On Wed Feb 29 12:57:02 2012, tom christiansen wrote: > pod2html generates illegal UTF-8 because it creates HTML pages that > claim to be UTF-8: > > <meta http-equiv="content-type" content="text/html; charset=utf-8" > /> > > But then generates strings afflicted with the Unicode bug. Code > points 128-255 > come out as simple illegal bytes, unless there's a larger code point > in them. > > The right fix is to binmode the output handle to :utf8. > > Here's a list of pages to test. Note that you won't get a wide char > warning > if it is only 128-255; you'll simply get illegal output. > > perlebcdic.pod > perlgit.pod > perlhist.pod > perlpodspec.pod > perlthrtut.pod > > perl588delta.pod > perl5100delta.pod > perl5120delta.pod > perl5121delta.pod > perl5122delta.pod > perl5123delta.pod > perl5124delta.pod > perl5140delta.pod > perl5141delta.pod > perl5142delta.pod > perl5150delta.pod > perl5151delta.pod > perl5152delta.pod > perl5153delta.pod > perl5154delta.pod > perl5156delta.pod > perl5157delta.pod > perl5158delta.pod > > perlcn.pod > perljp.pod > perlko.pod > perltw.pod > > Notice also that you get differently wrong answers running with > PERL_UNICODE > set to 0 vs to SD. The program should not be sensitive to whether > that variable is set, because it knows the encodings of its input and > output, and should set things accordingly. > > --tom > I have a fix in a local post-5.16 branch to be pushed out after the release unless I hear this should go to blead now. Steve --- via perlbug: queue: perl5 status: new https://rt.perl.org:443/rt3/Ticket/Display.html?id=111446Thread Next