foreach my $file (@testitems) {
my $clean = 1;
- my $doc = PPI::Document->new($file) or do {
+ my $source;
+ {
+ # due to a bug in PPI it cannot determine the encoding of a source file by
+ # use utf8; normaly this would be no problem but some people instist on
+ # putting strange stuff into the source. as a workaround read in the source
+ # with :utf8 layer and pass it to PPI by reference
+ # there are still some latin chars, but it's not the purpose of this test
+ # to find them, so warnings about it will be ignored
+ local $^W = 0; # don't care about invalid chars in comments
+ local $/ = undef;
+ open my $fh, '<:utf8', $file or die $!;
+ $source = <$fh>;
+ }
+
+ my $doc = PPI::Document->new(\$source) or do {
print $fh "?: PPI error for file $file: " . PPI::Document::errstr() . "\n";
ok 0, $file;
next;