[PATCH] www: diff: fix encoding problems when showing diff

[Eric Wong <e@80x24.org>]

We need to use the utf8 layer when writing files to be diffed,
and utf8::decode the `git diff' output.  Furthermore, do the
CRLF > LF conversion early to avoid showing CRLF vs LF
differences in the diff, since that doesn't matter to MUAs
(nor our normal HTML views)
---
 lib/PublicInbox/MailDiff.pm |  8 +++++---
 t/psgi_v2.t                 | 40 ++++++++++++++++++++++++++++++++++++-
 2 files changed, 44 insertions(+), 4 deletions(-)

diff --git a/lib/PublicInbox/MailDiff.pm b/lib/PublicInbox/MailDiff.pm
index a0ecef9f..7511144c 100644
--- a/lib/PublicInbox/MailDiff.pm
+++ b/lib/PublicInbox/MailDiff.pm
@@ -19,8 +19,10 @@ sub write_part { # Eml->each_part callback
 	my $ct = $part->content_type || 'text/plain';
 	my ($s, $err) = msg_part_text($part, $ct);
 	my $sfx = defined($s) ? 'txt' : 'bin';
-	open my $fh, '>', "$self->{curdir}/$idx.$sfx" or die "open: $!";
-	print $fh ($s // $part->body) or die "print $!";
+	$s //= $part->body;
+	$s =~ s/\r+\n/\n/sg;
+	open my $fh, '>:utf8', "$self->{curdir}/$idx.$sfx" or die "open: $!";
+	print $fh $s or die "print $!";
 	close $fh or die "close $!";
 }
 
@@ -66,9 +68,9 @@ sub next_smsg ($) {
 sub emit_msg_diff {
 	my ($bref, $self) = @_; # bref is `git diff' output
 	# will be escaped to `•' in HTML
+	utf8::decode($$bref);
 	$self->{ctx}->{ibx}->{obfuscate} and
 		obfuscate_addrs($self->{ctx}->{ibx}, $$bref, "\x{2022}");
-	$$bref =~ s/\r+\n/\n/sg;
 	print { $self->{ctx}->{zfh} } '</pre><hr><pre>' if $self->{nr} > 1;
 	flush_diff($self->{ctx}, $bref);
 	next_smsg($self);
diff --git a/t/psgi_v2.t b/t/psgi_v2.t
index f709c3c7..5b197a9f 100644
--- a/t/psgi_v2.t
+++ b/t/psgi_v2.t
@@ -13,6 +13,36 @@ require_mods(qw(DBD::SQLite Search::Xapian HTTP::Request::Common Plack::Test
 use_ok($_) for (qw(HTTP::Request::Common Plack::Test));
 use_ok 'PublicInbox::WWW';
 my ($tmpdir, $for_destroy) = tmpdir();
+my $enc_dup = 'ref-20150309094050.GO3427@x1.example';
+
+my $dibx = create_inbox 'v2-dup', version => 2, indexlevel => 'medium',
+			tmpdir => "$tmpdir/dup", sub {
+	my ($im, $ibx) = @_;
+	my $common = <<"";
+Date: Mon, 9 Mar 2015 09:40:50 +0000
+From: x\@example.com
+To: y\@example.com
+Subject: re
+Message-ID: <$enc_dup>
+MIME-Version: 1.0
+
+	$im->add(PublicInbox::Eml->new($common.<<EOM)) or BAIL_OUT;
+Content-Type: text/plain; charset=utf-8
+Content-Disposition: inline
+Content-Transfer-Encoding: 8bit
+
+cr_mismatch
+pipe \x{e2}\x{94}\x{82} or not
+EOM
+	$im->add(PublicInbox::Eml->new($common.<<EOM)) or BAIL_OUT;
+Content-Type: text/plain; charset="windows-1252"
+Content-Transfer-Encoding: quoted-printable
+
+cr_mismatch\r
+pipe =E2=94=82 or not
+EOM
+};
+
 my $eml = PublicInbox::Eml->new(<<'EOF');
 From oldbug-pre-a0c07cba0e5d8b6a Fri Oct  2 00:00:00 1993
 From: a@example.com
@@ -53,6 +83,9 @@ my $cfgpath = "$ibx->{inboxdir}/pi_config";
 [publicinbox "v2test"]
 	inboxdir = $ibx->{inboxdir}
 	address = $ibx->{-primary_address}
+[publicinbox "dup"]
+	inboxdir = $dibx->{inboxdir}
+	address = $dibx->{-primary_address}
 EOF
 	close $fh or BAIL_OUT;
 }
@@ -221,8 +254,13 @@ my $client1 = sub {
 	}
 	like($raw, qr/\b3\+ messages\b/, 'thread overview shown');
 
-	$res = $cb->(GET('/v2test/a-mid@b/d/'));
+	$res = $cb->(GET("/dup/$enc_dup/d/"));
 	is($res->code, 200, '/d/ (diff) endpoint works');
+	$raw = $res->content;
+	like($raw, qr!</span> cr_mismatch\n!s,
+		'cr_mismatch is only diff context');
+	like($raw, qr!>\-pipe !s, 'pipe diff del line');
+	like($raw, qr!>\+pipe !s, 'pipe diff ins line');
 };
 
 test_psgi(sub { $www->call(@_) }, $client1);