hval: to_attr: support wide characters

We need to escape wide characters when making attribute names from
filename-looking things in diffstats.

3 files changed, 15 insertions, 1 deletions

diff --git a/lib/PublicInbox/Hval.pm b/lib/PublicInbox/Hval.pm
index 7e007027..39256ee0 100644
--- a/lib/PublicInbox/Hval.pm
+++ b/lib/PublicInbox/Hval.pm
@@ -139,10 +139,12 @@ sub to_attr ($) {
         return if index($str, '//') >= 0;
 
         my $first = '';
+        utf8::encode($str); # to octets
         if ($str =~ s/\A([^A-Ya-z])//ms) { # start with a letter
                   $first = sprintf('Z%02x', ord($1));
         }
         $str =~ s/([^A-Za-z0-9_\.\-])/$ESCAPES{$1}/egms;
+        utf8::decode($str); # allow wide chars
         $first . $str;
 }
 
@@ -155,6 +157,7 @@ sub from_attr ($) {
         }
         $str =~ s!::([a-f0-9]{2})!chr(hex($1))!egms;
         $str =~ tr!:!/!;
+        utf8::decode($str);
         $first . $str;
 }
 
diff --git a/t/solve/0001-simple-mod.patch b/t/solve/0001-simple-mod.patch
index c6bb1575..c55fe310 100644
--- a/t/solve/0001-simple-mod.patch
+++ b/t/solve/0001-simple-mod.patch
@@ -3,9 +3,11 @@ To: meta@public-inbox.org
 Subject: [PATCH] TODO: take expert web design advice
 Date: Mon, 1 Apr 2019 08:15:20 +0000
 Message-Id: <20190401081523.16213-1-BOFH@YHBT.net>
+Content-Type: text/plain; charset=utf-8
 
 ---
  TODO | 2 ++
+ Ω    | 5 --
  1 file changed, 2 insertions(+)
 
 diff --git a/TODO b/TODO
diff --git a/t/solver_git.t b/t/solver_git.t
index 92402c3a..92c07334 100644
--- a/t/solver_git.t
+++ b/t/solver_git.t
@@ -154,7 +154,16 @@ EOF
         my $non_existent = 'ee5e32211bf62ab6531bdf39b84b6920d0b6775a';
         my $client = sub {
                 my ($cb) = @_;
-                my $res = $cb->(GET("/$name/3435775/s/"));
+                my $mid = '20190401081523.16213-1-BOFH@YHBT.net';
+                my @warn;
+                my $res = do {
+                        local $SIG{__WARN__} = sub { push @warn, @_ };
+                        $cb->(GET("/$name/$mid/"));
+                };
+                is_deeply(\@warn, [], 'no warnings from rendering diff');
+                like($res->content, qr!>&#937;</a>!, 'omega escaped');
+
+                $res = $cb->(GET("/$name/3435775/s/"));
                 is($res->code, 200, 'success with existing blob');
 
                 $res = $cb->(GET("/$name/".('0'x40).'/s/'));