about summary refs log tree commit homepage
path: root/lib/PublicInbox/MID.pm
diff options
context:
space:
mode:
Diffstat (limited to 'lib/PublicInbox/MID.pm')
-rw-r--r--lib/PublicInbox/MID.pm12
1 files changed, 10 insertions, 2 deletions
diff --git a/lib/PublicInbox/MID.pm b/lib/PublicInbox/MID.pm
index 78952b95..1c2d75cc 100644
--- a/lib/PublicInbox/MID.pm
+++ b/lib/PublicInbox/MID.pm
@@ -6,7 +6,8 @@ package PublicInbox::MID;
 use strict;
 use warnings;
 use base qw/Exporter/;
-our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime/;
+our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime mid_escape/;
+use URI::Escape qw(uri_escape_utf8);
 use Digest::SHA qw/sha1_hex/;
 use constant MID_MAX => 40; # SHA-1 hex length
 
@@ -25,6 +26,7 @@ sub id_compress {
         my ($id, $force) = @_;
 
         if ($force || $id =~ /[^\w\-]/ || length($id) > MID_MAX) {
+                utf8::encode($id);
                 return sha1_hex($id);
         }
         $id;
@@ -36,7 +38,9 @@ sub mid2path {
 
         unless (defined $x38) {
                 # compatibility with old links (or short Message-IDs :)
-                $mid = sha1_hex(mid_clean($mid));
+                $mid = mid_clean($mid);
+                utf8::encode($mid);
+                $mid = sha1_hex($mid);
                 ($x2, $x38) = ($mid =~ /\A([a-f0-9]{2})([a-f0-9]{38})\z/);
         }
         "$x2/$x38";
@@ -44,4 +48,8 @@ sub mid2path {
 
 sub mid_mime ($) { $_[0]->header_obj->header_raw('Message-ID') }
 
+# RFC3986, section 3.3:
+sub MID_ESC () { '^A-Za-z0-9\-\._~!\$\&\';\(\)\*\+,;=:@' }
+sub mid_escape ($) { uri_escape_utf8($_[0], MID_ESC) }
+
 1;