use raw header for Message-ID

[public-inbox.git] / lib / PublicInbox / MID.pm

diff --git a/lib/PublicInbox/MID.pm b/lib/PublicInbox/MID.pm
index 1b03fdce72e1333132f0884371754b290ac6ed24..78952b95e02f1eb94a6876017bc6c8e6a911978c 100644 (file)
--- a/lib/PublicInbox/MID.pm
+++ b/lib/PublicInbox/MID.pm
@@ -1,10 +1,12 @@
-# Copyright (C) 2015, all contributors <meta@public-inbox.org>
+# Copyright (C) 2015 all contributors <meta@public-inbox.org>
 # License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+#
+# Various Message-ID-related functions.
 package PublicInbox::MID;
 use strict;
 use warnings;
 use base qw/Exporter/;
-our @EXPORT_OK = qw/mid_clean mid_compress mid2path/;
+our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime/;
 use Digest::SHA qw/sha1_hex/;
 use constant MID_MAX => 40; # SHA-1 hex length
 
@@ -19,18 +21,13 @@ sub mid_clean {
 }
 
 # this is idempotent
-sub mid_compress {
-       my ($mid, $force) = @_;
+sub id_compress {
+       my ($id, $force) = @_;
 
-       # XXX dirty hack! FIXME!
-       # Some HTTP servers (apache2 2.2.22-13+deb7u5 on my system)
-       # apparently do not handle "%25" in the URL path component correctly.
-       # I'm not yet sure if it's something weird with my rewrite rules
-       # or what; will need to debug...
-       return sha1_hex($mid) if (index($mid, '%') >= 0);
-
-       return $mid if (!$force && length($mid) <= MID_MAX);
-       sha1_hex($mid);
+       if ($force || $id =~ /[^\w\-]/ || length($id) > MID_MAX) {
+               return sha1_hex($id);
+       }
+       $id;
 }
 
 sub mid2path {
@@ -45,4 +42,6 @@ sub mid2path {
        "$x2/$x38";
 }
 
+sub mid_mime ($) { $_[0]->header_obj->header_raw('Message-ID') }
+
 1;