This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
This test depends on the version of Unicode supported by the product being tested. In version 5.2 of Unicode, the codepoints listed below are in class No, and not NdNd. <regex value="[\d]"> <string value="፩" cp="4969" expected-match="true" actual-match="false" /> <string value="፪" cp="4970" expected-match="true" actual-match="false" /> <string value="፫" cp="4971" expected-match="true" actual-match="false" /> <string value="፬" cp="4972" expected-match="true" actual-match="false" /> <string value="፭" cp="4973" expected-match="true" actual-match="false" /> <string value="፮" cp="4974" expected-match="true" actual-match="false" /> <string value="፯" cp="4975" expected-match="true" actual-match="false" /> <string value="፰" cp="4976" expected-match="true" actual-match="false" /> <string value="፱" cp="4977" expected-match="true" actual-match="false" /> </regex>
In fact these characters (the Ethiopic digits) are still in class No rather than Nd in Unicode 6.2.0. (Apparently they are not classified as Nd because there is no '0' in the family). This seems to be a plain error in the test.
There is consensus: the Ethiopean digits are no digits in the sense that they do not belong to Nd, but to No. I have removed them from the test. See rev#442, 2014-05-23 in the repository. Resolving this bug as fixed.
THanks.