Re^2: Perl 5.8.? and ParseExcel

[sth]

in reply to Re: Perl 5.8.? and ParseExcel
in thread Perl 5.8.? and ParseExcel

...Yes, I should have said > 255(unsigned char).

..maybe the module should be using 'U*'.

I think I did see that idea posted, somewhere in all my searching. I will try to change it and see if it works. Any idea why this is occuring with 5.8? Did Unicode not work in 5.6 ?

[sth]

Well the 'U*' didn't work, the code is parsing the xls file and adding it to a tab in another xls. Spreadsheet::WriteExcel complains about the characters being to wide. I think there has to be something I can set that will avoid this warning. Parse is defaulting to 'C*', because it is not UTF8, as far as I can tell. I saw something about setting LANG=en_us.UTF8. I'll do some more searching, since I really don't know to much about UTF8, in case you haven't noticed. :-)

[BrowserUk]

Perl's implementation of unicode changed quite a lot between 5.6.x and 5.8.x. It is, so I read, very much improved in 5.8.x but with that improvement came the need for a few non-backwards compatible changes and some extra warnings it seems. I've yet done precious little with it.

[sth]

-Thanks for the replies BrowserUk. Right now I put in a hack until I can figure out or find an answer,

 
    {
        local *STDERR;
        open(STDERR, '>/dev/null');
        $oBook = $oExcel->Parse($disclaimer) || die "Unable to parse $
+disclaimer";
    }

..not the right solution, but it works for now.

Thanks Again, sth

[BrowserUk]

Pragmatic! You could also do:

{
    local $SIG{ __WARN__ } = sub{};
    $oBook = $oExcel->Parse($disclaimer) 
        || die "Unable to parse $disclaimer";
}

ps. Dropping a line to the module owner might be good too.