Subject: | Tests fail (depending on default parser?) |
CPAN Testers shows a mixture of pass and fail reports: http://matrix.cpantesters.org/?dist=XML-Filter-NSNormalise%200.04
Statistical analysis of reports from my smokers suggests that the problem depends on the default parser --- it passes if the default parser in XML::SAX::PurePerl (theta=1) and fails with XML::LibXML::SAX (theta=0):
****************************************************************
Regression 'qr:(.*default parser.*)'
****************************************************************
Name Theta StdErr T-stat
[0='const'] 0.0000 0.0000 0.00
[1='eq_ XML::LibXML::SAX 2.0125 (default parser)'] 0.0000 0.0000 0.00
[2='eq_ XML::LibXML::SAX 2.0126 (default parser)'] 0.0000 0.0000 0.00
[3='eq_ XML::LibXML::SAX 2.0128 (default parser)'] 0.0000 0.0000 0.00
[4='eq_ XML::LibXML::SAX 2.0129 (default parser)'] 0.0000 0.0000 0.00
[5='eq_ XML::SAX::PurePerl 0.99 (default parser)'] 1.0000 0.0000 15234298105144452.00
R^2= 1.000, N= 33, K= 6
****************************************************************