import: switch to URL-safe Base64 for Message-IDs

Hexdigests are too long and shorter Message-IDs are easier
to deal with.

diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index 4c007b610dc3fc58760a91b815b1ab5405f58c1e..77e74c13af6cf66797685f770838bb6ab1995a87 100644 (file)
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -401,7 +401,16 @@ sub atfork_child {
 
 sub digest2mid ($) {
        my ($dig) = @_;
-       $dig->clone->hexdigest . '@localhost';
+       my $b64 = $dig->clone->b64digest;
+       # Make our own URLs nicer:
+       # See "Base 64 Encoding with URL and Filename Safe Alphabet" in RFC4648
+       $b64 =~ tr!+/=!-_!d;
+
+       # We can make this more meaningful with a date prefix or other things,
+       # but this is only needed for crap that fails to generate a Message-ID
+       # or reuses one.  In other words, it's usually spammers who hit this
+       # so they don't deserve nice Message-IDs :P
+       $b64 . '@localhost';
 }
 
 1;

diff --git a/t/v2writable.t b/t/v2writable.t
index c6bcefd4f0e1833a0432211177b1e40a0815818b..bbe6d14b7fecf3e7720f3e37fe1835c65430fc87 100644 (file)
--- a/t/v2writable.t
+++ b/t/v2writable.t
@@ -68,6 +68,7 @@ if ('ensure git configs are correct') {
                [ $mime->header_obj->header_raw('Message-Id') ],
                'no new Message-Id added');
 
+       my $sane_mid = qr/\A<[\w\-]+\@localhost>\z/;
        @warn = ();
        $mime->header_set('Message-Id', '<a-mid@b>');
        $mime->body_set('different');
@@ -75,13 +76,14 @@ if ('ensure git configs are correct') {
        like(join(' ', @warn), qr/reused/, 'warned about reused MID');
        my @mids = $mime->header_obj->header_raw('Message-Id');
        is($mids[1], '<a-mid@b>', 'original mid not changed');
-       like($mids[0], qr/\A<\w+\@localhost>\z/, 'new MID added');
+       like($mids[0], $sane_mid, 'new MID added');
        is(scalar(@mids), 2, 'only one new MID added');
 
        @warn = ();
        $mime->header_set('Message-Id', '<a-mid@b>');
        $mime->body_set('this one needs a random mid');
-       my $gen = content_digest($mime)->hexdigest . '@localhost';
+       my $gen = PublicInbox::Import::digest2mid(content_digest($mime));
+       unlike($gen, qr![\+/=]!, 'no URL-unfriendly chars in Message-Id');
        my $fake = PublicInbox::MIME->new($mime->as_string);
        $fake->header_set('Message-Id', $gen);
        ok($im->add($fake), 'fake added easily');
@@ -90,14 +92,14 @@ if ('ensure git configs are correct') {
        like(join(' ', @warn), qr/using random/, 'warned about using random');
        @mids = $mime->header_obj->header_raw('Message-Id');
        is($mids[1], '<a-mid@b>', 'original mid not changed');
-       like($mids[0], qr/\A<\w+\@localhost>\z/, 'new MID added');
+       like($mids[0], $sane_mid, 'new MID added');
        is(scalar(@mids), 2, 'only one new MID added');
 
        @warn = ();
        $mime->header_set('Message-Id');
        ok($im->add($mime), 'random MID made for MID free message');
        @mids = $mime->header_obj->header_raw('Message-Id');
-       like($mids[0], qr/\A<\w+\@localhost>\z/, 'mid was generated');
+       like($mids[0], $sane_mid, 'mid was generated');
        is(scalar(@mids), 1, 'new generated');
 }