X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FContentId.pm;h=9d9be4179b988ad518d19d7a32eff24b937cccca;hb=9bd675d33ad1e49bd2ebe12a1d216216e61380de;hp=9082b7694002168467ab3d2610a36631348116d2;hpb=dd83a896a7eb718dcd49560d0aab967f1f481883;p=public-inbox.git
diff --git a/lib/PublicInbox/ContentId.pm b/lib/PublicInbox/ContentId.pm
index 9082b769..9d9be417 100644
--- a/lib/PublicInbox/ContentId.pm
+++ b/lib/PublicInbox/ContentId.pm
@@ -1,16 +1,30 @@
-# Copyright (C) 2018 all contributors
+# Copyright (C) 2018-2019 all contributors
# License: AGPL-3.0+
+# Unstable internal API.
+# Used for on-the-fly duplicate detection in V2 inboxes.
+# This is not stored in any database anywhere and may change
+# as changes in duplicate detection are needed.
+# See L manpage for more details.
package PublicInbox::ContentId;
use strict;
use warnings;
use base qw/Exporter/;
our @EXPORT_OK = qw/content_id content_digest/;
use PublicInbox::MID qw(mids references);
+use PublicInbox::MsgIter;
# not sure if less-widely supported hash families are worth bothering with
use Digest::SHA;
+sub digest_addr ($$$) {
+ my ($dig, $h, $v) = @_;
+ $v =~ tr/"//d;
+ $v =~ s/@([a-z0-9\_\.\-\(\)]*([A-Z])\S*)/'@'.lc($1)/ge;
+ utf8::encode($v);
+ $dig->add("$h\0$v\0");
+}
+
sub content_digest ($) {
my ($mime) = @_;
my $dig = Digest::SHA->new(256);
@@ -21,29 +35,61 @@ sub content_digest ($) {
# in SearchIdx, so treat them the same for this:
my %seen;
foreach my $mid (@{mids($hdr)}) {
- $dig->add('mid: '.$mid);
+ # do NOT consider the Message-ID as part of the content_id
+ # if we got here, we've already got Message-ID reuse
$seen{$mid} = 1;
}
foreach my $mid (@{references($hdr)}) {
next if $seen{$mid};
- $dig->add('ref: '.$mid);
+ $dig->add("ref\0$mid\0");
}
# Only use Sender: if From is not present
foreach my $h (qw(From Sender)) {
- my @v = $hdr->header_raw($h);
+ my @v = $hdr->header($h);
if (@v) {
- $dig->add("$h: $_") foreach @v;
- last;
+ digest_addr($dig, $h, $_) foreach @v;
}
}
-
- # Content-* headers are often no-ops, so maybe we don't need them
- foreach my $h (qw(Subject Date To Cc)) {
- my @v = $hdr->header_raw($h);
- $dig->add("$h: $_") foreach @v;
+ foreach my $h (qw(Subject Date)) {
+ my @v = $hdr->header($h);
+ foreach my $v (@v) {
+ utf8::encode($v);
+ $dig->add("$h\0$v\0");
+ }
}
- $dig->add($mime->body_raw);
+ # Some mail processors will add " to unquoted names that were
+ # not in the original message. For the purposes of deduplication,
+ # do not take it into account:
+ foreach my $h (qw(To Cc)) {
+ my @v = $hdr->header($h);
+ digest_addr($dig, $h, $_) foreach @v;
+ }
+ msg_iter($mime, sub {
+ my ($part, $depth, @idx) = @{$_[0]};
+ $dig->add("\0$depth:".join('.', @idx)."\0");
+ my $fn = $part->filename;
+ if (defined $fn) {
+ utf8::encode($fn);
+ $dig->add("fn\0$fn\0");
+ }
+ my @d = $part->header('Content-Description');
+ foreach my $d (@d) {
+ utf8::encode($d);
+ $dig->add("d\0$d\0");
+ }
+ $dig->add("b\0");
+ my $ct = $part->content_type || 'text/plain';
+ my ($s, undef) = msg_part_text($part, $ct);
+ if (defined $s) {
+ $s =~ s/\r\n/\n/gs;
+ $s =~ s/\s*\z//s;
+ utf8::encode($s);
+ } else {
+ $s = $part->body;
+ }
+ $dig->add($s);
+ });
$dig;
}