view release on metacpan or  search on metacpan

lib/Devel/StackTrace/AsHTML.pm  view on Meta::CPAN

our $VERSION = '0.15';

use Data::Dumper;
use Devel::StackTrace;
use Scalar::Util;

no warnings 'qw';
my %enc = qw( & &amp; > &gt; < &lt; " &quot; ' &#39; );

# NOTE: because we don't know which encoding $str is in, or even if
# $str is a wide character (decoded strings), we just leave the low
# bits, including latin-1 range and encode everything higher as HTML
# entities. I know this is NOT always correct, but should mostly work
# in case $str is encoded in utf-8 bytes or wide chars. This is a
# necessary workaround since we're rendering someone else's code which
# we can't enforce string encodings.

sub encode_html {
    my $str = shift;
    $str =~ s/([^\x00-\x21\x23-\x25\x28-\x3b\x3d\x3f-\xff])/$enc{$1} || '&#' . ord($1) . ';' /ge;
    utf8::downgrade($str);