mid: add `mids' and `references' methods for extraction

We'll be using a more consistent API for extracting Message-IDs
from various headers.

1 files changed, 23 insertions, 1 deletions

diff --git a/lib/PublicInbox/MID.pm b/lib/PublicInbox/MID.pm
index 2c9822f4..786c056f 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 mid_escape MID_ESC/;
+our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime mid_escape MID_ESC
+        mids references/;
 use URI::Escape qw(uri_escape_utf8);
 use Digest::SHA qw/sha1_hex/;
 use constant MID_MAX => 40; # SHA-1 hex length
@@ -48,6 +49,27 @@ sub mid2path {
 
 sub mid_mime ($) { $_[0]->header_obj->header_raw('Message-ID') }
 
+sub uniq_mids {
+        my ($hdr, @fields) = @_;
+        my %seen;
+        my @raw;
+        foreach my $f (@fields) {
+                push @raw, $hdr->header_raw($f);
+        }
+        my @mids = (join(' ', @raw) =~ /<([^>]+)>/g);
+        my $mids = scalar(@mids) == 0 ? \@raw: \@mids;
+        my @ret;
+        foreach (@$mids) {
+                next if $seen{$_};
+                push @ret, $_;
+                $seen{$_} = 1;
+        }
+        \@ret;
+}
+
+sub mids { uniq_mids($_[0], 'Message-Id') }
+sub references { uniq_mids($_[0], 'References', 'In-Reply-To') }
+
 # RFC3986, section 3.3:
 sub MID_ESC () { '^A-Za-z0-9\-\._~!\$\&\';\(\)\*\+,;=:@' }
 sub mid_escape ($) { uri_escape_utf8($_[0], MID_ESC) }