| sy crisp 2005-07-29, 9:04 am |
| I have a web page with a textarea into which users can enter text
including symbols such as =E1 =E9 =A3.
When I enter the text 'a =E1 =A3 $' and submit the form under CGI I get
'a \xc3\xa1 \xc2\xa3 $' returned via cgi->param.
Running the same code under mod_perl returns 'a =C3=A1 =C2=A3 $' which
looks as though every utf8 byte has been translated into latin1.
Below is an example script to demonstrate the problem.
Can anyone tell me what I need to do to stop this double encoding of
form data under mod_perl?
Or failing that, how I can convert from 'a =C3=A1 =C2=A3 $' back to 'a
\xc3\xa1 \xc2\xa3 $'?
Thanks,
Simon
#!/usr/local/bin/perl -w
use strict;
use CGI;
binmode STDERR;
my $q =3D new CGI;
print $q->header("text/html; charset=3Dutf-8");
print "<HTML><BODY>\n";
my $msg =3D $q->param('AAAA');
print "<br>AAAA: $msg\n" if $msg;
print STDERR "AAAA: $msg\n" if $msg;
print $q->start_form(-method=3D>'GET');
print $q->textarea(-name=3D>'AAAA');
print $q->submit(-name=3D>'DOIT');
print $q->end_form();
print $q->end_html();
exit;
CGI-> AAAA: a \xc3\xa1 \xc2\xa3 $
MOD_PERL-> AAAA: a =C3=A1 =C2=A3 $
|