected by the introduction of x0BE6, Tamil digit zero. This was introduced in Unicode 4.1. More information on Tamil numbering is at http://www.unicode.org/notes/tn21/tamil_numbers.pdf Again, I propose to remove this character from the test since the results vary depending on the Unicode version supported.
In addition, test re00505 has a similar problem, caused by the addition of character x023F in Unicode 4.1. I have resolved these problems by removing the relevant characters from the tests.