From @hakonhagland

( Note​: I first tried to send this as a followup to [perl #88592], but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id 88592" )

This bug​:

( https://rt.cpan.org/Public/Bug/Display.html?id=88592 )

is still present in my perl 5.22.1. From the last post in the ticket it
seems like it was fixed with a pull request to the Encode module in october
2014...

$ perl -v
This is perl 5, version 22, subversion 1 (v5.22.1) built for
x86_64-linux-gnu-thread-multi
(with 58 registered patches, see perl -V for more detail)

Here is a test script​:

use open qw( :std :utf8 );
use strict;
use warnings;
no warnings 'utf8';
my $bytes = "Test\x{E5}End";
my $fn = 'test.txt';
open ( my $fh, '>​:raw', $fn ) or die "Could not open file '$fn'​: $!";
print $fh $bytes;
close $fh;
open ( $fh, "<​:encoding(utf-8)", $fn ) or die "Could not open file '$fn'​:
$!";
my $str = do { local $/; <$fh> };
close $fh;

the output on my machine (x86_64 GNU/Linux)​:

utf8 "\xE5" does not map to Unicode at ./p.pl line 11.

From @jkeenan

On Sun Jul 31 14​:31​:05 2016, hakon.hagland@​gmail.com wrote​:

( Note​: I first tried to send this as a followup to [perl #88592], but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id 88592" )

This bug​:

( https://rt.cpan.org/Public/Bug/Display.html?id=88592 )

is still present in my perl 5.22.1. From the last post in the ticket it
seems like it was fixed with a pull request to the Encode module in october
2014...

$ perl -v
This is perl 5, version 22, subversion 1 (v5.22.1) built for
x86_64-linux-gnu-thread-multi
(with 58 registered patches, see perl -V for more detail)

Here is a test script​:

use open qw( :std :utf8 );
use strict;
use warnings;
no warnings 'utf8';
my $bytes = "Test\x{E5}End";
my $fn = 'test.txt';
open ( my $fh, '>​:raw', $fn ) or die "Could not open file '$fn'​: $!";
print $fh $bytes;
close $fh;
open ( $fh, "<​:encoding(utf-8)", $fn ) or die "Could not open file '$fn'​:
$!";
my $str = do { local $/; <$fh> };
close $fh;

the output on my machine (x86_64 GNU/Linux)​:

utf8 "\xE5" does not map to Unicode at ./p.pl line 11.

Reproduced in blead at commit ec7784b on Ubuntu Linux 16.04 LTS.

--
James E Keenan (jkeenan@​cpan.org)

The RT System itself - Status changed from 'new' to 'open'

From zefram@fysh.org

Hakon Haegland wrote​:

( Note​: I first tried to send this as a followup to [perl #88592], but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id 88592" )

Probably because it's not a Perl 6 bug. That email address looks like
it's specific to Perl 6.

-zefram

From alex@chmrr.net

On Mon, 1 Aug 2016 07​:44​:26 +0100
Zefram <zefram@​fysh.org> wrote​:

Hakon Haegland wrote​:

( Note​: I first tried to send this as a followup to [perl #88592], but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id 88592" )

Probably because it's not a Perl 6 bug. That email address looks like
it's specific to Perl 6.

That email address works with anything in rt.perl.org, which
includes both perl5 and perl6 -- but the bug you're referencing is on
rt.cpan.org, which has its own bug numbers. You'd have needed to
email bug-Encode@​rt.cpan.org.

But I'm also the person who pops up halfway into that ticket. The test
I provided, and the solution it arrived at, looks to not have affected
the `open(..., "<​:encoding(utf-8)") codepath, only the Encode​::encode
ones. So the latest Encode does not actually resolve that bug.
- Alex

[1] dankogai/p5-encode@07c8adb5

From @xsawyerx

On Mon Aug 01 00​:37​:06 2016, chmrr wrote​:

On Mon, 1 Aug 2016 07​:44​:26 +0100
Zefram <zefram@​fysh.org> wrote​:

Hakon Haegland wrote​:

( Note​: I first tried to send this as a followup to [perl #88592],
but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id
88592" )

Probably because it's not a Perl 6 bug. That email address looks
like
it's specific to Perl 6.

That email address works with anything in rt.perl.org, which
includes both perl5 and perl6 -- but the bug you're referencing is on
rt.cpan.org, which has its own bug numbers. You'd have needed to
email bug-Encode@​rt.cpan.org.

Considering this is an upstream CPAN module, I think we should close this.

From @cpansprout

On Mon Aug 01 08​:06​:00 2016, xsawyerx@​cpan.org wrote​:

On Mon Aug 01 00​:37​:06 2016, chmrr wrote​:

On Mon, 1 Aug 2016 07​:44​:26 +0100
Zefram <zefram@​fysh.org> wrote​:

Hakon Haegland wrote​:

( Note​: I first tried to send this as a followup to [perl #88592],
but I
got reply from
bugs-correspond@​bugs6.perl.org : "Could not find a ticket with id
88592" )

Probably because it's not a Perl 6 bug. That email address looks
like
it's specific to Perl 6.

That email address works with anything in rt.perl.org, which
includes both perl5 and perl6 -- but the bug you're referencing is on
rt.cpan.org, which has its own bug numbers. You'd have needed to
email bug-Encode@​rt.cpan.org.

Considering this is an upstream CPAN module, I think we should close this.

I have confirmed that it is a problem with Encode and have just opened <https://rt.cpan.org/Ticket/Display.html?id=116629>.

--

Father Chrysostomos

@cpansprout - Status changed from 'open' to 'rejected'

From @cpansprout

On Mon Aug 01 08​:49​:14 2016, sprout wrote​:

On Mon Aug 01 08​:06​:00 2016, xsawyerx@​cpan.org wrote​:

Considering this is an upstream CPAN module, I think we should close
this.

I have confirmed that it is a problem with Encode and have just opened
<https://rt.cpan.org/Ticket/Display.html?id=116629>.

Oh, wait. I had my one-line diagnostic patch still applied to my local perl repo and ran the tests. I got these failures​:

# Failed test 'FB_WARN message utf8'
# at t/fallback.t line 110.
# ''
# doesn't match '(?^​:does not map to Unicode)'

# Failed test 'WARN_ON_ERR message ascii'
# at t/fallback.t line 124.
# ''
# doesn't match '(?^​:does not map to Unicode)'
# Looks like you failed 2 tests of 50.

which suggests to me that PerlIO​::encoding (which is core-only) may well be asking for warnings without checking hints.

I think this is something that needs looking into, so I am reopening the ticket.

--

Father Chrysostomos

@cpansprout - Status changed from 'rejected' to 'open'