5031 2007-09-11 06:35:55 +0000 Doctype detection fails if root element includes non "word" character 2007-09-11 06:50:39 +0000 1 1 1 Unclassified Validator check 0.8.1 All All RESOLVED FIXED http://qa-dev.w3.org/HEAD/dev/tests/5031-root_dash.xml P3 normal 0.8.2 1 ot ot www-validator-cvs oldest_to_newest 16453 0 ot 2007-09-11 06:35:55 +0000 The doctype detection routine in preparse_doctype() has the following regexp to detect FPI and SI: m(<!DOCTYPE\s+(\w+)\s+(?:PUBLIC|SYSTEM)\s+... the first (\w+) is the name of the document type, which has to be the root element (ref: http://www.w3.org/TR/xml/#vc-roottype ) but the \w+ is incorrect, as the root element can (among others) have a dash or dot. (ref: http://www.w3.org/TR/xml/#IDANQDS ) This half-breaks detection of the doctype for languages with root element including non "perl word (alphanum plus _)" characters. 16454 1 ot 2007-09-11 06:50:39 +0000 Patched in http://lists.w3.org/Archives/Public/www-validator-cvs/2007Sep/0071.html Adding test case. The validator should now report on this test case as: This Page Is Valid -//Recordare//DTD MusicXML 1.1 Partwise//EN! rather than This Page Is Valid XML!