[PATCH 06/12] v2: generate better Message-IDs for duplicates

["Eric Wong (Contractor, The Linux Foundation)" <e@80x24.org>]

While hunting duplicates, I noticed a leading '-' in some
Message-IDs as a result of RFC4648 encoding.  While '-' seems
allowed by RFC5322 and URL-friendly (RFC4648), they are uncommon
and make using Message-IDs as arguments for command-line tools
more difficult.  So prefix them with a datestamp to at least
give readers some sense of the age.  And shorten the "localhost"
hostname to "z" to save space.
---
 MANIFEST                      |  1 +
 lib/PublicInbox/Import.pm     | 18 ++++++------
 lib/PublicInbox/V2Writable.pm |  6 ++--
 scripts/dupe-finder           | 54 +++++++++++++++++++++++++++++++++++
 t/v2writable.t                |  5 ++--
 5 files changed, 71 insertions(+), 13 deletions(-)
 create mode 100644 scripts/dupe-finder

diff --git a/MANIFEST b/MANIFEST
index 58b3634..00a0970 100644
--- a/MANIFEST
+++ b/MANIFEST
@@ -124,6 +124,7 @@ script/public-inbox-watch
 script/public-inbox.cgi
 scripts/dc-dlvr
 scripts/dc-dlvr.pre
+scripts/dupe-finder
 scripts/edit-sa-prefs
 scripts/import_maildir
 scripts/import_slrnspool
diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index 9e8900f..c7a96e1 100644
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -14,6 +14,7 @@ use PublicInbox::Address;
 use PublicInbox::MsgTime qw(msg_timestamp msg_datestamp);
 use PublicInbox::ContentId qw(content_digest);
 use PublicInbox::MDA;
+use POSIX qw(strftime);
 
 sub new {
 	my ($class, $git, $name, $email, $ibx) = @_;
@@ -330,7 +331,7 @@ sub v1_mid0 ($) {
 	my $mids = mids($hdr);
 
 	if (!scalar(@$mids)) { # spam often has no Message-Id
-		my $mid0 = digest2mid(content_digest($mime));
+		my $mid0 = digest2mid(content_digest($mime), $hdr);
 		append_mid($hdr, $mid0);
 		return $mid0;
 	}
@@ -445,18 +446,19 @@ sub atfork_child {
 	}
 }
 
-sub digest2mid ($) {
-	my ($dig) = @_;
+sub digest2mid ($$) {
+	my ($dig, $hdr) = @_;
 	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';
+	# Add a date prefix to prevent a leading '-' in case that trips
+	# up some tools (e.g. if a Message-ID were a expected as a
+	# command-line arg)
+	my $dt = msg_datestamp($hdr);
+	$dt = POSIX::strftime('%Y%m%d%H%M%S', gmtime($dt));
+	"$dt.$b64" . '@z';
 }
 
 sub clean_purge_buffer {
diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm
index 66f8a8a..0dcdeda 100644
--- a/lib/PublicInbox/V2Writable.pm
+++ b/lib/PublicInbox/V2Writable.pm
@@ -174,19 +174,19 @@ sub num_for_harder {
 
 	my $hdr = $mime->header_obj;
 	my $dig = content_digest($mime);
-	$$mid0 = PublicInbox::Import::digest2mid($dig);
+	$$mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
 	my $num = $self->{mm}->mid_insert($$mid0);
 	unless (defined $num) {
 		# it's hard to spoof the last Received: header
 		my @recvd = $hdr->header_raw('Received');
 		$dig->add("Received: $_") foreach (@recvd);
-		$$mid0 = PublicInbox::Import::digest2mid($dig);
+		$$mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
 		$num = $self->{mm}->mid_insert($$mid0);
 
 		# fall back to a random Message-ID and give up determinism:
 		until (defined($num)) {
 			$dig->add(rand);
-			$$mid0 = PublicInbox::Import::digest2mid($dig);
+			$$mid0 = PublicInbox::Import::digest2mid($dig, $hdr);
 			warn "using random Message-ID <$$mid0> as fallback\n";
 			$num = $self->{mm}->mid_insert($$mid0);
 		}
diff --git a/scripts/dupe-finder b/scripts/dupe-finder
new file mode 100644
index 0000000..1402237
--- /dev/null
+++ b/scripts/dupe-finder
@@ -0,0 +1,54 @@
+#!/usr/bin/perl -w
+# Copyright (C) 2018 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
+#
+# ad-hoc tool for finding duplicates, unstable!
+use strict;
+use warnings;
+use PublicInbox::Inbox;
+use PublicInbox::Over;
+use PublicInbox::Search;
+use PublicInbox::Config;
+my $repo = shift;
+my $ibx;
+if (index($repo, '@') > 0) {
+	$ibx = PublicInbox::Config->new->lookup($repo);
+} elsif (-d $repo) {
+	$ibx = { mainrepo => $repo, address => 'unnamed@example.com' };
+	$ibx = PublicInbox::Inbox->new($ibx);
+} else {
+	$ibx = PublicInbox::Config->new->lookup_name($repo);
+}
+$ibx or die "No inbox";
+$ibx->search or die "search not available for inbox";
+my $dbh = $ibx->search->{over_ro}->connect;
+my $over = PublicInbox::Over->new($dbh->sqlite_db_filename);
+
+sub emit ($) {
+	my ($nums) = @_;
+	foreach my $n (@$nums) {
+		my $smsg = $over->get_art($n) or next;
+		print STDERR "$n $smsg->{blob} $smsg->{mid}\n";
+		my $msg = $ibx->msg_by_smsg($smsg) or next;
+		print "From $smsg->{blob}\@$n Thu Jan  1 00:00:00 1970\n";
+		$$msg =~ s/^(>*From )/>$1/gm;
+		print $$msg, "\n";
+	}
+}
+
+my $sth = $dbh->prepare(<<'');
+SELECT id,num FROM id2num WHERE num > 0 ORDER BY id
+
+$sth->execute;
+my $prev_id = -1;
+my ($id, $num, @nums);
+while (1) {
+	($id, $num) = $sth->fetchrow_array;
+	defined $id or last;
+	if ($prev_id != $id) {
+		emit(\@nums) if scalar(@nums) > 1;
+		@nums = ();
+	}
+	$prev_id = $id;
+	push @nums, $num;
+}
diff --git a/t/v2writable.t b/t/v2writable.t
index 85fb6a6..d37fb06 100644
--- a/t/v2writable.t
+++ b/t/v2writable.t
@@ -68,7 +68,7 @@ if ('ensure git configs are correct') {
 		[ $sec->header_obj->header_raw('Message-Id') ],
 		'no new Message-Id added');
 
-	my $sane_mid = qr/\A<[\w\-]+\@localhost>\z/;
+	my $sane_mid = qr/\A<[\w\-\.]+\@\w+>\z/;
 	@warn = ();
 	$mime->header_set('Message-Id', '<a-mid@b>');
 	$mime->body_set('different');
@@ -82,7 +82,8 @@ if ('ensure git configs are correct') {
 	@warn = ();
 	$mime->header_set('Message-Id', '<a-mid@b>');
 	$mime->body_set('this one needs a random mid');
-	my $gen = PublicInbox::Import::digest2mid(content_digest($mime));
+	my $hdr = $mime->header_obj;
+	my $gen = PublicInbox::Import::digest2mid(content_digest($mime), $hdr);
 	unlike($gen, qr![\+/=]!, 'no URL-unfriendly chars in Message-Id');
 	my $fake = PublicInbox::MIME->new($mime->as_string);
 	$fake->header_set('Message-Id', "<$gen>");
-- 
EW